All right. Thank you everybody for joining us, both in here and online. We have a great day of content today. I hope you all find it useful. Before I kick off the program today, I'm going to read this in excruciating detail and then go right to the agenda. This is our lineup for today. One procedural thing before I hand the mic over to Jeff. We will take a few questions during the customer panel. We prefer to handle all Q&A at the end when we'll have mic runners in the room and we'll have all the execs up on stage. With that, let me turn it over to Jeff.
Great. Thank you, Greg. Hey, everybody. Great to see many of you again. Great to meet some of you today. Thank you all for coming out and spending your afternoon with Twilio. I see it as this great opportunity for us to tell you more about Twilio, help you understand what Twilio is all about, what we see as the long-term opportunity, and what it is that we're trying to accomplish, and how it is we're going to get there, and take questions. You'll get a chance to meet the team. A lot of great information sharing to go on today. Thank you for coming. We started Twilio nine years ago, almost 10 years ago. With this basic idea that communications was going to transform because of software, that software was going to fundamentally reconfigure how the world communicates and reconfigure how this industry works.
We came to that conclusion because of I, myself, as a software developer, had always needed communications. I'm a serial entrepreneur. I had started three companies before Twilio, and I was one of the first product managers at Amazon Web Services. What I realized was that at each of my previous companies, we were using the power of software to iterate quickly, to constantly listen to our customers and ship new versions of our products, and fundamentally build differentiation because of our ability to move fast and understand our customers. That's the superpower of software. Also in every single one of those companies, we had needed to communicate. We needed to talk to our customers, have a dialogue with them in order to build a great customer experience. Every time we needed that, we looked at ourselves and said, "Well, I'm a software developer.
Making a phone ring, that's magical to me. I don't know how to make that happen." We turned to the existing communications industry at the time, and we'd say, "We have this idea we want to build to how we can engage with our customers. How can we go about doing that?" Every time we got this answer that was very unfulfilling. It was like, "We're going to come in and we're going to bring in and integrate all these different esoteric technologies together, and that's going to cost millions of dollars, and it's going to take years to actually get it in front of your customers." Every time this happened, I remember thinking, "This is insane." Everything we do in software is measured in sprints. It's measured in weeks. Customer feedback, always shipping.
In communications, things take years and millions of dollars before a customer can ever touch a solution or give you feedback and tell you if you're going in the right direction. That seemed crazy to me. Kind of realized that this world of communications is just diametrically opposed to that ethos of software. In 2008, Evan, John, and I started Twilio to change this, to migrate communications out of its legacy, which was in hardware and physical networks, and migrate it into its future, which is software. We told you at the roadshow 18 months ago when we went public. We told you the story about how in order to unlock this vision, we had to do a number of new things. We talked about how this is a new era of software, where the way companies are unlocking software value is actually fundamentally changing.
It used to be about selling software that ran on a server and selling you, licensing it to you on a per server basis. Then came SaaS, where it's about licensing seats to you. Now the new era of software is actually all about platforms and all about giving APIs to developers in order to build the next generation of software. This is the new era of software that's emerging, empowering those builders. We talked to you about a new category of communications that we had to invent in order to unlock this opportunity, that of a cloud communications platform, building APIs to give to developers to allow them to actually build all these ideas and to build this future of communications, which was going to be based in software. It's about a platform. It's not about an app.
We talked to you about a new model, a new business model that we had to create. This fundamentally more efficient developer first go to market, where we win the hearts and minds of developers, and they bring us into businesses of every shape and size. Now, 18 months later, I'm here to say we obviously keep learning and we learn some more things. First of all, the new software era, it's absolutely confirmed and it's only growing. You look at all the platform businesses are getting stronger and stronger every day, and the software ethos is permeating every kind of company out there. We talked about we had to create a new category. This one I think we were actually wrong about. I think we're going to create multiple new categories to fulfill on our mission of building Twilio and fueling the future of communications.
When I talked about a new model, a developer first go to market, it's working great. We're doubling down on it, and you're going to hear more about that later. That's the quick update. As I walk through what I wanted to talk about today, there's three things I want you to think about. First of all, this is day one of a huge opportunity to reshape communications, which is one of the world's largest markets. This is a developer-first approach that we're employing, and it drives both efficiency for how we're growing the company, as well as allows us to tap into a very large market, a broad market reach that we're able to touch with this approach. The last thing is how our leadership is being driven via software differentiation, and building trust in the market and trust with our customers.
Let's start with the first one, where we say this a lot. In fact, on the day of our IPO, every Twilio around the world wore a T-shirt that said it's Day one. Day one of a large opportunity to reshape communications. There is a large shift going on in pretty much every category of technology and every category of IT spending, and it's a shift in how software is actually delivered and how value is delivered to customers. It's not just about on-prem to cloud, like the SaaS people will tell you about, but it's also about apps to APIs, applications to platforms, because that's how you reconfigure very large markets.
If you look at how cloud platforms are reshaping almost all of the huge markets that are out there, every major category is getting reinvented through software and as building blocks to power the next generation of apps, whether it's Amazon Web Services with compute and storage, whether it's Google changing how people build maps, Twilio, of course, with communications, New Relic with analytics, and Stripe for payments. Every one of these categories is getting reshaped with a platform developer-first approach. These are the pillars of every app that we use, the key ingredients that go into every app, both mobile apps, web apps, B2B apps, every one of those apps. Cloud platforms are the way to tackle these large opportunities. The one difference between Twilio and all of these other markets is actually in size.
If you look at global IT spending, $3.5 trillion, nearly half of it is communications. 40% of all IT spending worldwide is spent on communications. An investment in Twilio is an investment in this idea that software is going to fundamentally reconfigure almost half of every IT dollar spent worldwide because of the power of software. If you look at how this money is spent today, how do companies traditionally spend this huge chunk of their budget? It's in this legacy approach, the one I described earlier that frustrated me before we started Twilio. You start off with the network layer. Use a vendor like AT&T or Verizon or someone like that. Then on top of that, you need to plug in some hardware. You bring in some hardware that's pretty inflexible.
The thing about the network and the hardware is they don't scale very easily. You got to keep adding more whenever you need to do more. They are geographic by nature. They sit in a place in the world, and the carriers are all geographically bounded, and the box sits in a closet somewhere. If you're going around the world, you got to recreate all this infrastructure everywhere in the world that you go. On top of that basic, those fundamentals, you want something that does something. You pile on an application of some variety. You bring in one of the big app vendors, and you put that on top, then the app needs to be personalized because it doesn't do what you want. You got to customize the hell out of it.
You bring in all these experts with one or more of these certification logos and these various stacks to come and integrate the whole thing together. You're like, this model just doesn't work. It's fragile, it's slow, and it's expensive. You look at it, these are huge market cap companies. AT&T, $180 billion. Cisco, $200 billion. Oracle, $200 billion. It seems like $200 billion is the magical number in this world. The communications part of this world, that value is going to migrate to somewhere because of the power of software. I'd love for it to be Twilio. It doesn't matter. You can change out the companies for different kinds of stacks. You can have different geographic components, even different types of applications inside the enterprise that they're trying to build, whether it's sales, whether it's support, whether it's marketing automation.
There's all these different use cases where communications is a part of it, but the model stays the same. How we've thought about this changing is over time. This is all going to convert into software. What we started doing is taking the bottom layers, the network and the hardware, and virtualizing it and moving it into the cloud. Taking something that was inflexible and expensive and slow and hard to scale and moving it into something that's virtual, is in the cloud, that scales elastically on demand, that works everywhere in the world. That's what virtualizing that does. You take that application layer.
The way we see it, this becomes application platforms that provide the flexibility that customers need, that allows them to actually take these applications and deploy them in a way that integrates with all the other systems and is very pliable, so they can change it to exactly how they want, because these are customer-touching systems, very important that companies are able to customize it to build their brand and build their customer experience the way they want it. You take those experts, the certified logos of the sort of small workforce of people who are experts in these esoteric technologies. What we do is we democratize that and make it all the developers in the world, 20 million developers in the world who know web and mobile technologies. They should be the ones that companies are turning to in order to build these applications.
By doing so, by virtualizing, by componentizing those applications, by democratizing the workforce, this is how we fulfill on our mission, the mission that we came up with, to fuel the future of communications. How do we block and tackle this big idea, this big future of communications that we're trying to fuel? Well, what we did is we came up with, a number of years ago, this master plan. Here's how we're going to do it. What we said is, step 1 is build a platform. Because a platform is the way you get exposure to the full breadth of opportunities that exist in a market as large as communications, which has so many different facets to it.
You build a platform, you put it in the hands of the world's developers, you say, "Can't wait to see what you build," we let developers show us what are the most burning problems inside of companies. What are the most valuable things that they need solutions to? What needs to get solved in the world? Developers start building. That's not enough, because after you build a platform, you start to see these big, juicy areas, the world's most pressing problems that need to be solved in communications, you move on to step 2, which is go deep and ensure that Twilio and our ecosystem wins. Make sure that we take those workloads that are the most important ones and that are migrated onto Twilio's cloud and into our ecosystem.
By doing so, you go deep and you win in the areas that are most important and most valuable and the biggest, juiciest areas of this market. Then you finish with step 3, which is repeat. Build more platforms that expose you to more problem areas, more use cases, more solutions that need to be built, you repeat this whole process. That's our master plan. That's how we intend to tackle this very large market in the fullness of time. If you think about the history of Twilio and where we started, back in 2008, we launched our very first product, Twilio Programmable Voice.
Allowed developers to make and receive phone calls, but also give them the basic building blocks of things you can do during that call, like saying text and playing back audio and recording audio, things like that started to unlock the key use cases. What's going on in the world of voice that needs to be updated? Call centers, call tracking, IVRs, things like that. A couple of years later, we launched our second major product, Twilio Programmable SMS, the ability for a developer to send and receive text messages with any phone in the world. We started building all the intelligence services that sit on top of that. How do you make sure the deliverability is high and they don't get blocked by carriers and things like that? We started building all these intelligence into those products.
After several years in building out these products, we asked ourselves, okay, we've got a great voice product, we have a great SMS product, we're going to keep investing in those and building those, what else is the future of communications going to consist of? These technologies are both existing technologies. We looked at smartphones coming online. We looked at LTE, 4G, 5G networks coming. We looked at the ubiquity of mobile apps. We said, clearly, there's going to be more to the future of communications than just voice and SMS. We started building out what are the other components that developers need that may constitute this future of communications. What are the other building blocks that developers need to experiment with to figure out how we're going to communicate 10, 15, 20 years from now?
We started building those out, and we started a few years later. We built out Twilio Programmable Video and brought that to market in 2015, allowing developers to build multi-party real-time video into their applications. Right. Then we did the same thing with chat, allowed developers to build chat into their web and mobile applications with a very modern chat experience. Then we said, "Okay, well, what else?" There's all these social apps out there, things like Facebook Messenger, WeChat, Viber, and WhatsApp. Developers are probably going to want to tap into those. We built our Channels product. Then we started incorporating virtual assistants into that because you've got Alexa, Siri, Cortana, and all those personal assistants.
This year, we started adding those into the mix because those might be places where developers want to interact with customers. Right. What else? We built Twilio Programmable Fax this year, which believe it or not, there's actually huge workloads for fax out there in the world. Then we said, "Okay, well, what else?" AR and VR are starting to come online. Maybe the future of communications will incorporate augmented reality or will take place in virtual reality. We're going to start offering some building blocks for developers to explore what the use cases of communications might be in AR and VR. Who knows? I'd love it if one day in the future, Twilio had Twilio Programmable Holograms as another one of these building blocks. Definitely not a forward-looking statement there, by the way. Right.
Whatever the building blocks that developers may need to experiment and build this future of communications, Twilio is going to be there building those and giving those to the developers of the world. That's how we see this opportunity. Then we started building out the product wheel, if you will. We built out all these capabilities for developers. After several years of building out all these capabilities, we took a step back and we said, "Okay, maybe it's time to move to step 2 of our master plan." What are all the things that developers are building with all of these mediums of communication, and how should we think about where to go next? When we sat back and looked at it, we saw that all of the world's communications seem to fall into 4 categories. First of all, there's consumer, C2C, consumers chatting with each other.
Things like Facebook Messenger, iMessage, WhatsApp, and WeChat, these types of applications, right. There's been a lot of, obviously, activity there. We've got customers doing amazing things in this category. Okay, well, what else is there? There's also the collaboration market. B2B. How do people inside of companies collaborate with each other? You see things like Slack, and Microsoft Teams, and Lifesize, and Zoom, teleconferencing, and all this, and software is making it easier and easier for us to collaborate with each other. This space has been moving really fast, and we've got customers doing amazing things in this category. Then you look at the third, machine to machine, IoT. This is obviously an exciting space. If you believe the hype, every grain of sand on the beach is going to have its own IP address one day. Right.
You've got 60 billion devices are going to be connected in just a few years. We see this world growing immensely. The last category was the most interesting to us. B2C, business to consumer, customer engagement, how companies talk to their customers. Well, first of all, we looked at our customer base. We looked at our revenue. The vast majority of Twilio's revenue comes from companies, developers building apps that are trying to engage with their customers better. It's the number 1 set of use cases. We said, "Why is that?" Well, why? It's because it's incredibly broken. We feel this all the time as customers. When we do business with companies. We feel like the experience in dealing with that company is kind of broken and disjointed, and they don't really know who we are, and it takes five calls to get anything done.
We experience this all the time. Engaging with companies is broken. Guess what? It is not getting better. It is getting more broken because of the speed at which all these other technologies are moving. By contrast, when you've got a mobile phone and the apps get updated every two weeks, and you've got Apple launching new products every year, Facebook doing the same, Google with RCS, all these platforms are moving at lightning speed. Companies can't keep up. The feeling is that it's getting worse and worse when you try to engage with these companies. When we talk to companies, we said, "Is this a problem?" I talked to the C-level executives. I talked to the CEO of a major bank recently. I was trying to understand, what's on your mind? How can I connect Twilio to what's on your mind?
I said, "Is having a conversation with your customers strategically important to you?" He said, "It's one of the top three things that I think about." Why? "Because having a conversation open with my customers is the number 1 thing that I need to do. I can't get disintermediated by social networks and all that sort of stuff. I've got to have a strategy to stay relevant with my customers." One of my top three strategic priorities for this major bank. I said, "Interesting. What's preventing you from doing that?" "Well, let me tell you. A few years ago, it was going to be a mobile app. We invested all these resources in building a mobile app.
By the time we finished with that mobile app, the world had moved on. Now it was about Facebook Messenger. Then it was about Siri. Then it was about iMessage. Then it was about Facebook Messenger. Then it was about Alexa. The world kept moving on. We just can't keep up. By the way, it's not just one team who has to deal with this. It's all these different teams. It's my sales team that's trying to engage. It's my marketing teams who are trying to be relevant. It's my support and my operations teams.
All these different functions around the company are all trying to keep up with this pace of technology, and it is just not possible. You take this problem of the speed of technology, multiply it by the number of teams that are in the company who all need to answer this problem in order to have a compelling customer engagement story, and it's a totally untenable problem. That's why we've decided to double-click here for step 2 of our master plan. It's broken, it's getting worse, and we can solve it, and customers are already coming to us for it. That's what the Engagement Cloud is all about. It is an application platform for customer engagement. The funny thing is, customers are organically coming to us with touch points across the entire customer life cycle.
Various teams inside of various companies in our customer base have all come to us organically saying, "Well, I'm trying to build for this part. I'm trying to build for the marketing touch point. I'm trying to build for the sales touch point. I'm trying to build for the support touch point." These are all big markets. Sales, marketing, service and support, logistics, operations. These are big markets, and that's why we're starting to see competitive deals with companies like Cisco and Avaya and Eloqua and Marketo. These various solutions that companies have had to stitch together in the past that never quite worked so well together, they're starting to say, all these customer touch points across the entire customer life cycle, how do I make it so this thing works?
How do I make it so that I've got a single system of engagement for all the ways and all the teams that actually touch my customers? How do I get rid of this hodgepodge, this stack that barely stays standing, hodgepodge of solutions, and how do I get something that actually makes sense and is holistic? A system of engagement. What we've seen is that we can take and combine the flexibility that customers love from our APIs and deliver them to the speed of buying a solution that they like for something that's off the shelf. That's what we consider to be the value proposition of an application platform. It can give you both flexibility and speed. That's what we're investing in with the Engagement Cloud. You're going to learn more about our Engagement Cloud from Pat in a little bit today.
We're excited to talk about that. You may be asking, okay, great, you build the Engagement Cloud. What comes after that? First of all, the Engagement Cloud is a big undertaking. We're going to take down multiple huge markets. That's our goal. We got a lot of work to do there, obviously. We do think about the future. You may also be asking yourself, where's Twilio Programmable Wireless on this chart? I haven't seen that yet. That was a big deal, right? Absolutely. I think of that as actually the beginning of our next platform. That's the repeat step. We've started, we've seeded. Programmable Wireless is the next platform that over time will show us what are the big problem areas in IoT that need solving. What are the big challenges that every company faces?
We're starting at it from the network side that one day may turn into the Twilio IoT cloud as our next big app platform. We're not worried about that today. Today, we're driving deep on the Engagement Cloud. That ultimately is how we're going to fulfill on this mission of fueling the future of communications. We're riding a huge wave here. The disruption of one of the world's largest markets and 40% of all IT spend, the disruption by software of this enormous market. We're systematically attacking the major workloads and the major use cases in order to fulfill on this opportunity. If you believe, as we do, that software is going to fundamentally transform how the world communicates and improve it and make every company better at communicating with their customers and make all of those categories better, then you'd agree with us.
It's day one. Day one of this huge opportunity. Twilio is making the investments to unlock this opportunity, but it's a long play. This is a big market and a big transformation that's going to happen, but we're excited to be making these investments. Now, I want to talk about the second part, how our developer-first approach is driving efficiency in our go-to-market and also giving us broad reach into this very big market. We started with this idea, ask your developer, that developers would be key to this migration. We want to remove friction, make it easy for developers to get up and running on Twilio, put Twilio in the tool belt of every developer in the world. It also allows us to efficiently grow our revenue because of that, because developers become our big advocates inside of companies.
We now have over 1.9 million developers on our platform. Those developers adopt Twilio, and then they bring us into the companies that they work for to solve the business problems. One of the things I think that we underestimated when we started this company was just how well this would work in companies of every shape and size. Look at this example. This was a recent win of ours, Morgan Stanley. Morgan Stanley wanted to empower their private wealth managers to be able to talk to their customers in a modern and engaging way. Text message, right? Wow, mind-blown, right. To do so in a compliant way. Previously, they'd said, "Sorry, you're not allowed to text with your clients. It's not compliant. We can't track it." They wanted to enable this, right?
They moved to a solution where everyone, I think 16,000 private wealth managers around the world, are using an app they built powered by Twilio to now allow them to text with their customers. Here's the amazing thing. The champion of this product was the Chief Digital Officer, Naureen Hassan, and she's going to come out here on our customer panel later and talk to us. This project got started by the developers. They prototyped it. They said, "We think this is possible. We'll come back to you." They prototyped it. That's what got the ball rolling inside of Morgan Stanley, one of the largest banks in the world. Right? This implementation is just the first step. Naureen's going to share her vision for the full roadmap of what they want to do. That ball got rolling.
The whole thing got started because the developers, they asked their developers and said, "What's possible here?" The developers said, "We know. We think we got this. Let's build a prototype." That's how it got started. We all know that Twilio's model, this API-first approach, works well in software companies. Your Silicon Valley disruptors. We all know that. It also works well with solution partners, ISVs. Those ISVs, those solution partners who build SaaS products on top of Twilio are amazing because they allow us to reach into SMB markets, SME markets, or enterprises that don't want to build, they just want to buy. Great. We've got a great answer for that part of the market as well via our solution partners. The question that everybody has had, what about the enterprises? Are enterprises really going to build? Well, yes.
Here's the thing. Enterprises have always been builders because they have these big customization requirements, and they've always invested huge amounts of money, huge amounts of time. Twilio as a platform is ideally suited to help them answer that question to get exactly the customized environment in a complex environment that they operate in to get that solution. Between these three, we have software companies, solution partners reaching into the SME, SMB audiences, as well as the enterprises. Twilio can cover the whole market, and Twilio is actually a great solution for all three of these. You see this coming through in some of the analyst reports that are coming out. In fact, DMG, a market analyst around the contact center, recently published this report.
It said, "Contact center infrastructure platforms like Twilio will attract the attention of large enterprise contact center environments that are accustomed to building their own solutions." We can allow them to get exactly what they want, the reason why they've historically built, but get that value faster, get that value cheaper, and get more customization than they've ever had before. By the way, I want to point out in this same DMG market research report on the contact center, they pointed out that Twilio is estimated to be the second-largest cloud-based contact center infrastructure based on the number of customers, and the third largest based on seats. Twilio is the third-largest cloud contact center, and we don't even have a product in this space. That's the power of the Twilio ecosystem. The power of a platform. We haven't even tried, and we're the third-largest vendor.
That's the power of this ecosystem and this platform-based approach. The other thing, in addition to opening up the whole market, is this model is efficient. Why? Because developers do a lot of the heavy lifting for us. They're our champions. They build prototypes, and they bring Twilio into companies of every shape and size, and then they become our heroes inside of the company. Instead of listening to slideware, they listen to their own developers. That's ask your developer, right? The developers become our hero, and that's why we've got this very efficient revenue acquisition model. Only 23% of our revenue is spent on sales and marketing to grow at the rate we have. If you think about why, a lot of it has to do with our usage-based model.
I showed a chart like this during our roadshow, I explained, here's how customers adopt Twilio. A developer might get started. In that first month, it might be a dollar, maybe $3, maybe $5. This is an actual customer revenue history here. One of our sort of typical customers, nothing extraordinary. They get started. They build that prototype, and they spend a few dollars on Twilio, and you know what? That's great, because once they feel like they've got that prototype dialed in just right, they put it out in front of some of their customers in a beta. You see the usage start to tick up as they roll it out slowly to more and more of their customer base. Customers like it. They turn it up all the way. They roll it out to the whole customer base, the revenue grows.
That's the first of three expansion vectors we have in every single account. As a product goes through the product development life cycle, from prototype to beta to GA, global release, it's going to touch more of their end users, it's going to generate more usage, our usage-based revenue model means we make more money as they go through that. The second vector of growth is that as our customers are growing their end user bases, they have more people to interact with, drives more usage, drives more revenue to Twilio. We benefit indirectly from the combined sales and marketing expenditure of all of our customers combined. That's pretty cool. Then the third expansion vector we have in every one of these accounts, once we get the developer on board, is saying, "Great.
You know Twilio can be used in all those different life cycle events of the company and by all these different departments inside of the company. What other problems need to be solved?" The developers or the business say, "You know what? We need to solve this problem over here. Let's do that." You see this big spike of growth at the end here. That's the deployment, that's the global deployment of their second biggest use case. This product development life cycle, coupled with repetition of it, doing it again and again for all the different departments that need it, this is helping to drive our growth very efficiently. I like to summarize our business model as this. We acquire developers like consumers, but we let them spend like enterprises. That's kind of what it's about.
We win the hearts and minds of developers, then because of that, we win the hearts and minds of businesses, we solve a really juicy strategic problem they have of how do I talk to my customers, right? Even in large deals with major banks, right, I mean, people like this is the same dynamic we see playing out in a wide variety of companies. Guess what? The Engagement Cloud is helping us do it even more. Why? Because we're connecting Twilio's API-based approach to the key strategic things that are important to our customers, customer engagement. You don't just have to believe me. I think this is represented in our numbers, right? This is our revenue expansion rate ex Uber for several years. This developer first go-to-market is showing here, right? Remarkably consistent revenue expansion rate, taking out the outliers.
Through this period of time, the numbers are getting quite large. The numbers keep growing, but the expansion rate has been very consistent. George will talk more later in our go-to-market section about the things that we're doing to continue to drive our go-to-market and our business model for innovators. Okay, the last thing I want to chat about was our leadership, our leadership position that we built and continue building through software differentiation and by building trust in the market. See, I think there's this interesting dynamic in the cloud, which is that cloud markets tend to develop a dominant leader. If you think about it, right, for CRM, it's Salesforce, for infrastructure, it's Amazon, right? There's a far and away dominant leader, generally speaking, in all these markets. Why is that? I think it's because of trust, right?
With the cloud, what you're doing is you're carving off a part of your business and handing it off to another company and saying, "Here, I trust that you're going to execute this better than I am. I trust that you're going to execute this better than some other vendor." That's a lot of trust. Therefore, the dominant player continues to grow in dominance because that's where all the trust ends up residing. Leadership begets leadership in the cloud. Our approach is software, to invest in software, to build that trust, and to build that leadership. See, we're a software company. Almost half of our headcount of our company is software developers, R&D organization. The company was founded by three software developers.
What we believe in is building features and functionality and capabilities that allow our customers to do amazing things with software and fulfill on their business requirements, and to do so with a high degree of trust, right? We call this agility with resiliency, right? Some companies believe that you can either be agile, always be shipping, right, ship fast and break things, or you can be resilient, right, and say, oh, a lot of enterprise companies will ship two, maybe three times a year, a big update, and that's how they get reliability. But at Twilio, we reject this as a false dichotomy. We believe you can move fast. You can always be shipping, listening to your customers, being agile, and you can do so with reliability, right? You can ship new business value every few days, and you can do it with five nines availability.
That's what we believe, and that's what we've been executing on. Ott and Pat and Christine are going to talk later today about how we fulfill on this promise of shipping, listening to our customers, and being able to deliver the value that they need and doing so with five nines availability and all the compliance things that you can imagine. They're going to go into detail about that. What I wanted to show you is how I believe this is a virtuous cycle that we've invested in. Some people ask, "Why do you invest so much in trust? Why do you invest so much in reliability?" Because here's how I think this virtuous cycle works. You build a product, you get started, you're a startup, you build a product, customers like it. Customers start coming to you. What do those customers do? You get traction.
They show you the market. They show you what they need. They give you visibility into what the market is going to look like in the future. They show you what you need to build in order to fulfill on that potential. The other thing is, as customers flock to you, they give you their requirements for security and reliability. They set a high bar for you because you're the only option. They say, "I want to use you. You got to hit this bar." You invest in that. You invest in those capabilities that customers give you visibility into. You invest in that high bar of quality and reliability and trust that they are setting because they need it or else they can't adopt you. Because you invest in that, it gets this flywheel going.
Trust that you start investing in as a young company leads you to customers and leads you to build an ecosystem. That ecosystem paints the picture for you that allows you to drive your product velocity. They give you insights into where the market is going. They let you see into the future and build faster and better than anyone else. Also, they require you to build the resilience into your product in order to adopt you. Because of all that future visibility and resilience you built in, that leads to more trust. Trust leads you to more customers. The virtuous cycle completes itself. I think that's how the cloud works. That's why we've invested so much in building out agility and resiliency, delivering software value to our customers at scale with an extreme focus on reliability and resiliency. Leadership begets leadership.
That's how you build an ecosystem in the cloud. You can listen to me or not listen to me or Ott or Pat or Christine. We can listen to a leading industry research firm who recently said Twilio is the leader in Communications Platform as a Service globally. Undisputed. We'll provide you a copy of that report. To summarize, right, this is day one. Huge opportunity. Never seen an opportunity like this in my entrepreneurial career to reshape one of the world's largest markets with the power of software. Our developer-first approach allows us to do it very efficiently. Also cast a very wide net as we tackle this big opportunity. Our leadership is gained via our investments in software, agility, resiliency, and building trust with the market that leads to a flywheel of growth and dominance in the market. Thank you very much.
Now, actually, I have one more factoid actually I was going to share. Before I invite up Christine. Did you know, this is a total coincidence, that this week is the 25th anniversary of the very first text message ever being sent. I thought that was kind of cool. I also thought it was kind of cool that we've recently started hitting peaks where every day we're delivering 100 million text messages. I think that's pretty cool that we're hitting these peaks on the 25th anniversary of the first text message we just hit this week. That's really cool. With that, let me invite up the head of Twilio's Super Network, Christine Roberts. Christine.
Thank you. Thanks, Jeff. I'll skim over the legal disclaimer as well. My name's Christine Roberts. I'm the vice president and general manager of Twilio's Super Network. When I was considering coming to Twilio about a year and a half ago, I came from a little bit of a telco background, a lot of my questions to the executive team and to the engineers I talked with were, how are you approaching this differently? Because I've lived in the world of telco, and I know what it looks like, and if we're just going to do that again, then maybe not so much. What became very clear and very obvious to me very quickly is that Twilio was approaching it completely differently.
They were approaching it from a software-first mentality, they looked at this massive global telcos all over the world, there's thousands of providers everywhere, and they said, "Look, the only way to harness all of this is to harness it with software." If you combine the power of software with all of that underlying connectivity, that gives us a differentiator. That leads to something very different than what a traditional telco would do to approach this market. It's also important that if you think about how developers look at the world, they don't look at the world along the lines of geopolitical lines. They look at the world through the eyes of software, and an application should work in the U.S., it should work in Brazil, it should work in Indonesia. It's software, right?
Software needs global reach, we need to abstract away the complexities of achieving that global reach. Then I asked about, okay, how are we approaching this? Because there's a lot of folks in the market who are looking to put in least viable solutions and lowest cost, I said, "Listen, I believe very strongly that customers are number one." It also became obvious to me that that was a tremendous alignment with how Twilio was approaching this entire world. All of this creates this amazing flywheel of greatness that I could not be more thrilled to be a part of. Jeff talked about the master plan, the step one of that master plan is building platform, luckily for me, I am part of that platform. The Super Network is the platform upon which the Programmable Communications Cloud and the Engagement Cloud sit.
Their expectations of the Super Network, they expect us to be able to scale, not scale linearly, scale for burst, scale for great opportunities that we didn't see coming. They expect the Super Network to provide global coverage. George and his team say, "Do we have coverage here? Do we have coverage there? Can we reach this place? Do we have numbers in this place," right? They expect the Super Network to provide that level of global coverage, our customers and developers expect that as well. They expect the Super Network to get smarter to learn to adapt to grow, they expect us to allow our developers to write software instantaneously, part of that is providing instant provisioning of phone numbers. Right?
Having phone numbers where developers need it, when they want it, how they want them, when we say we have a phone number available, actually having a phone number available, not, "We'll get back to you in a couple weeks." Of course, phone numbers in general are as unique as all of us who own them, for sure. Jeff has talked a lot, I think a lot of you have seen and heard the analyst calls and heard the road shows. He talks about why Twilio was born. It's that fundamental problem of how difficult it is for developers to reach the world. So part of the first problem Twilio had to solve was establishing a globally distributed communications network, and we spent the last nine years building that network out.
In case anybody tells you otherwise, it's really hard work. We've got some amazing teams that have built on my team, that have built that level of connectivity. We're so proud of that level of connectivity and the global connectivity we've built and that network of connectivity that we had to come up with an amazing name, and of course, that is the Super Network, as you know it today. Despite the fact that when I asked Jeff if I was going to be required to wear a cape to work every day and he didn't get it, I think it's incredibly apropos. What I've learned in this year and a half is that the secret sauce that actually makes the Super Network unique, because I think a lot of people today are calling their networks super.
The secret sauce that's underlying it all is the fact that it's actually a software network. We manage our network with software. There's complexity associated with all of these carriers and all of this connectivity and all these different connectivity types and all these countries and all of that, and it's impossible to manage on a scalable basis. To think about the scale that Jeff's talking about, to think about that world and be able to scale, it's impossible to manage that without software. This is a big data problem, to approach it otherwise, I think would be not the smartest move.
As I said, I came from a bit of a telco background, I spent a little too much time in a few master switching offices, I'm pretty sure at one point or another, I was associated with one or the other of these blue boxes, which are represented with hardware. The lines in between them are copper and fiber and all the things that connect our networks globally. If you think about a telco approach to solving the world's connectivity, this is kind of the first thing that you would do, right? You'd draw a network diagram, you'd figure out what underwater cables you're going to leverage, you do all this kind of stuff.
If you go back to the fundamental problem that Twilio was trying to solve, the fundamental problem Twilio was trying to solve is: how do I use software to make a phone call? That's the underlying problem. If you solve that problem first, your approach is completely different. As such, the Super Network is a network of networks. We have connectivity to lots and lots and lots of carriers, but we virtualize that complexity away from the developers and the enterprises that want to leverage Twilio software and Twilio APIs. They don't need to care, nor should they have to care about all of those boxes. They just want to be able to make a phone call with software, and that's what we've done with the Super Network, is we've abstracted all that complexity away.
That's just one piece of it because underlying all of that complexity and all those networks is the realities of the world of telco. This is one of my favorite charts. This is a three-day, I think the x-axis is three days, and the stack bar are minutes of outage. Every single one of those colors is a carrier we're connected to that had outage and downtime. Every one of them. You can imagine if our customers were subjected to this variability all day, every day, every week, every month, every year. This is the reality.
I was talking to one of my providers in Brazil, I think about two weeks ago, I was trying to understand the postmortem of what went wrong, and they said, "Well, somebody climbed a telephone pole and removed all of the copper wire between that pole and the next pole so they could sell it." That happens. That happens maybe not so much in the United States, but in other countries, it actually happens. That mess is abstracted away by the software that the Super Network brings. Our customers felt zero impacts from all of that downtime, from all of those outages. Zero impacts because our software says, "Oh, look, there's a problem. Let's put your messages over here. Let's make sure they get delivered." That's the power of software, that's the power of how Twilio has approached making the Super Network incredibly resilient.
Almost all of our telco issues at this point, we catch them before our carrier providers do. What typically happens on a normal day is we say, "Okay, I see that this traffic was rerouted from here to here. We better let that carrier know there's a problem." We pick up the phone, and we call that carrier and say, "Hey, we've rerouted our traffic. You might want to look into some issues that we're seeing." Right? That's our job. Excuse me. That's our job. That's not the job of businesses or enterprises or even developers, right? That's my job. That's my team's job. The way that we're able to do that is we monitor over a billion data points every single day. This is how we do it. Right? You can imagine in the old world of NOC, right?
Where you have lots and lots of people sitting behind terminals and looking at graphs and charts. This is the only way to manage a global network. Back when Twilio and I were still dating in SIGNAL 2016, I had the amazing opportunity to listen to Nancy talk about, as part of the .org initiative, talk about her team at Crisis Text Line. She manages a team of 24/7 counselors and offers a free nationwide text line for people in crisis. She stood up on stage and said, "Look, we can't afford downtime. When these people text, they expect someone at the other end to answer them. They expect someone to be there when they're probably not having a great day." Right? This is why businesses build on Twilio. This is why developers build on Twilio, right? This is the kind of things that developers do on Twilio.
Connecting calls and messages around the world is just one piece of the puzzle. If that's all I had to do, maybe I'd get to go home at 4:30 or 5:00. Instead, another piece of the puzzle that we're responsible for is phone numbers. Phone numbers all around the world. In order to leverage voice, in order to leverage text, you start with a phone number, right? If you've been to our website, first thing we do is say, "Buy a phone number." Pat likes to say that phone numbers are the planet's most important identity. I'm not sure if I believe that or not, but I do know that our businesses and our developers rely on my team to be able to deliver on that identity for their apps and their businesses worldwide, and our customers worldwide.
It's core to the success of their business is having the right phone number for the right application. It's very personal. It's very local. I don't know if any of you recognize this phone number. This phone number is very personal to our head of investor relations, Greg, and you'll find out I don't think it's his home phone number. You'll find out more about this phone number in Pat's presentation. Part of that phone number being personal, and part of it being local, and part of it being specifically applicable to a use case, is making sure that developers all over the world have access to this massive breadth and depth of phone numbers that we've collected over the years.
Prior to SIGNAL London this year, my team made sure that every single phone number was perfectly curated to be able to fit any possible use case that that number could possibly be forced to handle. What our developers said to us, and what our enterprises said to us is, "Look, we don't need every number to be perfectly capable of every use case. We need our phone numbers to be perfect for our use case." As part of the innovation and the software innovation that goes on within Super Network, we released the Global Phone Numbers Catalog, which allows developers, which allows businesses to have access to the broad choice of phone numbers that we've cataloged over the years. There's millions and millions. It's use case driven.
You can put in your use case it'll say, "Hey, we've got the perfect phone number for you. Here's some suggestions." It's a powerful search engine. It allows you to keep track of your phone numbers and what attributes they have, and we're incredibly proud of being able to expose the breadth and coverage and amount of phone numbers we've looked at all over the years. Obviously, having phone numbers is important, but having a breadth of phone numbers is really important. We also announced at SIGNAL London that we have coverage in 100-plus countries. What's really great about this is we don't just say we have numbers in 100-plus countries. We have tested numbers in 100-plus countries. We have provisioned numbers in 100-plus countries. We have these numbers on demand when a developer needs them.
If you think about what that allows our developers to do and how it allows them to build out their applications in a global scale, it's pretty awesome. We don't stop. We don't rest on our laurels. We're continuing to build out reach to other countries. As our sales team tells us, as our customers tell us, we'll continue to expand. 100 was a pretty great milestone for the team. Breadth of coverage isn't enough. My daughter's a teenager, I had the interesting opportunity to get her her first phone number. You talk about personal, it was very personal. This is going to be her personal phone number for probably a very, very long time. I picked up the phone, I called our telco provider, I said, "Hey, I'm getting my daughter her first phone number.
Can you help me pick and choose?" She's like, "Sure, I'd be happy to help you pick and choose a phone number for your daughter. I see that you live in California, where do you want the phone number to be from?" In a very strange way, it's a very valid question as it relates to our phone bill because our cell phone numbers, mine and my husband's, happen to be Chicago-based. I was in Chicago for 20-plus years. When I moved to California, we kept our phone number because it's very personal, right? I'm not going to get rid of all my contacts. People need to contact me, even my mother. Now my mother and my father, who happen to live in Michigan, have Northwest Indiana phone numbers, which is a weird thing I won't go into.
My in-laws, who are also on our phone bill, are in northern Wisconsin, which I think is a massive amount of area covered by one area code. You can imagine, I'm on the phone with a customer service agent, she says, "Where does your daughter want her area code to be?" We said, "Well, at this point, she's a California girl, let's make it a California-- Let's make it a 415 number," and we got her a 415 area code. Imagine that same conversation if the answer from that customer service rep is, "Yep, we're super excited to get your daughter her first phone number. Does she want a North Dakota number or a South Dakota number?" I'm like, "Wait a minute. That doesn't make any sense." That's why depth of coverage matters.
While we can relate to North Dakota and South Dakota and the difference between a California number and the United States, you can imagine that a business in Basingstoke, U.K., doesn't want a London city number. Right? Twilio focuses a lot of energy on making sure that within given countries, we have a depth of numbers to provide across all those different locales. Just as an example, here's the depth of coverage one of our direct competitor provides. You can imagine that that's not a very local or personal service if those are your only options. What makes me incredibly excited to get up every single day and to come to work and to continue to build this amazing network is that, Jeff talked about the flywheel of trust.
This is my flywheel, as we have more connections, we have more data points, we lower our cost and we raise our quality. As we lower our cost and we raise our quality, we get more customers. As we get more customers, we get more volume. We get more volume, we get more data points, we get more connectivity, and it just gets better and better and better. Software intelligence combined with connectivity, it's the only way to manage a network of this size. It's the only way to grow it's the only way to scale it, and trying to do it any other way is ludicrous. Global reach, whether it's depth, breadth, connectivity, the ability to terminate in any country you need to terminate calls or messages to, will never stop being important and will just get better.
That flywheel of continuous improvement, is what gets me up every day. With that, I'd be thrilled to turn it over to Ott Kaukver who will talk to you about the Programmable Communications Cloud.
Thank you, Christine. Hi, my name is Ott. Important thing, legal disclaimer. My name is Ott Kauker. I'm the VP of Engineering, and General Manager for voice and video business here in Twilio. I'm an engineer, as a background. I've spent 15 years in communications industry. Before Twilio, I spent eight and a half years at Skype when I started scaling Skype when Skype was in one small room, from 20 persons to saw the growth up to 2,000 people. Built up the operations department first and then ran the engineering for seven-ish years. I've had the privilege to be part of the Twilio team for the last four and a half years and built up the R&D organization that we have here, really excited to be part of the next wave of growth. Today, I'm going to focus on Programmable Communications Cloud.
We've had remarkable growth over the last nine years, but I believe that we are just starting. We're going to talk about the opportunity ahead of us. Secondly, I'm going to talk about the world-class developer experience that has helped us to build an efficient go-to-market model. Finally, how those investments into R&D and software have helped us to develop a structural advantage for Twilio in the marketplace. Jeff talked about step number one in our master plan, build platforms. Christine just covered how we have taken the complexities of the telecommunications network and really abstracted that with the software APIs. Programmable Communications Cloud is the software layer on top of that Super Network that helps developers to access that network using the tools they know through software APIs, through communication APIs. Programmable Communications Cloud is a general-purpose platform that consists of two things.
First, multipurpose building blocks for virtually any use case on the planet, for any communications use case on the planet. Secondly, a software intelligence layer that helping to build advanced use cases and also save developer time for our customers. It's designed to get customers on the platform. Jeff already walked us through our history a little bit. We started with a voice primitive in 2008. We added SMS in 2010, and then later we had IP-based primitives, and then wireless, and even fax, believe it or not. This enables customers to build a wide array of applications on top of the Super Network. Communication preferences change, like new channels, new communication applications pop up every day. If I open up my phone, there's probably seven or eight different applications if I need to communicate with people.
That also forces businesses to change how they communicate because they have customers now everywhere. That's why we added additional channels underneath our communication primitives. It's using the same API that customers can communicate now through a variety of channels, like for example, Facebook Messenger. That's important because it enables omni-channel communication. Businesses can now talk to the customers where they are. Secondly, it signals that the Twilio platform is future-proof. We keep innovating, we keep adding all the channels so the customers they wouldn't have to do new integrations. This is possible in the area of software. In the hardware, you will have to buy new boxes and do new integrations with expensive consultants, but we keep getting better every day. We do more than, estimated this year, more than 30,000 production deployments to our platform.
That's sometimes more than 200 deployments per day. Twilio gets better for our customers 200 times a day. That's pretty powerful. This platform approach is a great possibility to unlock new markets. It's great for opportunity creation. Flexible building blocks enable us to build almost literally any communication application in the world. For example, in contact center, you can build omni-channel contact center experiences like IVRs and with bots. Logistics. Customers can choose to coordinate the deliveries, and they connect the agents to the people who they want those deliveries to get delivered to. E-commerce, product managers can use smart notification to increase engagement. It's the flexibility that enables the innovation and unlocks whole new markets. That's why we built Twilio to be global from day one. We have an infrastructure in nine different regions, which enables a local experience.
With one API, with one software, with one application, whatever your customers are building, they can build this for all the countries and powered by 100 countries' phone numbers that Christine has provided. That's great. You can do it all from your keyboard, and developers love it. Speaking of developers, like Jeff said, we have 1.9 million developers on our platform. How have we achieved this amazing developer acquisition model? It stands on 3 pillars. The first one, where we have invested more than 9 years now, is self-service business model. You put the sign-up page up there. It's low friction. It enables customers to start prototyping as soon as possible. You provision your first phone number, you put the credit card in, and you start experimenting. Experimenting is important because experimentation allows innovation. Through innovation, new businesses get born.
You think that's super easy, everybody can do that, but actually making something easy requires hard work beneath that. If you target the global market, you need to deal with localization, different countries' credit card payments, and also fraud. You need to build machine learning to prevent fraud that's happening. We've invested 9 years into tools and platforms to make sure that we have low friction onboarding experience for our customers, and that's super powerful. We acquire customers like consumers. Second, our world-class developer experience. It was really designed keeping developers in mind, designing intuitive APIs, providing helpful libraries in all the major software languages, quick start tutorials. Not to just get people started quickly, but also help them to really retain them on a platform. The 3rd thing is our runtime platform, and I want to dig in a little deeper here.
One of the key metrics that we measure is how much time does it take for a developer, whom we have acquired like a consumer, to develop that first application. Sometimes it can take months because you have to deal with hosting and infrastructure and service and all this kind of stuff. How can we make this quicker? How can we make sure that developers are getting their apps to production faster? That's why we created our service platform. Our functions and asset products that enables to reduce the time to production from experimentation from months, sometimes to minutes. This accelerates and helps our funnel conversion. Our investments into monitoring tools and debuggers have put really the control in the hands of the developers. Not only it's just better experience for developers, it actually helps us to reduce the support costs.
The 3rd thing, we've built the platform to be enterprise-ready. If enterprises consider to have a development platform, they need to have a certain set of requirements that are must-haves in the platform to be considered, like single sign-on, role-based access controls, auditability, the traceability built into the platform. That's what we have in Twilio Runtime platform. Why do customers choose us? We talk to our customers often, for developer evangelist events, for our customer meetings, and one thing that always comes up from our NPS survey results is quality and reliability. Jeff already mentioned that, too. If we acquire our customers efficiently, we need to retain them on the platform, too, because the platform that isn't able to guarantee uptime, resiliency, and massive scalability, it's really going to affect the customer experience for the applications built on top of that.
What enterprises and what customers need is five nines of reliability. I just want to put this into perspective a little bit. Look at this chart. Five nines reliability means that you can have less than six minutes of downtime per year. Six minutes. That goes by pretty quickly. Five nines is hard. You cannot achieve this with humans. You need to invest into R&D, build systems, automated recovery, failover, and that is what we have done. Over the course of last almost decade, we invested more than $200 million into R&D, both building products, building security, but also building resiliency into our platform. This has enabled Twilio to achieve the number one thing that customers want from us, and what customers want from their strategic communication platforms. That is the five nines of availability.
Second thing that we measure is not just availability, it is success rate, because the API that is up and serving errors is not really useful. We also measure all the requests that we process, how many of them we process successfully. I am happy to report that that metric exceeds five nines. From day one, Twilio has never had a thing called maintenance window, either planned one or unplanned one, which is surprisingly quite common in the enterprise world. Even with the big providers, we have a five nines availability, however, we have a maintenance window of two hours. It does not fit into my head, at least. Is not it?
How can you? If you build a reliable application, if you have something which is life critical, like even the Crisis Text Line, and like, oh, I could not send you a text or connect your call because I was having a maintenance window. Does not really fit in my head. What we have done here is that we have really built our systems in a way that there is no maintenance windows ever. Five nines without maintenance windows. We have also invested in security. ISO 27002 for Twilio, SOC 2 certification for Authy, and we are working towards GDPR support for our platform in May 2018 to keep our global customers happy. Second thing what customers choose us for is to really provide them building blocks so that they are able to build their use cases they want. Let us look back at 2008. Programmable Voice.
A simple API, able to connect calls, play media. We put that API out there. We learned how customers use it, what they use it for, and nine years later, after lots of investments in R&D, we have evolved from supporting a simple use case to just connecting a call to delivering a voice message to something far more powerful. We have learned from thousands and thousands of customers, incorporated feedback into our platform, and developed the capabilities they need. We built the most advanced programmable real-time communications network. If somebody wants to replicate that, they will have to spend decades of investment into R&D to do that. It is not just calling, connecting phone calls, or connecting legacy SIP infrastructures anymore. It is about building rich applications using mobile or web technologies. Programmable Voice platform today provides capabilities to build almost any imaginable communication workflow on Twilio.
This really gives the developers the advanced control and unlocks the flexibility of innovation to build really great customer experience. Let's have a look at the messaging product. 2010, also a relatively simple API. You were able to send messages, you were able to receive messages, which was great for sending one or 10 or 100. Once you put it out in the real world, again, you learn how customers use it, and you learn that the customer use cases are different. The workloads on messaging are very different. From marketing to phone verification, the PIN delivery, the peer-to-peer communication, there are different requirements for performance and reliability. You need to deal with scale, queuing, carrier filtering, finding spam, all those kind of things. It's not just anymore about sending a simple message that customers need.
We have invested into software layer, intelligent software layer that takes 95% of the complexity of developing a global messaging application out of our developers' hands. Now they can focus on really building their applications and leaving Twilio to worry about getting the deliveries done, staying compliant, and really making sure that it's working. It's a software layer on top of a messaging platform that handles our customers' delivery from the first message to the billionth so that they wouldn't have to write code. Talking about scale, the third thing that customers want and they will demand from us is operator interfaces. Once your customers get at scale, they need to have knobs and dials because developers probably are doing something else already, and the operators who have to operate the product, they need tooling to do that.
It's not just about making call, just making the phone ring or sending a message. That's why we built the analytics product set, to give customer visibility and put operators in control. Messaging product, voice product, and even insights into the Super Network are now available using our insights product line. The operators can now scale their business with confidence. Almost a decade of investment, starting with communications APIs, then adding channels, listening, learning from our customers what they want so that they can save time, enable new use cases, and operate in production scale. That's why we created the intelligent services on top of that as a second pillar of the Programmable Communications Cloud. Today, now we can build almost any use case, communications use case on top of Twilio. Communications in the future probably is going to be a little bit different.
It's good for today, tomorrow, we already know machine learning and AI is going to change how we are communicating. Bots are already in the contact center, changing the agent productivity numbers. Product managers use machine learning in their applications to decide where to reach out to the what customer and using what channel and what message to use in order to improve conversion or reduce churn. Twilio here is in a really interesting position because we have a lot of data flowing through our system, and we can learn from it, apply machine learning on top of that, and then give this power to the developers to create even better customer experiences. We had taken the first step in this field this May when we announced the Twilio Understand product at SIGNAL. It's a natural language understanding product for communications.
It's integrated into the platform, for example, you can build a dynamic IVR instead of pressing, like press one for sales, press two for support. You can do it in a way that asks customers, "What do you want to do today?" Or, "How can I help you today?" You can speak and Twilio Understand will extract the intent out of it, really reducing the load again for the developers to build a great experience. Understand adds a third pillar to the Programmable Communications Cloud. It's enabling communications of the future. It makes all the programmable communication primitives that we already have better because it's integrated into the platform. Us investing into the AI and machine learning is really betting on the future.
Twilio Platform isn't just great for building the great or any communication application of today, it's actually dynamic and strategic platform for our customers so they can bet on the future, too. To recap, the Programmable Communications Cloud, we believe it's the most advanced platform to build a real-time application upon. Our platform is better because we have a fundamentally better underlying network, software network called Super Network that Christine talked about. We have the wide breadth of communication primitives that unlock massive opportunity in many, many markets. We have a world-class developer experience that wins the hearts and minds of developers and is powering our efficient go-to-market model. We have high platform uptime and massive scalability so that once we have acquired those customers, we actually can keep them on the platform.
We have built the software differentiation that positions Twilio as a strategic platform for the future for the customers. This really fuels the flywheel of growth. The more products we put in front of the customers, the more we can learn from them, the better products we can build, and with more than 30,000 deployments per year, we're moving fast and we are accelerating. This is really flywheel of growth gives Twilio a structural advantage in the market. With that, thank you very much. I hope you enjoyed the first part of the day. Greg, if I'm not mistaken, we're taking a 10-minute break. Correct? 10-minute break. Thank you very much.
All right, everyone, if you could grab your coffees and grab a seat, we're going to get going with part two.
My name is Patrick. I am a GM of messaging here at Twilio. I've been at Twilio now for about six and a half years. Today I'm going to talk about why I'm super excited for the next six and a half years. A few years ago, we realized there was a common pattern that we were seeing from all of our customers. A common set of communication applications that they were building on top of the platform. Jeff mentioned it earlier. It was this B2C communications. The most broken part of many of our communication experiences. These common patterns that we saw across a vast array of customers, marketing teams at places like Nike and Walmart, they were sending out marketing messages for shoe deals, for sales promos.
We saw security organizations, at places like Box and Netflix, that were using our SMS voice offerings to send out pin codes or to enable mobile device verification. We saw organizations like Nordstrom and Morgan Stanley that were building ways for customers to interface with personal shoppers or financial advisors on top of the platform, helping drive sales. We saw Dell and Alaska Airlines enabling better operations through things like flight alerts and shipping notifications. Then we saw companies like ING and Zillow building contact centers on top of the platform. We kept seeing folks take our broad platform and use it in a variety of different ways. As Jeff mentioned, having observed all of these different use cases being built on top of the platform, we devised a strategy to go deep and win. We wanted to expand beyond just communication services and into application services.
Then our Engagement Cloud is how we're going to go about doing that. For the next 20 minutes, I'm going to talk through a number of things with you all. First, we're going to talk through the massive opportunity that exists in modernizing customer engagement. Then we're going to talk about how we're going to leverage our existing strength in the Programmable Communications Cloud and our Super Network that Ott and Christine talked about earlier, how we're going to increase our value capture and customers' time to value. Finally, we're going to talk about how we're tackling the build versus buy dilemma of customers in these markets today. First, the engagement opportunity. I want to share some numbers from a study we did a few months ago. We found that 81% of customers reported that it was difficult to communicate with businesses.
When you dig deeper into the numbers, you found 97% of customers would give more business to companies that communicate better with them. 33% said that in the last few months, they'd canceled a service or switched to a competitor based on negative communication experiences. With such an importance placed on communication, why are B2C communications so broken? Customers think it's critical, but businesses haven't been delivering. Why? It's because they've had the wrong tools for the job. IT organizations everywhere are like this guy trying to fill up that tire. It's never going to happen. They bought a tool that doesn't match the task they're trying to accomplish. Right? The existing tools, they have really four key problems. First, they tend to be single channel. Starting in channels like email or voice, they aren't designed for mobile customers of today.
The second big problem is they tend to actually still exist on premise, still using last generation's delivery technology. Third, they're siloed. Whatever engagement problem they solve, they solve it usually only for a single department, and it doesn't work across the entire customer journey. Finally, they're not flexible. Modern organizations want flexibility to deliver whatever customer experience they want. Looking back at the departments that I outlined earlier, there's a lot of vulnerable dollars here for the taking. Let's take digital marketing. You have the likes of Oracle starting from a single channel, email, and struggling to adopt many of the mobile channels, even though that's where their customers are today. Billions in IT spend, yet customers aren't getting what they want. In the contact center, you have Genesys, Cisco, Avaya. 90% of the wallet share here is still on premise.
There's one other important aspect about this market. They're not just very large markets that are vulnerable. They're also actually connected. Right? If you think about a customer experience, it goes through all these different steps in the life cycle of a customer. You acquire them on the marketing side, you validate, verify them on the security side, you sell to them, you need to service them, and then you need to support them. Right? That's why Twilio's approach to this is different. Twilio's approach is omni-channel from the start. You saw from Ott that we've been investing in multiple modalities for many, many different years, voice and messaging. We have a cloud-based service delivery model. Twilio was born in the cloud, we know how to run cloud services at scale. We shared some of those numbers.
We're API first, because we know that customer engagement is really just part of your customer experience. It's integrated into whatever service or solution you're offering. Finally, we're flexible, allowing businesses to customize their B2C interactions to their exact specifications, not their vendor's roadmap. That's a bit of what the opportunity space looks like. Now, let's talk about how we do value capture and get customers successful with our products. First, let's take an example of a contact center. If you look at the modern contact center, there are a few things customers are going to need, right? The first thing they need is they need it to be omni-channel. Right? We communicate with organizations through a variety of different channels today. With our Programmable comms cloud, we have support for every single one of those channels of communication.
After you meet the omni-channel needs of a customer, you need to add in support for automation and bots. Studio, which I'm going to show you a little bit later today, helps you build an IVR or a bot, but we also have support for natural language understanding with Understand and Notify for system alerts and automations. Next, you're probably going to need something for your agents, right? That's where our Frame product, which is a UI product that we built, comes in handy. We also have Sync that allows you to synchronize state between the customer and the agent on the other end. The next thing you're going to need is skills-based routing.
How do you get from the customer problem with a certain product to the agent that has the right expertise, or the customer that needs you to speak with a certain language with the agent that supports those languages? That's why we built TaskRouter, another component in the Twilio offering. Finally, you'll need things like workforce scheduling and a variety of other needs that enterprises have. That's why we built a marketplace. We have folks in our marketplace like Teleopti or Verint that enable you to add these capabilities to your Twilio-built solution. Our strategy is to leverage our strength in communication services on the far left and expand into application services over time.
Unique in this market is our low cost of sale, that allows us to lead with just a single component, getting customers successful with, let's say, a chat use case, and then allows us to add more and more components over time. We add those additional components, we're also building new components, increasing the portfolio of offering we have for customers. This component-first approach is exactly why ING selected Twilio for their contact center. ING started their contact center rollout actually with chat in the Netherlands. They had a problem verifying customers, and they needed to see them face to face, so they rolled out video in Spain. They added chat to France and Poland, and now they're doing voice in their contact center in the Netherlands as well with Twilio.
While all of this was happening, another team inside of ING started using Twilio in Australia for notifications. That's the power of our platform-based approach. You can adopt each and every one of these components over time and drive that strong net expansion that Jeff talked about earlier. Let's take another example of how our Engagement Cloud is helping us win business. Let's take account security. This is a space where we have our Authy product. Authy tackles the account security market with a packaged offering, directly solving a business problem of account verification and two-factor authentication. Why should you care as investors? With products targeted at business owners, you get a sales process that isn't just bottoms up and led by the developer, but it's also appealing to business owners in the organization. The result, example, like Pinterest.
Here's a blog written by their security team. The co-authors of this post are one of the security engineers, the person that did the actual implementation, and their head of security. They describe how they used Twilio's Authy to make this experience happen. The same thing played out with Namecheap, one of the world's most popular domain registrars. Our champion here was actually the CEO himself. He actually took to their blog to talk about Authy and their Twilio implementation, even before it was even live. If you even look at the quote in the blog here, it's really, really interesting. As CEO of Namecheap, "I'm going to give you my commitment that we'll deliver two-factor authentication within 60 days." That was just after a developer in the organization demoed a proof of concept of what they could do with Twilio. That's how confident they were.
As investors, Twilio's Engagement Cloud isn't just about good NPS and blog posts promoting our products from our customers, it's something we monetize directly. We priced our Engagement Cloud to be application services in addition to the communications that these products generate. That lets us capture more of the value that our products are unlocking for our customers, while continuing to capture the communication value. Great. Next, I want to talk about build versus buy. I've heard from investors that you have a great build option, but the large part of the market wants to buy. That's not what we hear from customers. What we hear from customers is that they want flexibility and they want fast delivery. It's this need for flexibility in every enterprise application service that has led that enterprises have been building for a very, very long time.
They use different words than build, but they've always been built. I'll just talk you through roughly what the current buy options look like. The current buy options usually involve buying some hardware. Usually that's oversold, plugging it into some other hardware that you got sold that was also oversold, and then you spend $millions and $millions on professional services and consultants to come in and customize the thing that you just bought. This is where the actual build in most enterprise software is happening. Then after you've bought something, done your build and customization, then you need to spend $millions more, usually in maintenance contracts, just in case you want to make changes to the equipment that you bought later on. This is what the enterprise cycle looks like today. Enterprise software has always been built.
The needs of the modern enterprise are just too complex to do anything other than build. This is why so many $ are spent on professional services and customization. APIs, though, are the right packaging for folks that want to build. Rather than packaging something up in a steel box, which is inflexible and expensive, APIs are flexible and easy to adapt to your business's need. When we think about this problem of build versus buy, we think about it in terms of customer needs, right? Twilio uniquely can give customers all of this while we continue to make it better every day. In fact, our latest product here to help organizations maintain flexibility while still allowing them to build is Twilio Studio. Twilio Studio is the visual builder for Twilio. We introduced this back in September at our SIGNAL conference in London.
With Studio, anyone who can write on a whiteboard can use Studio. You can maintain the same flexibility of our API-based approach, but now with the ability for anybody in your organization to use Twilio. Developers can use this to build something faster, but marketers and contact center managers can make changes after the fact. Since it doesn't require code, even though it's allowed, when we roll out new capabilities like speech recognition or natural language understanding, you don't need to wait for development teams for Twilio to be able to bring on these new revenue streams and customers to be able to take advantage of these new features. In fact, we recently used Studio to reprogram our investor relations phone number.
If folks have been to the investor relations website on Twilio, you'll see there's a Contact Us and there's a phone number down on that page. This phone number is actually today powered by Twilio Studio. What we're going to do for the next 10 minutes or so is we're going to reprogram our Twilio investor relations phone number live in front of everyone, through the power of Twilio Studio. Cool. If we switch over to the laptop here, we can take a look at what our current investor relations phone number is. You'll see there's basically three boxes here. You have an incoming message, and if you send a message to that phone number, you'll get a reply, which just says, "Thanks for contacting investor relations.
We'll be in touch shortly." This actually forwards the message to Greg so that Greg knows who has contacted the investor relations hotline. The second thing is if you call it'll say, "Thanks for calling Twilio investor relations," and it's actually going to go ahead and send that call to Greg's cell phone. Let's just give it a try now so you can see how it works. Pull up the phone and I will dial the investor relations hotline.
Thanks for calling Twilio Investor Relations.
You're seeing this flow. Greg's phone over in the corner there is ringing. If Greg were to pick that up, he would hear me talking to him. Let's go ahead and actually make changes to the investor relations hotline right now in front of everyone. The first thing we want to do, we want to actually add two new features to the investor relations number. It's great that we can talk to Greg whenever we want, but we want to be able to press a button and actually get a copy of the latest earnings. Then the second thing we want to do is be able to press another button and actually be able to listen in to the last earnings call. We're going to program that number to do that right now.
The first thing we are going to do is we are going to grab a gather here. What a gather does is basically listens for input on the call. It can listen for speech, it can listen for DTMF tones. We are going to go ahead and add that here. I have already written up this text, so you do not need to watch me type it, but I will just paste it in here. All this is going to say is, "Press one to talk to the IR team." We are going to make sure that that gets routed directly to Greg. "Press two to get a copy of the latest earnings release" or "Press three to hear our latest earnings call." I am going to have one setting here, which just says, once I get one digit, go ahead and route the call. Great.
That is first, and then we want to do is add a split. What a split will do is basically allow us to take whatever was pressed on that call, and route it in different ways in the application. I will want to go ahead and set the variable here, which will be the digits that were just pressed, and I will go ahead and save that. Now we want to add some transitions. There is, once again, three things that we want this phone number to do. One, connect to Greg, two, to allow you to get a copy of the earnings release, and three, to listen in on the last call. We will go ahead and say press one, and we want to route that to connect to Greg. We want to do two. We want to just put a placeholder there. We will wire it up later.
Then three, and once again, placeholder there. We can wire that up later. Great. There are two more things we need to do. Now if you press two, we want to make sure that it sends you the latest earnings. We are going to wire this up, and I have a URL of the latest earnings that I will take and paste into here. Great. One last thing we want to do here is actually let you listen to the last earnings call. We will wire in a Say/Play. You want to play a message, and this is a URL that we are hosting on the Twilio Runtime that is a hosting of the last earnings call. One last thing we need to do here is wire that up. Now everything should be wired up here.
We are going to do one more thing, which is we need to reroute this away from the current welcome greeting that just does the forward into our more complex greeting here. Let us go ahead and just refresh this and make sure everything loaded. Great. Let us call that number again.
Press 1 to talk to the IR team. Press 2 to get a copy of the latest earnings release. Press 3 to hear our latest earnings call.
Let's press 2, and we should get a text message of the latest earnings release. There it is. Cool. You can see earnings release comes through. It takes you right to the webpage. That was pretty neat. That's really, really neat. You can do that pretty rapidly, very quickly, super flexible. We talked earlier about the investments we've been making in machine learning and AI. The next thing we want to do here is we actually want to use Twilio Understand, our speech recognition engine, and actually allow you to talk to the Twilio investor relations phone number and actually do any of these actions as well. Let's go ahead and let's drag onto the canvas here, Twilio Understand.
I have a model that we built earlier represented by this SID, we want to make sure we take the speech result from that initial gather. Okay. Similarly, we need to pull out a split. This time we're going to split based on the intents our natural language understanding model has found. This is the model that we trained. Let's go do it based on our machine learning model's intent. Great. The names for the intents were Talk To IR, which is what will connect us to Greg. We're going to wire that up to Greg just like we did when you pressed 1. Now we're going to do the intent of get a copy of the last earnings call. If our natural language understanding model returns this, we're going to go ahead and send you that message that we sent before.
Finally, if we want to listen to the last earnings call, we will Route you to the earnings call. That is called Say Play. There it is. Great. One more thing to do is we want to make sure that we wire up this gather here, the user said something, to our machine learning model. Great. Let's refresh it, make sure it's been updated, and then give it a call. Great. Back in here, call the Twilio IVR hotline.
Press 1 to talk to the IR team. Press 2 to get a copy of the latest earnings release. Press 3 to hear our latest earnings call.
Listen to the latest earnings call.
Good afternoon, welcome to Twilio's Q3 2018.
Great. There you have it. We basically took a live working phone number, already running, we turned it into an IVR that was omni-channel, and then we added machine learning and natural language processing in about five, maybe it was 10 minutes, we got all the way through that. Why does that matter to you as investors, right? If you think about this, what would you have needed to have done in the past to have built something like this? Well, first, you would've needed to go get a deal with AT&T or a carrier to get connectivity. You then would've needed to get something from Avaya or Cisco to build the IVR and the phone tree. Then you would need to go find Nuance and train a model and license speech recognition from them. You'd need to put all these different things together.
It would take a ton of time, and it would cost a ton of money. With Twilio, you can do this very rapidly, and you can do it by each individual component. You can just add one component here, one component there. You don't need to have a one-size-fits-all solution to solve the problem, right? That's why we're so excited about our engagement opportunity. The ability to take all these different technologies, package them up as components, and let our customers adopt each and every one of them is what makes us so excited about the opportunity in front of us. Cool. Thank you all. Next up, George is going to talk to go-to-market and how we get these products in our customers' hands.
Thanks, Pat. It's great to be here today. My name is George Hu, and I am the Chief Operating Officer of Twilio. Prior to Twilio, I spent 13 years at Salesforce. My last three there, I was their COO, and I helped grow the company there from $20 million to over $5 billion. One of the reasons I'm here at Twilio is because I see the same type of opportunity, a massive market ready to be disrupted, an amazing set of technologies like you just saw from Pat, great visionary leadership team. I'm here to help accelerate this company's growth in the future. One thing that we didn't have at Salesforce, and is really, in my opinion, the secret sauce of why Twilio is poised for success, is the developer momentum, winning the hearts and minds of millions of developers around the world.
More and more, developers are important to the buying decisions, and especially in the area of customer engagement and cloud, which we'll talk about. This developer momentum allows us at Twilio to invest with tremendous efficiency and productivity in a sales motion to capture the opportunity in our customer base. I'll talk about how we're investing in that to capture that opportunity. I'm also going to talk about our competitive advantage. You saw from the demo, how compelling our technology is, how differentiated it is, and we're bringing that message every day and winning deals at massive rates over our competitors, at rates comparable or higher to than what I saw in my previous life. Finally, I'm going to talk about really that we're on a long-term journey here. A huge opportunity to disrupt a market.
Even right now, we're laying the foundation for the future with bets in areas that we think will pay off over time. We have a huge opportunity in enterprise, a huge opportunity in international, as well as with partners, and we'll cover all of that in the next 10 to 15 minutes or so. Let's talk about developers. This is really, as I said, the secret sauce of the Twilio formula. 1.9 million developers coming in, passionate about Twilio. This is not by accident. Twilio, one of the things I think about is that our self-service motion is actually a product. It's probably our number one product that we've invested to figure out how to attract, onboard, and make successful developers so that they go out there and can build projects, get them going, and then tell others about it. We accelerate that motion.
We pour fuel on the fire with in-person events. This is really something that I came to Twilio actually to learn about because I knew that Twilio had some secret magic. It's really about these 250-plus events that we do around the world every year. We're getting in front of developers where they are, at universities, hackathons, meetups, we're evangelizing the Twilio message and getting the technology in their hands. Then we're coming up with innovative ways to train them how to use the technology. At Salesforce, I saw the power of the Trailhead platform. At Twilio, we've just launched something called Twilio Quest, which is a different type of learning platform optimized for developers. You can see it's fun. It's designed like a game.
I think this is not only a transformational technology, in terms of getting developers to adopt Twilio faster, but I think that Twilio Quest has a similar potential to be a platform for learning for all types of developer technologies. I envision a future where companies like GitHub and Atlassian can be rolling out content on the Twilio Quest platform. All this innovation, everything we do from the evangelism, Twilio Quest, the sign-up process, results in just tremendous developer enthusiasm. You can see it here in the NPS, you can see it in the tweets, but I see it on the ground with the customers every day. Every time that I'm talking to a CEO, a COO, a head of a business line, CIO, almost every meeting, there's a developer in the room with me. The developer's the champion. I never saw that at Salesforce.
Salesforce, the developer's never in the meeting. At Twilio, the developers are there telling everyone why Twilio is the only choice. The good news for Twilio is that developers are becoming more and more important. This statistic, when I saw it, I knew it intuitively from being in the sales cycles, but when you see it written out like this, it's really compelling, and it really resonates as true to me, that today, over half the buying cycles for cloud developers are a key influencer. 22% of the time, they're actually the decision-maker. At Salesforce, we were really excited when we got line of business involved, that we were going to go around IT by going to line of business, but you can see here today, still for the most part, line of business is not involved in these decisions, but developers are.
I think we have something even more special at Twilio. How are we capitalizing on it? We're capitalizing it not just by working with developers, but taking that momentum and using it in a highly productive way to reach the business. When I came in, one of the first things I looked at was, what is the productivity of our sales team? It was honestly incredibly shocking, because I'd come from what I thought was one of the best sales organizations in the world, and I saw these numbers, and they were incredible. $1.5 million-$2 million of incremental business our reps were bringing in every year annualized for a ramped sales equivalent for us. Really incredible. That allows us to deliver the numbers we're delivering at tremendously efficient sales and marketing investment, 23%.
A number that's far below what I saw in my previous life. Very compelling. Over time, as we make these customers successful, as we grow, as we cross-sell, up-sell, as they make their own organic investments in their business, we see this tremendous dollar-based net expansion rate, 137%. It's interesting, when I joined Salesforce early days, I noticed two things. One was that there were a lot of naysayers about the model. I remember first time I talked to Gartner, they said, "Oh, no CIO wants to put their data in the cloud. That's insane." The second thing I noticed was that after we went public, that a lot of people just didn't understand our business model. Our stock went up, it went down.
People asked all sorts of crazy questions that showed that they didn't understand our business model. It's taken me a while, actually, to understand the Twilio business model, but I'm incredibly impressed. I think it's just as disruptive, that's why, because of the numbers I'm seeing, we've made the decision to, as we've grown our sales and marketing investment, we've decided to especially focus on quota-carrying capacity. When I joined the company, we had a meaningful go-to-market team, but the vast majority of the resources were not quota carrying. They were focused on other roles like operations and sales engineers and specialists and people that answer the 800 number, but not actually a lot of people out talking to customers and selling the product.
As we've grown, I've focused on most of our incremental resources on that quota-carrying number. What's really exciting is that that productivity you saw on the previous slide was from our recent Q3. This is a number that productivity has held steady even as we've invested. We're growing aggressively after all segments of the market. We think there's an opportunity in every geography, every company size, every industry. The future is limitless for Twilio. We're going after it with a very structured approach. Commercial accounts, we're mostly selling over the phone, less than 1,000 employees, very productive motion. We're also seeing a very successful motion in the enterprise. This is an area of much greater investment for us than it's probably historically been in the past. We're investing in both segments at the same time. Again, I'll tell you a story.
When I joined Twilio, my second week, one of our enterprise reps grabbed me in the elevator. She said, she was one of our West reps. She said, "George, you need to help me. I'm drowning in opportunities." Jeff walked by the elevator at that moment. I said, "Stevie, get out here right now and repeat to Jeff what you just told me." She goes, "Jeff, I'm drowning in opportunities." That gave me, that plus the numbers, gave me the conviction that we can invest to capture this opportunity. When you have a champion in the account, the developer saying, "Not only do I love Twilio, but let me show you it's actually working." What Pat demoed for you, that's what our customers are seeing from their developers. The product actually works. You cannot underestimate the power of that.
We're tackling the enterprise segment with a similar motion, and we have our strategic accounts, which we take special care of. These are our most largest and most highest-spending customers, our most important customers, and we want to really make sure we focus on them with also a special focus on solution partners. I think that the Twilio platform, if there was ever a platform that was ready-made for solution partners, it's Twilio, and we'll talk a little bit more about that. On the direct side, what's really interesting to me was that before I came here, Jeff says, "Ask your developer." Before I came, I actually asked my developer, I said, "What do you think of Twilio?" They said, "We love Twilio." These were developers mostly for around here, tech companies, and I thought that Twilio was mostly selling to tech companies.
The good news is that not only is Twilio successful with tech companies and digital economy companies, but to my pleasant surprise, we've been closing transactions in every industry, and with amazing brands, and you're going to hear from Naureen at Morgan Stanley, and healthcare like Blue Cross and Blue Shield, and retail like Lululemon, and real estate, and on and on. The future, as I said, is really limitless for this company because just like what I saw at Salesforce, we can sell to every company because guess what? Every company in the world is thinking about, how do we engage with our customers in a whole new way? I believed that there.
I believed the future was limitless there, I think that company has continued to prove that there is almost no bounds to the appetite for companies to engage their customers, and I see that same opportunity at Twilio. I want to take a second to talk about solution partners. Why are we so successful with solution partners? I realized this, that almost all of these companies, like the company I came from and all the companies we thought of at the time, like Workday, et cetera, and you can see many of the similar companies on this slide, like ServiceNow, Zendesk, almost all of these companies actually were designed from the ground up as systems of record. A lot of the people that started these companies came out of Oracle and PeopleSoft and companies that thought that way.
Not very many of them thought of themselves or were designed by people that think in terms of systems and engagement, and that's what Twilio really is. I think we have the opportunity to really partner so well because we provide a capability that companies that think like systems of records don't really know how to do. It's not really intuitive for the co-founders of Salesforce to think about how to send a text message or make a voice call. It wasn't in their consciousness. We do that. That's how we think, and that's why we're such a great complement for all of these systems of record, and I think that puts us in a very fantastic opportunity position. Let's talk a bit about competitive dynamics, because I get asked about this sometimes.
I can honestly say that I've never spoken to a customer that didn't tell me that Twilio was not their first choice, was the brand that they wanted to go with. You can see in the numbers here that we're winning, honestly, well over 80% of the deals that we compete in against competitors. What's exciting to me is the number that's really not on this slide, which is that in a very significant percentage of our transactions, we actually don't even see any competitors, which is kind of surprising to me, coming from my old world. The reason is that because of all the developer sign-ups, when the developers sign up, they get going, they're successful, they're using the products built.
You're not really at that point looking, you're not doing a bake-off, you're not doing an RFP, you're not bringing in four companies to try and compare prices with this. You're just figuring out, "Hey, I've started using Twilio. This seems great. What else can I do in the platform?" That is really the heart of the conversations we're having. They're picking us today because they trust our service, like Christine talked about. They love our omnichannel capabilities. I honestly think we're years ahead in terms of our breadth of channels and the depth and quality of our channels, as Ott talked about. They choose us over companies like Cisco, Genesys, and Avaya because of our flexibility.
What you just saw Pat do, there's no way you're going to do, you'd have to bring an army of Cisco engineers in to do what he just did in 30 seconds. The innovation of the Engagement Cloud, because companies generally want a platform that they can get excited about because they have unique requirements, they want a platform that can give them those requirements, they want to believe in a company that's going to innovate in their future. So many companies I talk to say, a great customer of ours, Centerfield, one of the first transactions I participated in when I joined Twilio, I was talking to Gary Pack, the CTO, and he said, "I'm betting on Twilio because I believe in your innovation.
I've seen how you've innovated in the past, and you're the company that I want to bet my future on." That's really important. Now, how do we get started? What does the progression of a customer look like? Here's a great example. Here's a great example of a new customer. We're landing a customer for the first time. It's a Fortune 500 medical devices company. They're using us for one of our most popular land or initial products, which is Programmable Messaging. Why did they pick Twilio? You would think that they could do this and get this from a lot of companies. A lot of companies say they have an API now in this space, but why did they pick us?
They picked us because it's an important use case, because it's a use case where they're getting messages from glucose monitoring machines and giving them to caregivers. Certainly not every use case is this sensitive, but I can say that the vast majority of our customers I talk to, reliability matters, deliverability matters. The things that we invest in matter. That's why we're winning these deals, we're winning them at a premium, and we're selling them a set of products, typically in the communications layer, to get started. Then over time, we want to start progressing them and evolving them with application services. That's where things start to get more fun. We start to talk about things like contact centers, call centers.
Here's a great example, National Debt Relief, where I think the linchpin technology for us was one of our application services, TaskRouter, which allowed them to do some very creative skills-based routing that they could not do with the legacy on-premise technologies. Daniel Tillman, who I met a few weeks ago in New York, who is their president, their CTO, co-founder, very much into technology, but also understands the business requirements, picked us because not only could he get the wide array of unique requirements he had met, but also because he didn't have to trade off cost and speed to get the flexibility. In fact, he got better cost and better speed. I hope you got that point from Pat. That is a really magic triangle. Usually, you can get one out of the three or two out of three.
Usually, you have something that's inexpensive, fast to deploy, or flexible. We deliver all three. I think that being able to do something unique because of the power of your technology model is something I saw at Salesforce, and I'm really excited about seeing it here at Twilio. Even that's not the end game. What's exciting is when a company says, "We want to go all in with Twilio." This is our first enterprise agreement that we closed last quarter, and it's a huge transaction for us. It involves every one of our products. This company has been a customer of ours since 2009, and they've started out that journey. They started with one communications channel. They expanded over time. They started using technologies like TaskRouter. Now, they want what Pat demoed. They want the entire Engagement Cloud vision.
They want to have complete flexibility because they view their contact center as a strategic weapon in how they engage their customers. They want to have the best, and they believe that this is their differentiation in the future. So we are focused on making every customer like this customer. That doesn't happen by accident. We have to invest to make sure that that, and be smart about making that net expansion rate go up. So how do we do that? At a most fundamental level, it starts with talking to your customers, engaging your customers. We preach customer engagement, and one of my priorities is engaging our own customers. Today, over 90% of our existing revenue is actually touched by a dedicated, or designated account team. That's really critical to get those conversations going. That's how you get the flywheel to continue.
We also use our own data to alert us to potential high-value customers that are signing up in the long tail, as well as any customers that are using any new technologies, thousands and thousands of alerts a month coming to us that we're working through to make sure that we don't miss any potential signal of a buying opportunity or even a potential issue that we can resolve within our customer base. What's also exciting is that our customers are not just using us for simple use cases, one-channel use cases. They're buying from multiple product families. 90%+ of our customers today are using products from more than one product family. Really, really exciting. I think we're just at the beginning.
We're just at the beginning because, as I said, I view this over a long period of time, while I love our developer motion, I think that the future is for Twilio really to be the first company that I think effectively brings together both developers as well as line of business and IT together in the conversation. We're going to start marketing that way. We've kicked it off with our new Twilio Engage Roadshow. This just is an example of the kind of marketing and kind of programs we're going to be doing in the future, bringing developers and business together, having a broader message around customer engagement, not just a tactical message. We did one in New York City, it was this past week, was tremendously well-received. We're going to be doing one in Melbourne soon.
I expect to replicate this program based on the initial results, throughout 2018. I think that we're just really at the beginning of the potential of making Twilio, at the end of the day, just like at Salesforce, we wanted people to think, if you're thinking about Siebel, if you're thinking about Oracle, if you're thinking about SAP, you need to be thinking about Salesforce. At Twilio, the way I look at it is, the shorthand I want to plant in the mind of every company and every CIO and every decision-maker is, if you're thinking about doing anything with Cisco, with Genesys, with Avaya, you need to be thinking about Twilio, that Twilio is the heir to those technologies, just like Salesforce is the heir to those other legacy companies.
I think that just based on things like that demo you just saw, every customer we can get in front of gets it, we've just got to do a better job of just continuing to execute that. We've got to also lay the foundation for the future. As I said, when I was at Salesforce, we grew it all the way, well, now through my time, it grew to $5 billion. Obviously, they're far beyond that now, the current management team, Marc and Keith, are doing a fantastic job there. It takes time to build a great company, I really believe that. You got to make the bets now because things take a long time. We're investing in the enterprise. You've seen that, some of the logos. We're building the sales team. We're doing the events, the marketing.
We're laying the foundation in international now. We're putting our developers, we're moving our developer evangelists out into the forefront of where our future is. Initially, most of our developer evangelists were in the U.S., and we still have a lot of them here, but we're starting to move them into places like Melbourne, for example, that event I talked about. Putting people in the frontier countries, having country managers focused on them. I'm particularly excited about partners, and you're going to hear from one of our partners on the customer panel coming up soon. I think that there's a huge untapped opportunity for Twilio in the SI channel, for example. When I was at Salesforce, I worked on building the Accenture relationship, the Deloitte relationship. Those took years to build, but the payout was huge.
All you have to do is, if you attended Dreamforce, if you looked at who are the premier sponsors of that event, it's Accenture, it's Deloitte, it's IBM, it's PwC. They've built that ecosystem, I think we're even more better positioned, we're even better positioned for that channel because those channels are all about delivering customized solutions to deliver unique business requirements for customers, and Twilio is really tailor-made for that. We are fundamentally functionally agnostic, channel agnostic. We can be used to deliver whatever dream a customer has and do it with a business model that is incredibly attractive for them, with our pay-as-you-go model, you can get started and try and grow over time. Very, very compelling. As you can see, I'm just very enthusiastic about being here. I see a lot of parallels to my previous opportunity.
I see the same type of upside. I think it's a little bit different, obviously, different in a good way. Different in a good way because we have this developer motion, something that we just never really had it exactly the same way at Salesforce. A huge opportunity through developers. People talk about software eating the world, I believe that, I think that developers are going to become more and more important. A bet on Twilio, in my opinion, is a bet on the power of developers, the bet on the future of customer engagement, the bet that communications is going to become ever more important in a world where we have bots and artificial intelligence. Who knows what the future will bring. I think we're positioned for that future better than anyone. I think we have an amazing competitive advantage in terms of our innovation.
We're trying to build a world-class go-to-market team to capture that opportunity, I couldn't be prouder to be here. Thank you very much.
All right, we're going to take a 10 or 15-minute break to assemble the customer panel. We're running ahead of schedule, we'll try to start at about 3:00. Thanks.
All right, let's get rolling again. Everybody, come on back, take your seat. Thank you for staying here and snacking out there. We're going to close it off now with meeting some customers, and then, of course, what you're all waiting for, hear from the CFO to give you all the juicy bits. I want to introduce two awesome customers who agreed to come and chat with us today and give you their insights into Twilio and why and what they're building on Twilio. Please welcome Naureen Hassan, the Chief Digital Officer of Morgan Stanley, as well as Ryan Nichols, the head of Zendesk Talk, which is their contact center product, both building on top of Twilio. Let's give it up for them as a thank you. All right. Awesome. First of all, let's start off.
Can you just give us the basics, like your name, the company, your role, what you do? Actually, I think I just did that.
You did that, Jeff.
Tell it better than I did it. Naureen, why don't you go through it?
I'm trying to think of how to say my name better than Jeff said my name. I'm Naureen.
You have more experience than I do.
This is true. Naureen Hassan. Morgan Stanley has two major divisions, the banking side, which you all know, and the wealth management side. The wealth management side, we are one of the largest wealth management businesses with 16,000 financial advisors around the country, and I am the Chief Digital Officer for that business.
Great.
My name is Ryan Nichols. As General Manager of Zendesk Talk product, I work with our customers on how to deliver an omni-channel customer support experience. How do you offer phone support in the same place that you engage with your customers over all sorts of other channels?
Great. Thank you. Let's start off just at the very beginning of time for each of you. How did you discover Twilio? Naureen?
Sure. At Morgan Stanley, one of the advantages of having the banking side is they organize for us something called Tech Week, where the rest of us go and meet new potential partners, startups, and other vendors. In 2014, Jeff blew us away in one of the Tech Week partnerships just on the capabilities that Twilio had. Everyone was really impressed with the ease, the capabilities, that's how we first heard about Twilio.
Great. What about you, Ryan?
Oh, for us, it was developer up. We had a developer and a product manager attend a Twilio event and came back to Zendesk, and they'd made the Zendesk help center ring, which we thought was pretty cool, kind of blew everybody away. We said, "We didn't know it could do that," and decided to productize it. Over time, as more and more of our customers, thousands of our customers, started to adopt this, we started to look at using more and more Twilio to build a real call center inside of Zendesk.
Awesome. Walk me through the very beginning. You discovered Twilio, and you sort of got this spark, like here's what's now possible because of Twilio.
Yeah.
Your developers started experimenting with these ideas. What, ultimately, business problem were you trying to solve that became the impetus for what then turned into the solution that you deployed?
Yeah.
Naureen?
Financial services, we have a lot of unique regulations. One of them is that we have to record and supervise all communications that we have with clients. With email, it means we just store them in massive files, and when somebody sues us, we present it, we read everybody's emails. Text messages provided a problem. How do we record what people are doing on their phone? We had two different solutions that we were looking at. One was give everybody a corporate phone, lock it down, and supervise it. It's a very expensive proposition when you have 16,000 financial advisors to buy that many phones and manage them. The other, the benefit that Twilio offered us was put another app on their existing BYO, bring your own device.
We put MobileIron on it, then with partnership with Twilio, developed this MS phone app where they could have a work number that rings on their personal phone. They could have text messages to their clients that we could record and supervise, because otherwise, we were telling financial advisors, "You are not allowed to text your clients," which in this day and age just isn't feasible.
Yeah, for us, it was really about seeing in our customer base, of 100,000 customers, 80% of them offer phone support as a way. It's a very popular support channel, obviously. It's how people want to engage with the companies they're doing business with. Most of our customers are offering phone support in an old-fashioned way. It's a siloed legacy technology. It's not connected to anything else in your business. When we wanted to bring that together with other support channels and deliver a more modern customer experience, there's a lot required to do that. You have to go get the phone numbers, you have to get the calls into the right agents. You have to make sure it's a really high-quality phone call and that everything's tracked in the right spot.
There's a lot to build out in that overall experience that was well outside of our core expertise, and that's where we realized we needed to partner in order to build out our solution.
Ryan, Zendesk has showed these great stats about, well, you started in email. Why did you go into voice? Well, it's because customer sat, and your customers actually demanded it of you. Is there any stats that, I know you've shared those in the past, about how customers love voice as a channel?
Yeah, it is the most popular support channel. What we're finding is that as channel preferences change and digital channels increase in popularity, it's on top of, not instead of, the core phone calls. We see the nature of the phone channel changing. People are using phone as an escalation channel. They're using phone for those high intensity, high emotional, super high priority moments. In some ways it's the moment of truth. When you actually have a customer on the line, you're either going to lose them in a sea of hold music while they're waiting to get in touch with the right agent, or you're going to make a satisfied customer.
Yeah.
Delivering a really good experience during that moment is really important to us.
Yeah.
I remember several years ago being really frustrated with a drone I had bought. It was exactly in that scenario. I was calling to get support, and when I called the agent, I heard the familiar doot doot, like Twilio Client sound.
They were a Zendesk customer, which is exactly what you're saying play out. There's that moment where you're really frustrated, and that's when delivering that superior experience, that ability to escalate and the value of omni-channel really becomes present.
When an agent can answer the phone call with the, "Hey, Jeff, you're probably following up on the email you sent us this morning. Sorry we haven't gotten back to you yet." That's a magical experience when you can deliver it.
Absolutely. Naureen, you had a great story. You said a lot, you had this problem you wanted to solve.
Yeah.
Of we've got all these private wealth managers who today say, and the client's like, "Oh, can I just text you?" They basically have to say no or be like, "Yeah, but don't tell anyone.
Exactly.
As a client of private wealth myself, I would say that speaking to bankers is something that if you can keep it to as brief a possible of interaction is of high value to someone like myself.
Jeff.
Every small question, they do want to schedule an hour-long phone call with you.
Yeah.
Text messaging. People know what I'm talking about. The idea that you want to use the right channel that customers prefer can literally make or break the relationship. I think that is extremely important. You had mentioned, you had this idea, like, is this even possible? Could we do this? The amount of time it took you to answer that question was actually pretty astounding.
Yeah. Our tech team worked with Twilio and we did a POC. One of the reasons we selected you as a partner for this, they did a POC in four hours.
Four hours.
Four hours to get it up and running. I often get asked, "Why did you guys choose Twilio as your provider?" One was you were able to manage through all of the security architecture, vendor management, and other things that come with serving a globally significant financial institution, but also because it was so easy to use and to integrate, right? If we can do a POC in four hours, we can get this thing up and running fast, and that was a key factor in our decision.
How long would it have taken the other things that you considered? Did you consider other things? If so, what were those options?
We did. We considered some mobile virtual network operators, other things that just They would've been longer than four hours. I don't have the existing stats, but the speed of being able to get your solution up and running, the API-based nature of it really was like, you got the big billboard, or you had it on the way to SFO, ask your developers. We asked our developers to say, "Which of these do you want?" They wanted Twilio, there we go.
What about you, Ryan? What else did Zendesk consider as an alternative here?
At the beginning, nothing, right? It was literally a decision that was, "Well, this works, so let's run with it." Of course, since then, we periodically look at options and evaluate our strategy for communications platform. Boy, I tell you, there are a lot of mornings when I wake up and I'm very glad that I don't have to manage a team to negotiate for additional toll-free numbers in Germany or manage mobile things in Italy.
It's a little interesting.
Absolutely. We get a whole lot more from Twilio than just phone numbers and minutes. As we consume more and more of not just the Super Network, but also your Programmable Communications Cloud, that's when it starts to say there really isn't a way of assembling other services from competitors, and it's certainly in a business that we want to be in on our own. As our partnership has deepened, it's really not something that we spend a ton of time thinking about.
What are the most important things you think your customers value in how we're jointly bringing this to market?
Well, I think that the global reach is a really important one.
Yeah.
Zendesk is a very global company.
Same, yeah.
We need to be able to support the needs of our customers to engage with their customers everywhere. At first, that was a challenge, right?
It was. It's actually worth. Stop me for a second. I remember back in, this was 2012, 2013, very early in the days of our global footprint for our Super Network when Zendesk, because we were founded in Denmark, I feel like there's always been a very global nature of your customer base. You pulled us way faster than we expected to become a global company.
Well, thank you.
In particular, phone numbers. This is one of the interesting things because you naively say, like, okay, in the early days you say, "Okay, well, we have phone numbers in all these countries. Great. Problem solved." You realize, okay, you have the phone numbers, but now the media is actually traversing the Pacific, round trip between your customers in Australia, going back to Virginia, and then back to Australia, and introducing 10 seconds of latency. Well, that's not going to work, right?
Yeah.
The more you double-click on it, you're like, well, the carrier partners have to have POPs in the right places on every side of that equation. We have to have all of our media services with POPs in all the right places to be able to service that traffic. This naive view that you just have the phone number and the problem is solved, when we worked on this together in the early days was like we were both out there chartering new territory. Because if you think about it, the notion of a Super Network that spans the whole world, that allows you in real-time to provision a phone number somewhere in the world, then call someone somewhere else in the world, and have everything just work, hadn't really been done before because you had had this geopolitically bounded industry, telecommunications.
With every country had its own carriers, companies would essentially put infrastructure in closets in those countries when they opened up in that country. You kind of did this, and it took years and decades, actually, to build up the global presence, and we were trying to do it essentially overnight.
Absolutely. When we did that launch, it was like, "Well, of course, we'll just launch globally." To fast-forward from those days to where we are today, where as we approach 100 countries.
Fast.
all those countries, it's amazing. At this point, it's not something that our developer team needs to think about at all. When I just think about the amount of things our team doesn't need to have expertise in, it's, again, those are the mornings when I wake up and I'm like, "Boy, I'm glad we don't need to think about this.
Naureen, were there other considerations that you thought about when you chose to move forward with us? You had speed, this proof of concept in four hours. Were there other considerations that you thought about from the enterprise perspective?
The global network. The fact that you guys did that with them first, thank you very much.
Thank you.
The global network, because we're such a global company, was another important consideration for us. Then also, I'd say the power to innovate and push our thinking. We realize that particularly, look, financial services is not at the bleeding edge of the latest and greatest. We need our partners to continuously bring to us new ideas, push our thinking, innovate, ask questions, and say, "Here's how else you can evolve your offering with our tools and what else is there in the marketplace." That culture and that ability to partner with us was another key factor.
Fantastic. What about now? What are the key things that we're doing for you now that you think are most valuable?
We have a vision. One reason we've been very public, I think, on our digital strategy is we think it is far broader than what the press kind of publicizes digital should be about. One of our core pillars is we have 16,000 financial advisors. We want to turn these guys into And they are mainly guys, but these men and women, into digital marketers. Right? If we can give them the tools that normally, somebody in a brand marketing group in headquarters is using to sell soap, if we can give those tools, the communication tools, the data and analytics, the intelligence, and put them in the hands of 16,000 people who also have the relationships in their local community, we think it's a really powerful way to grow our business. We're investing a lot in data and analytics.
It just can't be data analytics to the FAs. We also have to make it easy for them to get the messages out, to be able to communicate with clients at scale in the way that clients want to be communicated with in this day and age, texting, email, video, multimedia, and expanding beyond where we are today. That's where we're looking to partner.
We chatted, I told y'all earlier, the way we're seeing this organically rise out of our customer base is that there's this full, essentially customer life cycle.
Yeah.
Right? That often starts with acquiring a customer and then operationally communicating with them.
Yeah.
there's a support, there's marketing,
Yes. Service
there's sales, there's service, right? There's all these touch points that companies have, and it sounds like what you're saying is, you're going to start with one. We see this very commonly in our customers. We're going to start with one.
Yeah
of those touch points. There's the most important one, the one that's got the highest priority. We're going to start there, then after we make that successful, then we're going to start moving on to these other customer touch points. It sounds like you've started with the sort of one-to-one use case-
Yeah
of how do I connect the wealth manager to their client for those transactional messages.
Yeah.
Then that's just saying, well, great, now how do I take the marketing part of that life cycle?
Yes.
How do I empower them?
We're doing service as well. We think of it in the same way. We have a little chart. It's acquisitions of how do you acquire new customers, how do you then provide them advice, but that can't just be a meeting in a room where you have to go over pieces of paper. It also has to be, the day that Brexit happens, a text message that says, "Don't worry. Yes, the market's down blah, but your portfolio's fine because we diversified it." Then servicing, right? Because who wants to call in? You want to do it on your phone. You want to communicate back and forth on your phone. Then, relationship management and that engagement piece. We think of it the same way. We didn't even collude on that answer.
Yeah, no, it's something organically we see from so many of our customers, that whole life cycle comes into play. The other thing I think that's interesting about your business, and I think you've expressed this in the past to me, is that you have not just one client here, right? You've got the end customer-
Yeah
whose business you have to earn and keep-
Yeah
Satisfy. You also have the wealth managers.
Yes.
They can go work at other firms.
Yes.
Right? They can take their book of business and move it to a competing firm, the firm that provides them with the best tools-
That's right.
the best infrastructure to run their business and ultimately make them successful, has also a competitive advantage.
No, that's totally right. That's why we want to be the most modern platform out there, the easiest to do business with from both an FA and a client perspective. We think partnering with Twilio and other firms like this provide us a competitive advantage in the recruiting wars for financial advisors as well, in addition to just what we call the end client loyalty and retention.
Awesome. What about you, Ryan?
Well, I think, where we are now, it really is about how do you empower a developer team to make the same sort of evolution of going from one channel to that same developer team looking at SMS, and that was something we recently added to our platform, I guess, just over a year ago. It was a similar sort of transitioning from, "Hey, let's start with one channel. Let's look at other channels." We've been using a lot of Programmable Voice, Programmable SMS now. Looking at some of these higher level services that we want to bake into our offering to free up our development team to work on other tasks. For example, Voice Insights has been something that has just been a game changer for our support team, right?
To be able to work with a team of call center agents and be able to identify that, well, it's probably that team of agents that you have down in the basement floor with the low quality Wi-Fi, that's where you're dropping calls. To be able to change that element of our support experience, it just allows us to use more and more of the Twilio stack to enhance what we deliver to our customers.
Yeah, Insights is an interesting product. I actually think some of the early work we did with you really led us to understand the nature of this. Your developers go, you build a solution, you say, "Great, it's live," you market it, customers start using it, then they start writing into support, probably have used Zendesk. They say, "Hey, this call sounded bad," or, "My agent's here and dropped a call. What's going on?" The support team's like, "I don't know. What do I do? How do I deal with this?" What it ended up with was a bad customer experience for your end users because they couldn't get answers to their questions.
We built our Insights products, it's like x-ray vision into what's happening in Twilio's network, as well as the end user's network, that allows your agents to sit there and say, "Okay, great, I pulled up that call. I can see that there was a lot of drop packets at the edge. I can see that there was a lot of latency. Have you thought about this? Have you thought about plugging in with Ethernet?" Whatever. This is one that we discovered early on that we didn't know about, that it was some astounding number, it was like 30% of calls where customers would complain that there was no audio. It turned out their headset came unplugged from the computer. Right? You had your agents were sitting there being like, "I don't know.
I don't know what to tell you." Part of our Insights product, we actually had a thing that detected, yeah, there's no audio coming in from their microphone. Right? Then they would say, "Oh, yeah, I forgot to Sorry." Right? It was all of that early work that we did collaborating with you and your team, and understanding, and sitting down with your support agents that showed us the roadmap to say, "Yep, you need this Insights product in order to actually support these products at global scale like your team is doing.
Well, it turns our agents into superheroes, right?
Into experts, yeah.
It just allows them to have that insight and deliver just the right suggestion that makes the customer successful.
Yeah, it's also like there's all these different folks inside of Zendesk now who rely on Twilio to doing their job.
Yep. Not just developers anymore, right? It's our support team. Actually, as we start to look at, we're going to talk about this a little bit, just how early we are in this overall move of the call center into the cloud, and how when we look at the variety of organizations out there and what they're looking for, some of them love the, you flip a switch and you're able to offer a call center that's built right into Zendesk. Others want to have the time and the effort and want to build something that's specific to their organization. Most of the market's somewhere in between, right? This is, I think part of our ongoing partnership is figuring out how to serve that in-between market together by partnering. I think there's a lot of opportunity to do that.
When I look at some of the stuff that you're working on that's in the Twilio Engagement Cloud, I'm so anxious to build that into our products so that we can have our customers bring their own Twilio to Zendesk. I think there's a lot of opportunity to do that.
Yeah. We're excited to do that. We've been working together for a number of years now. Ryan, how have you seen the relationship evolve over time?
Well, it's just that from a convenient spot to purchase numbers and minutes and offer this service very quickly and easily, to, hey, this is a core part of our strategy for delivering a solution for omni-channel support to how do we partner together to accelerate this really once in a generation shift of this legacy industry into the cloud. I think that that partnership is kind of the next wave of our relationship together, going and tackling the market together.
Awesome. Now, Naureen, you answered this a little bit, but I'm curious if you have anything more to say. This notion that, as far as working together and the nature of the relationship with Twilio, you guys were very public about it, right? You did a press release. This is at the C-level. I remember I tried to get you to come to SIGNAL.
You said, "I'm actually presenting to the board-
Yep
of directors of Morgan Stanley. Sorry I can't be there." I'm like, "Poor excuse, Naureen.
He tried to get me to move the meeting, and he actually asked me why I couldn't video into our board meeting. You did.
I did ask our team how much a private jet would be, because I thought it was a good story, although there's no way Galardi would've allowed us to actually do that. Tell me, what was your strategy, right? You alluded to this, but what was your strategy for being so public about it? Because I do feel like there's a very intentional approach that-
Yeah
Morgan Stanley is taking when you talk about your digital disruption that you are driving.
I think the press, for those of you that follow financial services, has equated digital with robo-advice and created this storyline that technology is here to replace the human advisor. We wanted to change the dialogue. We fundamentally believe that, we think automated solutions have a role for investors, particularly earlier in their life cycle when their needs are simple. I think it's a great solution for an underserved market. We rolled out one of our own. That there's another opportunity out there, and that people need human advisors when their needs get more complex, when they have to deal with stock options, trusts, taxes. That there's a big digital opportunity in changing that segment of the market that isn't discussed, that needs to be discussed. That's about, you don't have to meet with somebody in person to do everything. You shouldn't have a fax machine.
We should leverage data and analytics. We should change communication with clients. We wanted to have that discussion. Part of our intentionality was let's shift. Let's change the topic here and talk about the big opportunity, which is in the core, in the bigger path of the market here.
Here's what I heard. There's an opportunity to sell them Twilio Programmable Fax. George.
Oh, yeah. Trust me.
That's what I heard. That's the next big opportunity with Morgan Stanley.
You should see our branches. There's a rumor going around that the last typewriter might have only left a few years ago. I don't know. We have been engaged, and digital is so important to Morgan Stanley that I present at either a full board or subcommittee every meeting on our progress and the partnerships and how we're making this forward. It's a big priority for us. Thank you.
Absolutely. Last question here for each of you. How are we doing, and what could we be doing for you in the future? Ryan, start with you.
I'll start with that. I think one indication of how much our relationship and how much our confidence and our customers' confidence in the Twilio service has improved over the years is how the concept of powered by Twilio is something that we didn't use to say, mostly because it was just, it would be confusing to people or they wouldn't understand it, and that is changing, right? I think that phrase is something that now builds confidence in our customers and that our sales team uses to express to our customers that there's carrier-grade infrastructure behind this service, and that it's a signal to our customers that we're going to be moving very quickly in changing how they do phone support. I think that that's as good an indication as any that you guys are doing really well, right?
That we literally wouldn't be in the business we're in with call center software, because there's so many areas of it that are just outside of our area of expertise, right? I think the partnership is really enabling us to help our customers change how they deliver this channel. I think that going forward, I love the idea of helping our team, as you guys move up the stack, helping our team move up the stack as well and solving more and more interesting challenges that our customers are asking for. We want to be spending our time in making voice, not a support channel that's offered over there by those agents that have special kit and specially trained in that area, but as a natural Just like a human conversation, right?
That might flow from an email that I get from my wife during the day coordinating plans for dinner to an SMS that I send her as I'm heading to the car to a phone call on, "Hey, can I pick anything up?" Companies can't have those sort of conversations with their customers today, and it's the technology that's getting in the way, and I think that we can partner together to change that.
Awesome. I know you've also asked for Studio.
Studio, for those of you who don't know, is just a beautiful kind of drag and drop IVR builder. It's one of those things, right, that this was my first request. When can I just put that in my product, right? I want to make that available to all my customers, that they can be building these incredibly sophisticated IVRs on their own.
That's one of the things that we focused on as a company. Everything that we build, we build with our solution partners in mind that you can take this component and then take it and embed it in your product and make your product better and just sort of rise the whole tide here for all of the cloud-based contact centers that are built on Twilio, that together, we're going to go and take this big market that's out there, this 90-plus % of contact centers that are still legacy on-prem, gathering dust in a closet somewhere, that we can all go tackle this huge opportunity together. Very cool. What about you, Naureen? How are things going, and what could we be doing for you in the future?
Things are going well. Our rollout has been very popular with our financial advisors, we've gotten great feedback. What I would say is just keep pushing our thinking, right? Our industry is the noted laggard, you guys work with all these different firms, push our thinking, give us ideas. We're eager for them. We're so heads down in so many things that we're trying to do. Like this IVR one's a great one. Our IVR sucks.
Pat can build one for you in about five minutes.
There you go, right?
If you want, he just did it.
Oh, yeah, multimedia video. Bring us new ideas because we're hungry to absorb them, but it's hard to pick your head up sometimes and see what else is out there.
Great. Well, we'll come by and work with the whole team on that ideation.
Yeah, that'd be great.
That's something our team does really well.
Yeah.
Very cool. Thank you so much, Ryan, Naureen. Thank you for sharing with us today. I think we've got time for maybe a couple of questions, not too much. There's a mic out there.
Thank you. Mark Murphy with JPMorgan. I'm wondering if I could get your views on just how big is the feature gap between Twilio and its competitors and maybe the reliability gap, between Twilio and its competitors, if you've ever had a chance to look at that. Secondly, how would you respond to someone who might say that SMS text is really not the future, it is the past, and we're going to be living in a world that is in-app and push messaging? Because I do believe that SMS text is a big part of your push, correct, Naureen?
Let me answer that. We are investing in all of them. We are investing in in-app, but SMS text, particularly for your one-to-one, right? From our perspective, it's a financial advisor talking to a client, which is a one-to-one relationship, and we are taking that from meeting to phone. To take that all the way to in-app is, like right now, our penetration of even people who use our mobile app, I mean, the numbers are shockingly low. We're just not even close. There's a huge market for us just in SMS right now, and then, we'll go from there.
Mobile apps, I mean, we're eight or nine years into this mobile app revolution.
Yeah.
I mean, this is not new.
We have a great mobile app, but we have a penetration problem.
Yeah.
If you think about both our FA client base, right, we have a two-step sell. We have the FAs, and then we have the clients, and so that's a long process to get through that.
I also love the user experience. I mean, you have this experience all the time with banks in particular, where they send you an email to tell you that you have an email waiting to be read.
Yes.
Right? Like, it's so annoying, right? SMS is just SMS.
Yeah.
Like I communicate like I would with my friends or my family.
Right
not get an email saying, "Click here to read your email." I think it's a great user experience there.
I'm happy to address the question about kind of features as well as quality and reliability. I think that when you look at the whole Twilio stack, I think that it's really when looking at the entire thing where it's impossible to find something that has each of those different elements, right? When we look at the breadth of phone numbers and the ability to quickly expand to new geographies with not just SMS, but voice, when you look at the strength of Programmable Voice and the increased features that are available at that level, and then the stuff on top that is like this Insights thing, right?
Every time there's a shiny new conversation to be had about, "Hey, why couldn't we be using this service or that service?" It's like, well, can we ask at the end of every call, how much was that call, get an answer, and then bill our customers in real time so that they know exactly how much that call costs. Those sort of things are so hard to build. Can you have a customer support rep on our service look at a call and tell that the audio plug isn't plugged in or tell that the network connection is crappy? Those are the sort of things that when you look at consuming all of those from one place and not having to assemble all the variety of component services, we just haven't found anything that measures up.
In terms of, I actually think one of the interesting elements of Insights in particular, and this gets to the quality and reliability thing, so much can happen between the mouth of a customer and the ear of an agent, right? There's so much technology and so much network in between those things that I think Insights has really given us transparency into what exactly do we mean by quality and reliability and helping our customers understand that as well. Helping our customers understand that, okay, well, here are some things that are really occurring at the carrier level, right? Here are some things that might be happening in Twilio infrastructure. Here are some things that might be happening in Zendesk infrastructure, and here's what's happening in your own network. Helping our customers understand that has made us realize exactly how strong Twilio's infrastructure is.
It's hard to diagnose these issues or tell exactly where these things happen in the stack. I think it's been one of the interesting side effects of us using Insights is realizing how strong Twilio is in that regard.
The other thing, you asked about sort of resiliency and Zendesk is unique. I feel like we've talked a lot about the long relationship that Zendesk and Twilio have had, and you were a customer of ours back in 2012, 2013, during some days where our reliability was not very good because we were going through hyper scaling, and our sort of V1 naive implementation of these things hit its limits. Those are some tough times for our relationship. I think I took Mikkel out to dinner more than once. What resulted is you've seen what it's like when it's not resilient, when you don't have that, and then you've seen where we are today, and I don't know if there's any commentary you would have about the investments that we've made over the years that you see valuing your customer base.
I think there's the things we've talked about, like global low latency and just getting the Twilio Super Network right. I think what I've been most impressed with, and actually what we've learned the most from, is your operational maturity model. It's less the things that you're doing and more how you're doing it, right? How deeply you've embedded that in the Twilio culture is something that we see every time we're interacting with one of your product managers on a new feature or new product. We get to see how deeply you've embedded quality and reliability into your culture, and that's something that frankly, we've said, "How do we do more of that," right? I think you've done more than anybody in the industry in that regard.
Great. Is there another question?
My name's Pat Walravens. I'm from JMP. My question for you is, I'm sure you both spend a lot of money on Twilio, my question is, how important is price in your selection process? As this relationship moves forward, how do you evaluate how much you're spending and whether that's the right amount?
Price is one factor, but honestly, as long as it's not don't tell my sourcing person I said this. As long as it's not out of the realm of-
I won't listen
I said it's out of the realm of reasonable, right? There's numbers where it gets to, like, "I can't do that." Within a range, the ease of implementation, the security architecture, the ability to meet all of our regulatory requirements ends up becoming an equally significant. It's not a price-only decision by any means. Again, there's some people, I think, in the market that have gotten a little bit ahead of themselves in terms of their value. Twilio's not one of them. It's one factor.
We're in an interesting position because there's an aspect of our business which is essentially reselling, repackaging and reselling Twilio minutes. There's a way in which price is a Twilio knows a lot more about the market price of a toll-free minute in Germany than I do, right? There's an aspect of which that I rely on Twilio for having, and the reason why that partnership is so important is our ability to penetrate the call center market in a particular geography is dependent on Twilio being able to secure and offer us compelling prices. Boy, that ends up as revenue for me, right? When I'm able to resell those minutes, and that's one part of our business model. Of course, most of what Zendesk does is agent seats, right? Per agent, per month agent seats.
That's where I think the partnership is important because we can go and look after Recently, right, with this, we took a crack at relaunching Talk in Japan, right? Said, "All right. We think there's a massive market there. Here's what Zendesk is doing more broadly in the Japanese market. How can we go and approach this? If we had this new type of phone number that our Japanese customers are asking for, if we were able to offer that at a compelling price, could we really penetrate this market?" That's increasingly how our partnership is working.
Okay, thank you.
Okay. I think we have time for one more. Who's it going to be?
Hi, Ryan. I was just curious if you could comment on any backups that you have to Twilio, if something is not working, and how hard it would be to switch. I think that's something a lot of us have struggled with over time on understanding that.
On backups, this is a conversation that we have with Jeff and team quite a bit is, Twilio is a backup for Twilio, right? Where there's so much resiliency built into the Twilio network that there's a lot of failover that's built in there. We do offer the ability for our customers to say, "Hey, if you want to route calls somewhere else, it can kind of bounce off at the edge of the Twilio network and into another system," which, if there's a, call it a brownout or an issue somewhere in the Twilio network, that provides our customers with some degree of comfort. Some of our customers take advantage of that, some of them don't. We don't have a second system that we're kind of failing over to. So far, haven't needed it, frankly.
Part of the reason for that is that we don't need it, but part of it is that that would be really, really hard to go and do. Again, this, I think, speaks to how much we've benefited from and how much of the Twilio stack we consume, where there's no such thing as a generic programmable voice to embed into an application like ours, right? I think we would be in a different situation if all we were doing was buying phone numbers in minutes, but we're not. There's a significant portion of our offering which is pretty tightly intertwined.
While, hey, look, I love the fact that every time we talk to Jeff about, "Hey, what should we be doing for our failover strategy?" You talk about how you're investing in making Twilio more resilient than ever, so that we would never even need to think about doing that. I like those sort of conversations a lot, because it would be a pretty gargantuan effort for us to do it otherwise.
It was some of the early conversations that we had with Zendesk around why we've been very public about our OMM, our operational maturity model. A lot of companies may would've said, "That's our secret sauce. We're just going to keep it inside." We actually wanted to talk about it because we want to build that trust, both with you and with your customers.
Also share our learnings with you, because if together Zendesk plus Twilio isn't a reliable solution for the market, then neither of us are going to win, right? That's why we believe in sharing this with our customers so that we can all build these great solutions together. Naureen?
Yep. No, you hit it.
Yeah. As far as his question, though, I don't even know the answer to this. Do you have some backup for us?
You know, I actually don't know the answer to that. Or not off the top of my head. It's not one of the core if a text messaging goes down, it's not the end of our world, so I would actually have to ask somebody what our business continuity plan is. I don't know.
Great. Well, I think that's all the time we have for questions. Please give a nice thank you to Ryan, Naureen. Thank you for spending the time with us.
Thank you.
Thank you, Jeff.
Now, for the gentleman you've all been waiting for, allow me to introduce Twilio CFO, Lee Kirkpatrick. Lee. Applause.
Thanks, Jeff. Again, my name's Lee Kirkpatrick, Twilio's CFO. My colleagues have introduced themselves and said their tenure at Twilio as a reference point. I introduce myself in terms of how revenue has grown since I've been at Twilio. I joined Twilio and we were a $3 million quarter, and we're now over $100 million. A lot of excitement, a lot of growth. Oh, thank you. From my perspective, it's very exciting to be part of a team working with such a great group of people, powerful business model. I think we said it several times here today, when I walk into one of George's sales all hands or see the product team working on Studio, it is very apparent that we are just still at the very beginning, scratching the surface of a huge opportunity. This is extremely exciting.
In terms of key takeaways, what I want to cover is this is a company that is growing rapidly at scale. We talked about in our last earnings call, our base revenue ex Uber is growing extremely fast, well over 50% in Q3 and in recent periods. We're growing quickly. We have an extremely powerful business model, we've talked a lot about that today. In fact, a lot of my thunder has been stolen in terms of some of the previous presentations about some of these great unit economics, our great business model, the developer go-to-led market. Another key part of that business model, I think Ott covered significantly, was we have made $220 million of investment in terms of R&D.
That software investment drives differentiation, which leads to great numbers, great unit economics, allows me to put charts up that go up and to the right. Last, we have a very attractive long-term operating model with target operating margins of 20%+. The platform model is a bit different than what you're used to seeing in typical SaaS, we'll get into that a bit later. Revenue drivers. As a level set, approximately 80% of our revenue is usage base. When you take this usage-based model where you bill by the minute, message or authentication, you charge pennies and sometimes less than pennies for each transaction, you have over 1,000 transactions per second across a customer base of 46,000 customers, you get a chart that looks like this. Steady, consistent revenue that's growing.
This is a snapshot of the software sector of companies over $100 million. Even with Uber, we grew over 41% in the third quarter, and that puts us in a really fine company. Across our revenue base, that's becoming increasingly diverse. 17% of our revenue was with our top 10 customers. That's improved from 31% over a year ago. Here's a little deeper look at this. This is our top 100 customer accounts, and you can see outside of our first two customers, we have very little concentration risk, and it's highly balanced. In fact, if I took this chart from a couple of years ago, same thing. If you took those top two customers, the percentages would be slightly the same. On an absolute basis, you would just see a rising tide across that whole customer group.
The business model, that's kind of been a theme today in terms of our great unit economics and business model. First item where you start, customer growth. We've had rapid customer growth. When I was speaking at the [IPO] roadshow, we had 29,000 active customers, and we now have 46,000 active customers on the platform. If we go a little deeper into that, the 46,000 customers, in terms of the last quarter, grew 35% year-over-year, strong growth across the overall customer base and the long tail. If we work our way up, in terms of what we look at as our mid-market size, customers with 10,000 ARR, $10,000 a year customers, over 3,700 customers growing 43% year-over-year. At the very top, we now have 47 active customer accounts spending over $1 million per year.
That number is up 74% year-over-year. If we reflect on this is a testament to what we've been talking about today, this $220 million investment in terms of R&D, reliability, quality, insights, service features. This is why customers are trusting Twilio to move their workloads over to Twilio and bringing on new workloads on Twilio. We're getting customers on the platform rapidly, and when we get that revenue on the platform, it stays on the platform. Our revenue retention rate is in the mid-90% range as of Q3, and has been consistent in that range over the last two years. We retain the revenue, and then we grow the revenue. This is a snapshot of our dollar-based net expansion rate with Uber, which has caused some ups and downs.
A cleaner look is removing Uber, and this is a slide Jeff presented earlier. I think the key takeaway here, it's been consistent over the period, and if you go back to early 2015, 140% relative to 137%. Revenue was a third the size, but we're basically maintaining the same level of expansion rate. Another look of our business is ARPU. This is our base ARPU ex-Uber. This number has grown $2,400 since the beginning of 2015. We're rapidly adding customers, and they're spending more on the platform. Combining the previous charts in terms of customer growth and expansion rate, you get cohorts that get larger and grow consistently over a certain time period. This is annualized revenue growth. We talked a bit also about our expansion rate of revenue acquisition earlier. Revenue efficiency is a key part of our model.
This is a comparison to other high growth companies. Hope Ryan from Zendesk stepped out of the room. We have a very efficient model here, and at Twilio, we drive high revenue growth, but we do not require high levels of investment in terms of sales and marketing to deliver that growth. I'm going to go a little bit deeper in terms of revenue and gross margins. Just again, we're all on the same page here. In terms of our cost of goods sold, we have three categories, but the main category for Twilio today is carrier expense, and that's our network expenses. That's 90% of our costs. If we go deep in terms of our revenue, we reported in our last earnings 53% non-GAAP gross margin.
What I wanted to do is just give you guys a little closer look in terms of what's under the covers of those numbers. If we take those numbers and pull out our variable customer accounts, Uber and our BeepSend acquisition, and this is a bucket of revenue that's been fluctuating significantly over the last few quarters. If we pull those numbers out, we're left with 46,000 base customers on the platform that drive the vast majority of our revenue. That gross margin has been 60%, and that number has been very consistent at 60% going back to prior to the IPO. We do see fluctuation movement in the gross margin. When we look at this core group, we've had high consistency at the high rate of 60%.
One thing we've talked about, I'm going to get in a bit later on, in terms of our long-term operating model and gross margin, is our target gross margin is 60%-65%. When we set that margin and looked at that and modeled this, and we talked about this prior to the IPO, one of the factors that made us very confident that we'd achieve this number is when we were looking at this core base of customers, we already were in that 60% range. I'm going to talk a bit deeper in terms of gross margin, and there'll be fluctuations and all that, but we do have a base of business that has remained very stable at the 60% target. Going a little deeper in terms of how we're looking at gross margin.
If you saw the roadshow, listened to our earnings call, had a one-on-one with me, you've heard me say many times, we're focused on reach and scale. We're not maximizing gross margin in the near term, and in fact, gross margins may fluctuate in the near term. What does that mean? Reach and scale means we're focused on acquiring customers, bringing them on the platform, and making them successful. Another way for reach and scale is revenue growth, and really, we've talked about this as our strategy, the measure of this strategy is, are we executing and driving revenue? Revenue growth is a measure of this strategy. If we go a little bit deeper and think about what is impacting the gross margin, gross margin's impacted by our products, what our customers are buying, what they're buying, and where that usage is taking place.
in terms of our customer base and our customers, and we saw Ryan on stage here, are primarily communicating to their customers through traditional voice channels and traditional SMS channels. That's how they're communicating with their customers. Accordingly, most of our revenue today is from products that support those, through core communications, traditional voice, traditional messaging, and there's a corresponding network expense associated with those. That's our current product line. As we look out into the long term, more of our revenue, we expect to be higher level in terms of what we call application services, which Pat talked about. These are services that do not have a network expense. We see more revenue coming from application services, higher level business logic, and this revenue does not have a corresponding network expense and has a much higher gross margin.
As we have this type of revenue, this higher level software revenue, again, it can sit on top of traditional channels like PSTN voice and messaging, and newer channels like IP messaging or video. Think of Twilio as our offering is software that sits on top of a channel. If you look at it from an international basis, those channel costs are higher, so we have a lower gross margin profile. If it's an application service, there's no network expense, the channel expense is less, so that's a high profile. International, lower profile, application services, higher profile. We'll jump deeper in terms of international. This is a number that we disclose every quarter, 24% of our customers are headquartered outside of the U.S., but this doesn't really tell the whole story from a margin standpoint.
We have a lot of U.S. customers on the platform that are U.S.-based, but they have a lot of usage outside of the U.S., and if we add this up, we see 53% of our revenue is from outside of the U.S. I'll take a step back here. This is an extremely important value proposition for Twilio, where a customer can be in one country and have a global opportunity with minimal development effort. If we look at our international revenue mix, it's a bit over 50%. International messaging is growing faster, and international revenue in total is growing faster than our overall business. This number, also at 53%, was uplifted a bit by our BeepSend acquisition in terms of Q1. What does this mean in terms of, as we're talking about gross margin?
For our core businesses, if international revenue mix increases, that does decrease our overall gross margin. As we look at the business, increasing international business is a very good thing. We are building the communication solutions for how our customers want to communicate today, and that's primarily voice and SMS, so that's our goal, to increase customer scale, customer reach, and get our customers successful using Twilio. Then in the long run, we'll be rolling out these application services and other products to layer on top of that large platform. That's the strategy. Jumping over to that, in terms of application services and the product. We've talked about in the past that these products are very software-based products, are a small part of Twilio's revenue, but growing rapidly.
Here we take a snapshot in terms of Q3 2017, and we see $9 million of application service products growing 100% year over year. If we look at how that fits in the entire piece of the pie, we see that application services, 9% of total revenue. From a platform model, this is not something that will change significantly overnight. As we roll out products, our customers adopt the products, we expect to see a steady increase of that green piece of the pie as we move forward, and that's how the platform model works. Another key thing about our overall model is application services pull through core communications revenue. Pat talked about our TaskRouter product, which is a really great tool for customers wanting to build a call center.
The TaskRouter generates application service revenue, but also pulls through core communication revenue, whether it's voice messaging or SMS. The majority of our revenue in the near term is coming from these core communication services, and then that's in the blue. As Ott talked about, nine years, $200 million plus of R&D investments have differentiated that product through quality, reliability, intelligent services. We talk a lot about the newer services. I really want to point out that that blue bucket, this is a great product. We have an edge, we have differentiation, and this is a big revenue driver for us. What's a great thing about this overall model, though, we're well-positioned to add to the application services on top of that, what Pat talked about.
We're increasing investment in terms of the Engagement Cloud, application services, and we see this as a great opportunity for Twilio going forward. One of the proof points George talked about in terms of his presentation was this ELA agreement that we signed, where it was a near $10 million commitment. Half of that commitment was the application services software revenue. The other half of the commitment was usage-based revenue. As we bring more and more of those type of agreements onto the platform, we'll see our application services increase. In terms of long-term operating model, and we'll come back to gross margin in a bit there. I will take a step back. Our operating model is different than your typical SaaS company. We have a lower gross margin profile today because of the network expense component. Our gross margin profile is lower.
We also have a far lower sales and marketing expense as a percentage of revenue. What we've talked about in terms of far greater sales efficiency, the developer-led go to market. That's far better at that end. Those offset. In terms of gross margin, right? We have a target gross margin of 60%-65%. Summarizing our gross margin discussion, like where we are today, how we get there going forward, if we look at today, pulling out these fluctuating items I talked about earlier, we've had a relatively stable gross margin in recent periods. There's multiple puts and takes on the gross margin, but the two biggest drivers are international mix and application service mix. Both of those are growing today, and both of those impact margins in different direction.
In the long run, the biggest driver is application services as we've modeled our business out and make those investments. As we look at the business, for us to achieve our 60%-65% target margin, which we're very confident with, we are modeling application services revenue at 30% of total revenue. Further down the P&L R&D, we will continue to make investments in R&D to keep our product differentiated. The target is 15%-18% in terms of the target model. As we grow revenue, we expect to see operating leverage on that line item. We've talked a lot today about sales and marketing. As discussed, this model is a far more efficient model in terms of generating revenue than what you typically would see with other high-growth companies.
Again, as we grow revenue, we expect to see the leverage down to the 16%-19% of total, and same thing with G&A, as we scale and grow the business, moving that to the 7%-9% range. A bit of housekeeping. 606 update. After hundreds of hours of review with KPMG and thousands of dollars of expenditure and a long 20-page memo, the impact is nothing. We'll go there, and then we'll wrap it up in terms of a key takeaway. Again, this business is growing rapidly at scale. We have a terrific business model with great unit economics, and we have an attractive long-term model, albeit it's different in terms of platform, but it's a great model. We're targeting 20% operating margins. Thank you.
Thank you, Lee. I love Analyst Day, where the 606 joke is the one that gets the most response from everybody. All right, cool. We're going to wrap up, then we're going to do a Q&A panel with the executive team, where you can ask us questions that you have. I think the key thing, when I put myself into your shoes, what do you want? You want to invest in a company that's operating in a large market, that has a leadership position and a differentiated product. We're here today to show you how we fit into that picture, right? I believe we check all three of those boxes. Right?
When you think about our Super Network, what we've built there is a software overlay over the physical assets of the world, of the world's communications infrastructure, that are powering a fundamentally new way for the world to communicate at global scale. Right? We don't own the copper wires. We don't own the fiber optics. You know what? Uber, Lyft, Airbnb, they don't own those things either. Right? Lyft doesn't own the cars, Airbnb doesn't own the apartments, but they're using software to put those assets, deploy those assets in the world in a fundamentally more efficient way, connecting supply and demand together, using software to optimize that, to measure what's a good apartment, what's a good driver, connect those riders and drivers and renters and hosts, right? That's sort of a new network business model, and that's what we've done for the world's global telecommunications with our Super Network.
Right? We focus a lot on what our customers say they want. Breadth of offering. You talk to Mikkel, you talk to Naureen, they've got their customers, they've got their wealth managers, they're all over the world, they want the coverage to be able to talk to their people wherever they are. They want depth. When they say they want a phone number, they want it now, not in 3 weeks. They don't want to have to pick up the phone and call someone. They want an API that returns a phone number in real-time, and they want that phone number to work, right? These are things that our software team building the Super Network has focused relentlessly on now for many, many years, and it's a flywheel that's getting better all the time. Right? Think about the Programmable Communications Cloud. Right?
The key to the Programmable Communications Cloud that Ott talked about. Is that number one, strategically it is driving our business model. The fundamental efficiency we get with our developer-first approach is driven by the quality of the building blocks, the ability for those building blocks to solve our customers' problems. It does the thing our customers need it to do, and the great experience the developer has when they're getting up and running. The fact that we invest so much in documentation, debuggers, our runtime environment. That's what allows developers to come on board, build that prototype in 4 hours, and then land Morgan Stanley as a customer. That's really powerful. Developers need that kind of sophisticated infrastructure and tooling of our runtime and our debuggers and all that kind of stuff in order to have those success stories.
That drives our business model, the fundamentally efficient developer-first approach to how we're tackling this big market. It's also key to unlocking use cases. I talked in the beginning about the trust flywheel. We've been in this market for almost 10 years now. What we put up on day one as the voice API would be an embarrassment today, right? Because what we put up back then, we then learned from the world's developers about these use cases, all that stuff I talked about with Ryan and Zendesk. We stubbed our toe, I think, on every sharp corner of telecom along the way here, figuring out how to make this thing work at scale globally with low latency, high quality, five nines availability, all of this stuff is really hard to do.
You get there by putting out a product, listening to customers, having them guide your roadmap, and that's what we've been doing for almost 10 years now. The mere existence of a voice API is not new, but it's the sophistication of that API and the iteration after iteration after iteration that we've undergone. In fact, Agent Conference, the newest flavor of our conference product, we just went to GA, I think yesterday. Right? That's the learning that we've done by working with developers and customers to understand the use cases of, you would think the idea of just bridging two calls together is a simple idea. Turns out it's incredibly nuanced if you want to unlock these workloads of global scale call centers, for example. That's what we've learned along the way.
We've built now, I think the third or fourth iteration of that product to be able to unlock that. That's our new Agent Conference product. Not all APIs are created equal. We've invested so much of our R&D energy. Lee Kirkpatrick had the stat, $200 million we've invested in our R&D in order to take that feedback from customers, learn from what they're trying to build, and continually plow it back into better and better and more sophisticated APIs. Customers have shown us this roadmap. Only by seeing thousands of customers across all these different industries and across all these different use cases, do you understand how these APIs are differentiated from each other because we've invested so much in learning from our customers and continually improving these APIs. Then it's the resiliency.
Just keep shipping and listening to customers is important, then the resiliency. Ryan told you in the early days of Twilio, we had Twilio Client. I'll admit it. We had availability problems with Twilio Client. We learned the really hard way what happens when you put something out there, customers embed your API into their product, then they start scaling wildly all around the world. We learned the hard way, we built the resiliency in. We developed this OMM, this operational maturity model, to make sure we never had those kinds of problems again. That investment now pays off when you get customers, enterprises like Naureen deploying this at global scale for a bank like Morgan Stanley.
That's what we've learned along the way, building the PCC and investing in that flywheel of trust that leads you to customers, that shows you the roadmap that leads you to more trust and that completes itself. Pat talked about the Engagement Cloud and what we're building as part of our first application platform. At the beginning of the day, I said we invented one platform, the Programmable Communications Cloud, actually, we need to invent more. This application platform is a brand-new invention that we're bringing to market. It's an innovation. Today, there isn't anything out there that really combines the power and flexibility of building a solution with the ease and speed of buying a solution. That's our goal to unlock, Pat showed you Twilio Studio.
That gives customers just that, the ability to have that complete flexibility, the ability to get up and running together. Therefore, together with our ecosystem of customers building on top of Twilio, we want to go take down these big markets, huge markets that use communications to drive business outcomes, marketing, automation, sales, support and service, logistics, operations, all the buyers, all the consumers of communications inside of companies throughout this complete life cycle that companies use to communicate. We think those are the markets. Those are the big opportunities that we're going after with the Engagement Cloud across that full life cycle. We're really excited to undertake that mission as the next step of Twilio's master plan. I know that platform business models are unique. It's not as straightforward as a SaaS company. It's not like downloading software.
I completely empathize with all of you trying to figure out Twilio, trying to figure out this brand-new business model. I know it's not easy. I know it's new, and it's fundamentally different than a whole lot of software companies out there. We think it is an incredibly powerful model to address one of the largest software opportunities on the planet in this huge industry that's not just communications, because also all those other ones I mentioned, marketing, sales, operations, support, service. It's pieces of software, it's pieces of hardware, it's pieces of the telecom networks. It's all that munched together into something fundamentally new, reconfigured for an era of software. That is what we're going after. I get it. It's not linear, doesn't map well to other stuff. It's hard to figure out. I get it.
That's why we did today, this event today, to try to explain how we're thinking about the opportunity and give you the opportunity to ask us questions about that. Here's what we're going to do. This is our vision that we put forward for all of you today. What we are going to do is we're going to focus on our customers. We're going to focus on what they need and focus on what they're telling us is important to their business. Scalability, global, trust, capabilities to unlock these markets.
We are going to focus quarter after quarter, unlocking more and more of this global opportunity and focusing on what our customers are telling us is the right way to do that, is to focus on those elements of trust, those elements of capabilities, agility with resiliency, and do that every single quarter after quarter, slowly eating away at what we see as an enormous opportunity. You heard from George, our go-to-market motion is working incredibly well. We've got a sales team that is engaging with customers, using data, driving the expansion rate, driving engagement with customers, and building that understanding with a high degree of confidence that making the investments to help us grow this market, to build those conversations with our customers, help them win in this new market, is the way to do that.
Quarter after quarter, expect for us to slowly eat away at this big opportunity. That's what we're all here to do for our customers because that's what we see as how we're going to win and how we're going to truly unlock this big opportunity. Thank you very much. I'll invite up the team, and we'll do a Q&A panel. Is it chairs or are we standing?
BYOC.
BYOC.
Very good.
We'll stand.
Oh, chair brigade. Thank you.
Bring your own chair.
While we're getting the chairs set up, I think we got mic runners out there. Greg, do you want to pick the lucky winners?
Let's just sit down.
I think we're good.
Let's just sit down. That's fine.
Thank you. Space ourselves out a little bit. Analyst crunched down there. Who's our first winner? Ittai.
Yeah. Hi. Ittai Kidron from Oppenheimer. First of all, thanks for the Investor Day. It was very helpful and informative and great detail, Lee. We will look forward for you to repeat that every quarter. I have a couple questions. First for you, George. You've talked about the sales productivity, and I'm impressed with that. I guess I'm trying to Help me understand how you're trying to fine-tune the sales effort to sell more application services, which is clearly something that's going to be helpful, not just on the top line, but from a margin as well. Are you building some specifics into incentive plans in order to drive that type of behavior? Is it too early to do something like that, just given the wealth of opportunities that are out there? The second question is for you, Lee, regarding the gross margin.
Again, thanks for the detail around that. I guess when I look at 2016, your gross margins were about 57% for the year. When I look at BeepSend, which closed in February, if I remember correctly, and you marginally moved down a point from March to June, and Uber is only half to a third of what it was at the same point last year. It feels like the impact of international then must have been massive. Three, four points associated just with the international mix on a year-over-year basis. You gave the mix of traffic international, which was just over 50%, but can you tell us what it was in the same point last year? Did it really double as a percent of mix international on a year-over-year basis? Is that the right way to think about it?
I'll go first as I think it's a more pertinent question. In terms of international mix, that has grown. We're not disclosing the specific increase year-over-year. We have seen international mix grow, and you're right with that BeepSend there. As I pulled out those numbers, if we look back, pulling out those heavily fluctuating items, which is our variable revenue, which is a significant amount of revenue, and a lot of that is international, Uber and the BeepSend, the number was very stable going back through last year. Now, there are puts and takes in those. The international mix does have a dampening impact on gross margin, and we also saw that application services increase, so that has a positive impact. Pulling out those fluctuating items, it's been very stable over that period.
Is it fair to say that on a year-over-year basis, it's really the geographical mix that has been the material change in your business from a contribution to the margin headwind?
If you look on a year-over-year basis, it's those fluctuating items we talked about that would have the impact. George?
Good question. I would say right now our primary focus, as you saw from the slides, is coverage. I think the way to sell more application services, to sell more Twilio, to sell more of everything, is in coverage and in the investments that we've shown you. I think that's the primary way we're going to sell more. In terms of incentives and things like that, I would say that I believe in generally keeping things simple versus trying to overcomplicate or overengineer things. We are seeing the uplift in application services that Lee showed you. We do think that things are working and trending in the right direction. We'll constantly tweak things as we go along, I would say the primary emphasis is getting more people talking to customers.
I think that's the number one way we're going to sell more application services, we'll continue to refine it over time, I don't think that's the primary lever.
Jonathan Curtis. Just quickly. First off, can you commit to giving some of these new metrics that you exposed to us every quarter? Please shake your head yes. If not, why not?
Yeah. Not committing every quarter. We do have our key metrics that we give out on a quarterly basis. These are the type of metrics, for instance, our progress in application services revenue, we'll give it out periodically in terms of milestones. We have a set of metrics, for instance, that we talk about at Twilio Signal, these are not necessarily regular metrics.
Okay.
Sounds like you want me to keep coming to Analyst Day.
Application services, how much did you have to touch to drive that growth in revenue? How much sales engagement was involved there versus natural pull-through from the model that we understand today?
Maybe you should talk about the ELA or.
I would say it's a mix. I would say that some of it, a lot of the initial number, the pre-doubled number is happening because I think there's synergy between application services and our communications. They kind of pull each other through, so I think that there's a little bit of that. Definitely, I would say that the sales effort has accelerated that, and that there's no way we would have closed that enterprise agreement and that application services component without the sales effort. I would expect that that would continue to be contributing to it more and more over time.
Just one more quick question. What is the impact of RCS on the messaging market and your opportunity?
With respect to RCS, it's something we've been following for quite some time. There's been a number of announcements out there. Single carriers have announced that they're launching RCS. For us, we see RCS as a natural fit in our channels offering. We have support already for Facebook Messenger, a number of other channels there, we would expect to plug RCS right into that part of our offering.
Okay. Thank you.
I would say from a customer perspective, I think Naureen's answer was quite instructive, actually, that was obviously not about RCS, it was about in-app. Most of the customers today, SMS for a lot of customers is still a new thing, honestly. Every customer wants a strategic platform where they can use all these channels, but, the truth of the matter is that for many of them, they're not driving material traffic on these new channels today. I think there's quite a good runway for even the products we have, frankly, or the core channels we have.
One question we just have around the application services, thank you, first off, for kind of breaking out and showing the revenue. I think beyond that, there's a question I think on a lot of minds around how much does the application services or these highly differentiated software components that the competitors we've all heard about don't have even, how much of the business are those products touching? I don't know if there's a specific answer for that, but trying to get a flavor for if you look at that 90% of the revenue that is SMS and voice minutes, how much of that volume is a customer who is either using TaskRouter and that's pulling minutes or using Notify and pulling SMS versus someone who's only engaging at the Super Network or programmable layer?
I'll speak to that. We haven't disclosed a specific number in terms of how many are using what. We do look at those numbers in terms of customers, the number of services for customers, which customers are using, for instance, voice intelligence services that Ott talked about. It's meaningful in a growing amount of customers, but that's not something we're talking about publicly.
Got it. Okay. Second one from me, then pass it off. If I take a step back on the point around channels and some of these digital channels versus your core SMS voice channel, if I look at that was a world that had hundreds of carriers offering effectively the same product that were very hard to engage with, very natural place to put a layer of software over it. If I compare that to some of these digital channels, I realize there's a number of chat apps and there's numbers of voice assistant, but they're measured in handfuls, right? Each one of those products actually has a lot of features that the others don't. If you want to leverage the full capabilities of iMessage or WhatsApp, there's value to going direct there that you may not have had going to a carrier.
As you think about if there is a shift towards these other IP channels away from phone number calls and SMS, what do you think is sort of the value prop of Twilio in that world, sort of aggregating those channels versus a developer who says, "I want to leverage the best of Facebook Messenger or the best of iMessage?
I'll give you my take on it. It is that proliferation of channels that is creating the complexity for our customers. Every business out there is trying to keep up with that. Now, you're right, it's not thousands of these apps that matter. It's going to be a handful. But even a handful of these things, I mean, these APIs are constantly changing, and for the most part, these companies aren't really focused on the tooling, the developers. You get a little bit of a, "Hey, we're a big company, people will adopt our APIs regardless." We see an opportunity at two layers of the stack. Number one is to provide a great experience for developers. To do things that developers need in order to be successful as they're adopting these APIs. Those are things like our debuggers and insights and all that kind of stuff.
The bigger thing that we think about is, and this is why we're investing in the application services and things like the Engagement Cloud, instead of capturing the desire to transit a communication, capture the intent of the customer. What is it they're trying to accomplish? Then we can translate that intent into the capabilities of the underlying channel. If you capture the intent of what the customer is trying to do, i.e., at the application platform layer, we can then bring any of those channels they want to bear because we know what they're trying to accomplish and effectively employ that channel for them in the most intelligent way and say, "Okay, well, this service supports cards, and this one supports images, and this one supports payments," and mix and match to fulfill on their intent properly.
That's where we see, in the application services arena and application platform, the ability to integrate with our customers at that more business intent layer than in the communications transfer layer. When I was at Salesforce, we acquired a company called Radian6, and they were focused on Twitter as the primary channel. I can tell you what a nightmare it was to try and even with an application that was really just single-threaded to even one channel that it depended on, what a nightmare it was to keep that up to date with all the changes in the API or Buddy Media with Facebook. It's quite laborious, even if that's your entire business, it's very difficult to do. I think what's working in our favor, and I see this when I talk to customers, is that they are really seeing just increasing fragmentation, right?
That they can't just focus on a solution partner is one thing, but now I'm talking to customers. I talked to the CEO and the CMO of a major, one of the top five online florists, flower delivery. They're trying to wrestle with conceptually, oh, there could be five, 10, 15 channels depending on what's happening in the world. I think fragmentation actually works in our favor, because in that world, yes, if I was single-threaded to one channel, even that's difficult, in a fragmented world, it doesn't make sense for me to do that level of investment to manage all these channels.
I think that more and more they say, "Oh, Twilio, if you can solve that problem for us, then that is very high value for us because we don't want to think about that." I think how easy it is to spin up one of these communication channels actually works in our favor, and it's kind of actually inverse, oddly enough, of the carrier world with some of the dynamics, there's still value creation for us there.
Yep.
Yeah, just coming back to gross margins, a couple of follow-ups. Lee, you alluded to some of the pressure from the variable customers and Uber. Is that now largely behind you, or do you expect there'll be some further downward pressure on that front? Then, I guess, longer term, maybe for Christine or Lee, as you think about the underlying network cost component of it, what are the opportunities on that front? I know you talked about business mix, but are there ways to cut costs to improve margins?
Sure. In terms of the first part of your question, we have talked about Uber in terms of calls, and we have made adjustments to their pricing to get their pricing commensurate to their volume. Uber is playing as expected. Generally, from a pricing standpoint, that's behind us. That's in terms of Uber. The variable customers, they're in that category for a reason. We've talked in the past how there's been a shift in terms of international mix and margin impact and mix, and that's where we can see pricing mix in terms of what countries are flowing to, in terms of usage and revenue and all of that. That one we don't really talk about going forward. That's a category both the revenue and the usage within there is volatile.
Does that answer your question? Yeah. I think the second part of your question was, as obviously, Lee puts up a chart that shows us that 90% of the cost of goods sold are sitting in Super Network. We actually focus on those costs every day. I think we were talking earlier about where we focus and where we put our time. We actually have access to all the rates. All of those rates are indicative of different kinds of workloads and different kind of payloads. Depending on what the workloads are, we talked a bit about there's a big broad space of messaging and voice being provided today. There's wholesale, but there's also what Twilio is providing.
While we have access to those rates, and we focus very hard on leveraging the volume that our rates get us and the advantageousness, if that's a word, that that volume begets. We continue to drive costs down worldwide, but it's an everyday focus for us. Yeah, it's always a focus. It's something we work on all the time.
Thanks for taking my questions, Pavan. Two questions. First, you put up the chart where you had sort of the very large customers up front and then sort of the long tail that came out. If you look at application services, I guess what I'm wondering is, how correlated are those with the long tail or with the larger customers? Or are the larger customers largely Programmable Voice and Programmable SMS because that came first? Are you seeing some of the newer customers adopt because of the application services? Just trying to understand if there is a correlation there between all these sort of very unique use cases and the end of that long tail, or if it's across the board.
Yeah. Without getting into specific numbers, if we look at our customer accounts, the customers using application services have a substantially higher ARPU than the ARPU. I can't speak to that chart, but in general, when they've adopted application services, there's that corresponding software revenue, and they're pulling through more of the Programmable Communications Cloud revenue. Those customers are larger.
Great. Just maybe one broader one. As you move to application services, it feels like there's almost a move, and I know you're still selling to developers, but a move towards more productized solutions, things I might sell at a higher executive level than a developer. Does that change the sales model? Does that change the go-to-market model? Because then I'm sort of selling to the head of a call center because I'm going to offer him an almost configurable call center application. I guess as you think about the future, not today, but down the road, as these application services become more productized, does that change the go-to-market model? Thanks.
I think that it changes some aspects of it and not others. I think that you will see, in terms of the marketing side, you would see us message more to the business decision-makers in that world. I think that you want to have a brand position in the mind of the customer, and I know a lot of times we talk about sales, but I think there's a big marketing aspect to that.
In terms of the sales motion, I still think that if the kind of innovation of Salesforce was to kind of bring line of business into the early part of the discussion, I still think that Twilio's innovation to bring the developer in still plays out in that world, because of just the nature of what we do, the value prop, we're going to be naturally aligned more with companies that see value in development, see value in customization. In terms of the nuts and bolts of the type of rep we hire, the sales cycle, I think you'll see some differences, mostly in the execution details. For example, today, I would say many of our sales cycles, we don't really actually need to even demo the product.
I think in a world where you saw, sorry, Pat's demo of Studio, if you're selling something that's like an IVR, you're going to have a very different type of motion. I don't think it fundamentally changes the nature of the process, developer up, value proposition. It may end up pointing to a different business decision-maker, but I think a lot of the mechanics are more similar than different.
Hi, thanks. I think in our business, one of the biggest things to avoid is owning any Amazon roadkill or future Amazon roadkill. You saw with the announcement of Pinpoint that people don't know really how to react or how to exactly interpret what Amazon's ambitions are in this space. I was wondering if you could expand upon the nature of that relationship and maybe help make the case to us why that won't evolve into more direct competition in the future.
Yeah, absolutely. I'll speak to the longevity of the partnership there and the strength of the partnership, right? We've been working with Amazon since the beginning of Twilio. We have a very strong relationship with the folks there, given that I used to work there, and they were a private investor in Twilio as well. While, of course, that's no guarantee, that's given us a lot of insights into their roadmap. It's allowed us to be at the right points in the conversations that as they were thinking about products, to be able to build a partnership wherein we're helping to power their communications offering. Now, if you look at what they've been doing is primarily investing at the sort of SaaS layer with Connect, with Chime, with Pinpoint, and that's very much like our solution partner offerings, right?
It's like for the contact center, it's like Zendesk or Salesforce, who've also built their contact center offerings on top of Twilio, and that to us represents a very great buy option in the market. Twilio is going to focus on a certain part of the market that wants a tremendous amount of flexibility, but our partners are going to focus on a part of the market that are more focused on buying a solution. Ultimately, we think that the full market opportunity here is actually migrating to the cloud and to software, all of these legacy systems that for the most part are still trapped in old phone closets.
In the contact center case, with Amazon Connect and with Twilio and with Zendesk and Salesforce and all of these software offerings that are part of our ecosystem, we are tackling the 90+% of contact centers that are still sitting in a dust-gathering box in a closet somewhere. That's the opportunity that I think we're all tackling together, and it's a very big market.
Thank you for putting up the revenue retention slide. Lee, could you just remind us what the definition is you're using on that slide? Then I think it was up there quickly, but I think it looked like the number had gone from 94 to 97, and I know it's not a big move, but what is it that's causing the business to get more sticky over time as opposed to less?
Sure, yeah. We've had strong, consistent revenue retention, so I wouldn't read too much into it in terms of a one or 2% uptick. We're very happy with that retention. Our definition is we take a look at our active customers in a period, and then we look at the active customers in a quarter, we look at those active customers in the following quarter and how much of that revenues were retained from those active customers. Another way to look at it is of the customers that left the platform, how much revenue did they walk with? The people that left the platform walked with 3% of the revenue in our last quarter.
When I looked at Bandwidth, their ARPU is almost 20 times as yours. Does that suggest massive upside potential for you to increase ARPU, or is there anything more structural that will make your ARPU always much lower than theirs? Second part of my question is in your target model, what kind of top-line growth can we assume? Because, for example, Salesforce will say when they have slow growth, they will get this margin. Wondering what's the top-line growth in your target model when you say you reach 20% operating margin.
Yeah. To the first part of your question, I can't really speak to Bandwidth. That's a very different model than Twilio's. In terms of ARPU, we've had significant ARPU growth, as we talked about, over since early 2015, growing up $2,400. As we release products and bring customers on board and offer them new offerings, we do see the ARPU grow, and it's growing consistency, and that's part of the power in terms of the model. We have not given out a specific target revenue growth rate yet for the target model.
Mark Murphy with JPMorgan. George, I had a couple for you. You mentioned seeing a strong pattern in the enterprise recently. I'm curious what you're aiming to see there in terms of engaging some of the larger SIs, the Accentures, the Deloittes, the Capgeminis. Those are the more influential SIs that are driving a lot of these digital transformations and the customer engagement modernization efforts. It would seem like they could drive a lot of pull-through for you. What has it taken when to kind of get them on board a little more forcefully? As well, I'm wondering what is your longer-term target for business mix, if you were to look between hoodies and suits, if you will. I think for Amazon, that's something.
Do you have that slide somewhere, the hoodies and suits slide?
Yeah.
Pull that one up.
We could have a nice visual for that. I think it's something like 60/40 for Amazon, right? That's a business that's at
What is it, an $18 billion run rate? What would be your longer-term target for that mix for Twilio?
Can you repeat the first part of your question, and then I'll get back to the hoodies and GSIs?
It was on how to get the Accentures, Deloittes, and Capgeminis on board.
All right. Okay, SIs. We're just beginning with SIs. This quarter, we launched our first-ever partner certification program, so we're certifying partners on the Twilio platform for the first time. Today, the truth of the matter is we're mostly doing that with smaller, more boutique SIs, the more nimble ones. That's where we're getting started. That's where we started with Salesforce, too. It's not a big surprise. When I started at Salesforce, I think that the Accenture, Deloitte relationships, or when I started owning the SI partnerships, excuse me, at Salesforce, I think those partnership sizes were roughly closer to the low double-digit teen millions. Now, I don't know the numbers, but I wouldn't be shocked if they were basically billion-dollar-plus practices now. It takes a long time to do it. First of all, you have to have people on it and invest in it.
At Twilio, we haven't invested in the GSIs really so far, and we plan to make an investment in that area to get started. I think the other key is focus. One of the things I did at Salesforce was we probably were working with seven or eight or nine GSIs in parallel. What I did was I actually had them just refocus back to Accenture and Deloitte and just get going with those two and build momentum on two and really focus the energy of our business on those two. We're going to be looking for a couple of, not too many, but a very focused number of GSIs, and we haven't selected them yet, and I can't tell you who they are, but a focused number of GSIs that want to really build a big practice around this.
I think that with the momentum we're seeing in specific use cases like contact center, I think that's a very clear lightning rod to get people's heads around what Twilio is and what the opportunity is. I think that, Jeff talked about building platforms and going deep, and I think in certain use cases, we see opportunities to go deep, and that's the area we're going to focus with a narrow set of GSIs to get started, build those practices. Once we show success with a couple of them, my experience is the rest will come over time. That's the playbook.
Richard Davis. On the application services, just so I'm clear, if that's going to go from 10% to 30%, how should we think about the pricing model? Because not all of those components strike me as minutes resale and stuff like that. Are they per seat, per developer? At least notionally, how should we think about that?
Do you want to?
Yeah, I'm happy to take that one. Yeah, the way that we're pricing those engagement products, it's really by the business problem that we're solving for customers. Authy, for example, it's per authentication, right? TaskRouter, it's per task. Really what we're doing is we're mapping to either the individuals in an organization or the parts of work in an organization that we're solving. That's how you should be thinking about those. Yeah, Authy is a per authentication, per verification. That's the way we price it. Then it pulls through an SMS or voice. If it's using an IP channel, then we still price per authentication, and you just don't have the comms pull-through.
Lee, quick one on clarification on the revenue retention. Did you say quarter to quarter, in other words, Q1 to Q2?
No, quarterly year-over-year.
Okay, good. That's helpful. Second question is, could you give us some color on what % of your revenue is subject to competitive pricing versus not, either because the customer's too small, so they're not going to shop, or because you have multiple hooks into that customer, in terms of engagement layer, application services, and therefore it's too hard for them to compare Nexmo or whomever to your service?
Yeah. I mean, I'll start off and hand off to George probably, who's closer to that. Really how I look at the business, I do look at that revenue retention rate and I look at the expansion rate. Those healthy factors show that we have a differentiated product, and our product is doing very well. That's the level I'm looking at it.
Without going to specific numbers, I could say with confidence, we track this. We do track this number, a significant majority of our customers during the course of a year, we don't enter any type of conversation with them in the nature of what you're talking about, for the reason you talked about, either they're small and/or they're using a particularly sticky use case, let's say.
Okay. Brian White, Drexel. Jeff, obviously Uber had the biggest and negative impact on Twilio this year. I think there's still a concern out there that other customers could go down this path. If you could just walk us through why Uber is different and maybe also why they didn't give you more of a heads up before it started to have a pretty significant impact in the near term.
Yeah, I think the first thing to look at Uber is they're an outlier in our customer base with the size and scale they got to. If you look at their annualized revenue that they got to in Q4 of last year, it was a pretty close to a $60 million ARR customer for us. That's obviously unique in our customer base. That was by far the biggest customer. That's pretty unique for any software company to have a customer that gets that big, especially for a company of our size. It was a pretty unusual situation. The hard thing to do with a usage-based model being what it is that that is growth as a function of their growth. When it gets to that size, there is not a great answer, right? If it keeps growing, then you've got a worsening problem.
If it stagnates, then the question is, well, why is your biggest customer stagnating? If it shrinks, well, we know what happens in that scenario, right? What we've focused our energy on is say, let's manage the Uber account as best we can. We have a great relationship with the team. We expect them to continue to be a fantastic customer of ours. We've had great relationships across the company. Let's make sure, though, that we manage that, make that a sustainable part of our business, and focus on the rest of our customers that make up really the core of the customer base, right? The 60 or the 46,000 customers beyond Uber, right? That's what we've really focused on as a business. I think it's the right move, right?
We're building this company from a lot more than just one customer, right? We're building this company for the long term, for a huge market opportunity, and for almost every company out there who needs communications. We've got to manage Uber, absolutely. I do feel we have a great relationship there. It is a little bit of a lose-lose situation once you get to where we were, unfortunately. That's how we're thinking about the future of the business, just building the rest of the customer base up, investing in the success of our big customer base that we have, and focusing on the fullness of this big opportunity.
Pat again, JMP Securities. Lee, I'll try one more time. Your gross margin's gone down three quarters in a row, right? The 60% is really helpful. We're pretty far away from the 60% now. Have we bottomed?
Yeah, Pat, we're not giving guidance or speaking in terms of forecasts, but we'll stay quiet again. We are focused on the strategy of reach, scale, growing, and adding to our customer base. That's what we talked about early on at the IPO, and that's what we're focused on, and fluctuations have happened in terms of the gross margin line. As we discussed today, if we pull out some of those items.
Yeah
We seem to be pretty stable. We're looking at a long-term business. When we hit that long-term operating margin, we want to make sure we have a very large company to apply that to. That's what the focus is.
Okay. Then a more fun question for George. George, how is working for Jeff different than working for Marc?
George will say is that Marc sometimes buys cars for people. Just saying.
I don't know about that. No, look, I think they're both great CEOs. It's interesting because I've worked with and sold to a lot of CEOs at Salesforce and here.
It's interesting, I was in one of our last quarterly business review, I texted one of my former colleagues at Salesforce because Jeff was presenting, I said, "Actually, this is the closest I've seen with the vision, the passion." I was like, "I feel like I'm having kind of a Marc moment here." They're very different, but the thing I think that I love is that similar ability to paint a vision of the future and a desire to win and create a great company. One of the reasons I'm here, one of the things I learned being at Salesforce is that at some level, you can go as far as your technology, et cetera. Also at some level, you can only go as far as the desire, will, and vision of your leadership.
I wouldn't have come to a company that I didn't feel like the CEO had a will to win and a vision that was huge. I feel very lucky to be part of the team. Yeah, it's great to be here.
Thank you, George. You'll get your car later.
Thanks, Jeff.
One more.
Remember that next time I mess up.
Last question.
It wouldn't be any fun if you didn't close with the gross margin question.
There hasn't been enough, has there?
No. If application services doubled year-over-year and the rest of the business was growing 50%-60%, that implies that non-application services margins has gone down 50-100 basis points. Is that international mix? Is that noise? I think we're all asking for the same reason. If international continues to grow quickly, that means margins will continue to decline. I think a lot of us are in here for the long term, but at the same time, it causes short-term issues with the public markets, which means we have to go back to our constituents and explain what's going on. The more clarity we can get on that beforehand, the better.
Yeah, I'll start. Jeff might want to jump in, too. Part of that is us talking about how the platform model and how the platform model works.
This is a model that's driving high revenue growth extremely efficiently. We as a management team, we manage that gross margin line item like a hawk, but we're not fixated on that on a quarter-to-quarter basis. We're looking at revenue growth, great unit economics, and looking to grow the business moving forward. That's how we're viewing the business. We also value the bottom line. Operating margin is extremely important, and you have to look at both those in context. You have to look at the gross margin in the context with sales and marketing mix and how that ties into the overall model. That's how we're running the business and focused on the business.
Yeah, what I'll say is if you look at the vast majority of our customer base, as Lee shared earlier, right, we have a gross margin that looks like that 60% gross margin with the vast majority of the customer base when you remove some of the noisy elements that the platform model does inherit, right? The way we look at it, we're going to invest in growing the business. We're going to focus on the top line, we're going to bring customers onto our platform, and they're going to be all around the world. The question of will application and the application services that customers really depend on, will that revenue grow faster than customers adopt us globally? Can't answer that question. Are they both the right things to be focused on as a company? Yes, I believe they are.
We can't give you exact visibility. I mean, we don't have exact visibility about when those lines are going to cross in any way. The way we think about it, and the reason why we're comfortable saying, "Yeah, look at this. This is what we think the appropriate number is for the long term." Does it take a huge leap of faith to say, "Well, the vast majority of the 46,000 customers are already there," and then you remove the noise. When is that 46,000 going to keep growing and growing and growing such that the noise becomes a smaller element as that diversification slide that we showed is that we're consistently doing that and diversifying the customer base. How long does that take?
Can't tell you, it's not a huge leap of faith to say that, yep, those numbers are going to keep growing and are going to continue to dwarf the sort of outliers in the platform. The second thing is it's a huge leap of faith to believe that the application services are going to, in the fullness of time, be 30% of our revenue, given that we believe, and I assume you all believe, the future of communications is software. I don't think that's a huge leap of faith. Can I tell you exactly when it's going to happen? Unfortunately, I can't. I think those are the dynamics that are at play, and you can come to your own conclusions about how those lines will play out.
We think that in the fullness of time, clearly software will start to dominate, and that clearly the size of our overall customer base that continues to grow, that will start to dwarf the noisy elements of the business. We don't think it's a huge leap of faith to arrive at that number. I can't tell you when it will happen, nor do I want to commit to you when it'll happen, because we want to do what's right for our first goal, which is driving the top line of the business and getting a foot in the door and building great relationships with customers like we shared with you today.
Yeah. One quick follow-up. Are there opportunities like Netflix did with Open Connect, or just what you've seen AWS do with effectively shutting out the market with their volume and scale? Obviously, Google and Microsoft are there. Do you see any tipping points or any other step function opportunities that you can take to help reduce the burden of those network costs?
We continue to work with customers to understand their needs and how they're scaling and how they're growing. Christine and her Twilio Super Network team continue to work with the carriers of the world and build more sophisticated software and more sophisticated algorithms. Yeah, this is a continuous process for us as we continue to build out. We do continue talking to customers about what their requirements are and what they need from us so that we can meet those needs. I would say there isn't a silver bullet that I would point to. If there was, I'd love to point to that and say, "That's it, and we'll get it in Q4 next year." There isn't such a thing. It is a continuous process.
We are going to continue to build a better and better network and service our customers based on the direction they drive us in and do what's best for our customers.
Thank you. I'll end with a product question. Can I build a cloud PBX today with Twilio APIs? If not today, when might I be able to do that? Thank you.
Yeah, you can.
Yeah.
In fact
There are companies that have.
Okay.
Yeah. I mean, Aircall is sort of that. That was one of the logos on our slide today, is a Irish company who has built a PBX-
French.
Oh, French. French, sorry. A company that's built just that. Fantastic. All right. Thank you, everybody. Really appreciate you spending your day with us. Enjoy the
Thank you.