Good morning, everyone. Welcome to Five9's Analyst Day. I'm Lisa Laukkanen. It's nice to see so many familiar faces, and I also want to extend a warm welcome to those of you joining us on the webcast. I'd like us to take a look at the safe harbor language, and if this will cooperate. Oh.
You're back.
Little preview. Before we get started, let's take a look at the safe harbor statement. As a reminder, during the course of today's presentations, Five9's management team will make forward-looking statements regarding future plans, industry trends, company initiatives, and expectations around financial performance. Our actual results may materially differ from what we discuss today, and you should not unduly rely upon such forward-looking statements. Please refer to our most recent Forms 10-K and 10-Q under the caption Risk Factors and elsewhere in the company's filings with the SEC. We undertake no obligation to update any such forward-looking information. Additionally, you can find reconciliations of our GAAP to non-GAAP results in the appendix of this presentation. The deck will be available on our investor relations website at the conclusion of today's presentations. We've got a great day ahead, and I'm going to begin with introducing Rowan Trollope, CEO of Five9.
He'll introduce the leadership team and tell you what they're going to focus on today. I'll turn it over to Rowan Trollope, CEO of Five9.
All right. Thank you, Lisa. Thank you. Thanks, Lisa. Thank you all very much for coming and braving the little bit of rain. I heard there's going to be snow later, we may have you for longer than you were expecting. We do appreciate all of you spending your time with us today. We do have a lot to talk about. Some of you were here at the last Analyst Day that we did, I guess, two to three years ago, appreciate you coming back. Frankly, thanks for all of your support, all of our many investors for the support over the years. That really means a lot to us. As Lisa mentioned, I wanted to start out the day by introducing our new leadership team.
We have added quite a few new faces to the organization as we prepare and ramp up for this next phase of growth in the company. We brought them all here, and they're all here. Some of them will be speaking and sharing some of the deep dives on the strategy and what we're doing with the product, and our go-to-market. Then we'll also have them up for a Q&A, and then at the lunches. You'll get a chance to meet the new leaders. Without further ado, let me just introduce them one by one, and I'll start out with the man, the legend, Dan Burkland, our President, who leads our go-to-market organization. Dan's been with the company for 10 years and really created the Five9 that we know today, along with Barry Zwarenstein, who's our CFO. Many of you know Barry.
Barry's here as our CFO. Jonathan Rosenberg, CTO and Head of AI. Jonathan's joined us about 10 months ago. You're going to hear from Jonathan talk about AI coming up. Ryan Kam, who's our Chief Marketing Officer. You can see some of our new branding around here. David Pickering, our new Head of Engineering, is going to talk about product. Scott Welch, our Head of Operations, keeps everything running. Scott's in the front here. Anand Chandrasekaran, our Head of Products. Anand is going to share with you the product deep dive. James Doran, our Head of Strategy and Operations. James. Tricia Yankovich, our Head of HR in the front. We've clearly added quite a few new faces, and we did that because there's a big opportunity in front of us that we want to go execute.
We're going to share that with you. Before I do that, I'd like to start out just with a quick note on the vision of the company. You're going to get a chance to meet 10 of the leaders of the company today. If you were to talk to any of the 1,000 plus employees at Five9, you would probably hear that it's a mission or purpose-driven company. People have been in this industry for a long time. We've had this idea that customer service should be better than it is. That we're all users of, and sort of consumers of customer service, and most of us don't like it. I don't know about you, but I certainly don't. We all sort of feel that it should be a much more human experience, that we want to be treated like people, not like numbers.
That the only way we can make that happen, and I go back all the way to my days in the contact center, the only way we can make that happen at scale for businesses all over the world is by using technology to drive efficiencies in these organizations. That's what we're all about. We're all about doing this, and you'll find that it's a very purpose and mission-driven company. Just a bit of an overview. If you stand back and look at the Five9 story, three things really stand out, and it's actually very simple, not complicated, the story we're going to tell you today. First, is we have an absolutely massive market opportunity in front of us as the on-premises contact centers transition into the cloud, and as millennials and Gen Z enter the workforce demanding better customer service experiences and digital transformation.
It's just a massive and long-term opportunity for us to continue to grow, and we're in the first innings. The second thing that you should know about Five9 is we have a strong and proven execution engine. We've been executing reliably and consistently for years, and that's going to continue for years into the future. I'll take you through all of these. That'll be sort of what some of the meat of the presentation is, and you'll hear mostly from my leadership team as we talk about their specific execution plans that we've put in place over the last year or so. Finally, Barry will close out the day talking about growth and profitability. We have a balanced approach to running the company, focusing on growth and profitability. I think you like to say that profitability is a quaint idea, Barry.
We really do focus on growing the company in a responsible and digestible way, and doing that in a balanced approach. Those are the three things that to me stand out about Five9. Incredible opportunity in front of us, proven execution engine, and a balanced approach to growth. I'm going to start out and take you through the market opportunity and give you a little more detail, and I'll do that in three parts. The first is the key trends. What do we see as the drivers and the trends in the industry? What's changing over time? Second, I'm going to talk about the positioning of Five9 in the market and why we're well-positioned and how we're positioned.
Third, I'll talk about the addressable market and the expanding TAM story that we've been sharing with you for some period of time, and we'll share a little more color on that. In fact, as we do that, we'll even dive in later in the day into a demonstration of the technology that we've talked to you about so many times. We're going to actually show it to you, and we're going to even have a customer on stage who can tell you what they've been thinking about it. I'll start it with the key trends, and with this beginning idea that what I call the CX paradox, the customer experience paradox, and that is that companies globally are spending nearly a quarter of a trillion dollars in customer service. They're spending a quarter of a trillion dollars to deliver something that basically no one likes.
We do surveys and we talk to customers, and we talk to end users, consumers. All of you can probably reflect on your own experience. Anytime you call customer service, it's usually not a delightful experience, and it could be, and it should be, and that's what we want it to be, but it isn't today. This is the backdrop in which we're executing. That's a lot of money spent to do something that no one likes, and we don't think that that needs to be that way or that it should be that way over a long period of time. What is it that people don't like? This is, again, one of the key trends in the industry is you're seeing this change over time.
84% of people say that they don't like being treated like a number or being treated like a person, not a number is very important to winning their business. If you just dive into the very high-level things, and again, you can relate this to your own experiences, what is it that we don't like about the current way that businesses implement customer service? It's hard to talk to someone. It's very frustrating when you need to talk to someone. If you can get the answer online, if you can self-serve or find out from the community, it's awesome. If you can't, and the company doesn't make it easy to talk to them, that's a real problem for people. I personally have had that experience numerous times in the last year, personally. Depersonalized and disconnected experiences. Businesses are complex.
They know a lot about their customers, but when you connect into a customer service contact center, and they don't know that you just called an hour ago saying that you were upset and want to return the product, they shouldn't be upselling you on the next time you get a hold of them. Right? That is something. Even things that are as simple as you call into a call center, and they transfer you around, and the next person doesn't know who you are, and you have to re-describe everything. Of course, labyrinthian IVR trees, an attempt to avoid customers. People use technologies to sort of have customers jump through all kinds of hoops to avoid talking to them.
This is really some of the things, and of course, you can expand on this in your own mind, but this isn't an option anymore for businesses. It might have been fine to do this when your customers had no choice about what product they would get, or the switching costs were super high, but this isn't an option, and this is becoming less and less of an option for businesses to allow bad customer service to exist because we are all empowered as consumers. Empowered consumers, we can write reviews, we can complain online, we have a voice, and we've got choice. When you think about this just from an economics perspective, it's switching costs are going down. In virtually every industry, there is more choice than there's ever been, and it's easier than it's ever been to switch around.
I don't think you have to look any further than your own mobile phone. Start with your mobile phone and look at how many food delivery apps are you using or have you tried? Now, maybe I'm biased because I live in San Francisco. There's a lot of them. If you're in New York, there's a lot of them. In the retail space, it's similarly true. Travel. Every category, it just seems like there's endless choices, and it's two taps and 20 seconds to delete one and get another. Very, very low switching costs for consumers. All of that means that businesses recognize, uh-oh, we've got to do something about this. That increased choice is driving the expectations amongst the consumer way higher. 84% of customers say that the experience a company provides is as important as its products and services.
Now that doesn't necessarily, I think, resonate as much, apparently, if you're a boomer, but for the more modern generations, experiences are more important than products in many cases. You can see that here. 70% has the increasing expectations for customer service. That is something that businesses are waking up to. Business owners, business operators, service professionals realize, in fact, 82% of them say that their company's customer service has to transform to stay competitive. This is one of the tailwinds that's driving our buyers. 82% of people in our industry recognize they've got to do something differently. What they're doing is not good enough. It's not just the cloud that they need to get onto from a technology perspective. There's a very, very critical business need for them to solve, and that's where Five9 comes in.
Let's now pivot and actually talk about Five9's positioning. Lots of market trends pushing this market in our favor towards we need a modern platform to help us transform the customer experience. That's what Five9 is. We are essentially a modern engagement platform that lets you deliver the thing that you know you need to deliver, a much better customer service experience. How do we do that? I'll just dive in a little bit here. On the left-hand side, the way that I think about this anyway is that there's sort of two parties in this picture. There's the customer and the business. The customer has needs. You can all put yourself in these shoes. What is it that you want when you think about customer service? Well, number one. You want to be able to use any channel that you're comfortable with.
You might want to be able to SMS, or you might want to make a phone call if something's more complicated. You might want to send an email. You might want to get service by yourself online and not talk to anyone at all. You might want to hit the company's website. The point is that you want flexibility and you want to be able to get service anytime and anywhere, however you want to get that service. That takes a modern platform. You can't solve that with a phone system. That's what most companies have, is just a phone system and a bunch of disconnected experiences. We bring all of those together. All of them together into something that's very simple and seamless and easy to use. On the right-hand side is the business, and they need to deliver.
They have the imperative to deliver better and better service every year. How do they do that? Well, it's a complex story. They have to take lots of operational data from lots of back-end systems. Increasingly, those systems are moving to the cloud, they've got to take all that together and weave it together to deliver a journey for the customer that is very seamless. Those are the two players. How does Five9 help here? Well, first, I mentioned we serve the customers in the channels they prefer. We're going to deep dive actually in this in Anand's section, we talk about the next generation, actually reimagining what is called in the industry today, omni-channel technology. We're going to talk about that. We are one of the best at bringing all these things together from a consumer lens.
Second is unifying the customer information into a single view. I'll dive into that in one second. The third is really having a modern platform that can move flexibly and quickly for the internal constituents, the people that deliver service. Let me take you through each one of these. The first one is, on the customer side, is all about serving customers in the channel they prefer, as I mentioned. I won't dwell on this slide because Anand is going to cover it in depth. Suffice to say, we have a plan to reimagine how this is done in business, and we're executing on that, and we'll be delivering it in one of our-- in fact, in our very next upcoming release in the first half of next year. We've got really great thinking coming into the company, new, fresh ideas.
In fact, Anand joined us from Facebook, where he worked on Facebook Messenger for business. As you know, it is one of the most modern ways that people want to engage with business. We've got a lot of fresh thinking in this, and we're driving that forward. That's part one. Part two is unifying all of it into a single view. If you walk into today a contact center and you watch what actually happens, you will see basically people that have headsets on sitting around in front of computers. What are they doing? Well, they're handling inbound requests from customers.
What customer service professionals, sort of the main thing that they do, is they're trying to enable that agent to deliver great service, whether it's live, like on a phone call or a messaging platform or what have you, or non-live, whether it's email or servicing customers in a community or a news group or what have you. What they need is they need a unified view of the customer. That doesn't really exist in most businesses today, or it's very hard to get your hands on. Our platform is a flexible and open platform that lets companies bring all of that data together into a single pane of view for the agent. Of course, we're experts at powering agents. We sell software by the seat for agents.
That UI, which we'll actually share some pictures of with you today, it brings that unified view together. In fact, the acquisition that we announced this morning that Anand's going to share more about is all about making this even more seamless and letting us do more and letting businesses do more here. The third area and the third real constituent in solving this problem is the internal customer service team themselves. What they need is analytics and reporting and dashboards and the ability to do very customized and fine-grained control of routing, the ability to start to plug in chatbots to automate conversations, whether it's in messaging or in SMS or what have you. There's a huge surface area of capability that Five9 has built up over time to deliver the right level of engagement to the service delivery professionals in that whole organization.
There are many different types of roles that exist there, from quality management to supervisors to the agents themselves. There's just lots and lots of people that have to be solved, and we have unique user experiences and UIs that help them do that. Increasingly, you're seeing us expose that up 1 level to the business owners, because business owners sort of are now able to directly tie the journey of the customer back to things like revenue. You want to be able to get a view of, okay, how are we actually doing in terms of driving revenue or retaining our customers or upselling our customers or and so forth. That's the third part of the picture, and we are doing very well at that. We've been doing this for a long time.
We'd like to say it's a two-horse race or a duopoly in this industry. That was reaffirmed this year in Gartner's latest Magic Quadrant. As you see on the left, where Five9 again for the fifth year in a row is a leader. In Forrester's Wave, where we also show up as a leader. I would also point out, because I'm going to come on to this next around addressable market, that Gartner's Magic Quadrant does not list the two market leaders by seat size, Avaya and Cisco. You can see that there's no Avaya and no Cisco dot on there. Why is that? It's because Gartner has concluded that the future is cloud. It's an open-and-shut case, and if you don't have a cloud platform, you don't get onto the Magic Quadrant.
While those companies continue to serve their base and their existing customers, any customer who's looking to transform customer experience that calls Gartner is going to be told, "Well, if you don't have a cloud strategy, you need to get one." That's very much working in our favor. Let's now talk about with the sort of positioning of Five9 at the high level out of the way. I want to talk about the addressable market for our company, which I have mentioned is expanding. If you've been following the Five9 story, you're familiar with something that looks like this picture, right? We operate in a really big market, a $24 billion market, that has barely moved to the cloud. It's just at the very first innings of transitioning to the cloud, 15%.
A new stat for you is that while the 15% is estimated by Gartner and others, the large enterprise transition to the cloud is sub 10%. I think the number is 9%. Large enterprises are even further behind average businesses, and we're seeing that in our numbers. You'll see that in every earnings call, that what's really driving our business is the large enterprise transition to the cloud. This is our current market opportunity, and that's going well. Here's the expanded market opportunity that we've been talking to you about. To explain this, I'll give you the backdrop. We started out talking to customers and asking them what help did they need? Now that you've got Five9, our customers, now that you've got Five9 deployed, what else can we do for you?
What emerged from And some of them are here today, so you get to hear from them. What emerged was, it's great that your software is now installed, and it's working, and the first thing we need for you is to keep it working and keep doing what you're already doing. The second thing that we need is, help make us more efficient. Go back to what I said earlier, we know that we need to deliver a better experience to customers, but we can't just go spend lots and lots of money. We have to be very efficient with our spend. If there's anything you could do to help us automate some of the calls that we get, and increasingly do that would be very helpful.
We put two and two together, that plus AI, and we said, "Wow, we think this is a big opportunity." Again, we're going to give you a demo. What this chart shows is essentially the $24 billion that we normally talk about with the red dot being our penetration, and the big black bubble is all of the spend in the contact center. That includes labor. Take the $24 billion, + $210 billion annually in labor globally. Now, McKinsey did a study recently and said that by 2030, 16% of customer service reps will be displaced due to automation. They believe, based on looking at the workloads in lots of companies, that we should be able to see around 16% displacement of agents. What does that mean for Five9?
On the one hand, you ask me all the time, "Well, isn't that bad for you because there's fewer agents?" It's only bad for us if we're not participating in the technology arbitrage of that labor, which we intend to do. We intend absolutely to help make businesses more efficient by doing that, and we think we're really well-positioned to do it. If we do the math, 16% of the 210 in labor equals $34 billion in automation opportunities or savings that are available for customers. That could be the value creation in this market. The question that we're obviously going to be exploring as we move forward with launching our products is, what is our opportunity to capture that value, that $34 billion? Great opportunity to create value in the world, and what can we capture as a company? We're working on that.
Very exciting. More than double potential opportunity for us to go after, not just in the existing market that we're already in. In fact, this problem becomes bigger the bigger you get. The larger enterprises have this problem at scale, and it's worth solving. In the smaller businesses, it's much less worth solving. For a large company, and you're going to hear from one of them, or two or three today, it's really worth solving. It means millions and millions of dollars per year if you can do simple automations, and we're really well-positioned to do that. Jonathan will tell you about it. That's the first section on the market. If I had to sum it up, it's strong. There's two tailwinds.
I've been in tech for 30 years now, and I've seen industry transitions where there's one tailwind, and there's clearly been this massive secular tailwind of just the legacy technology needs to move to the cloud, and that exists in our space. Sometimes we get the question about the why are we so poorly penetrated in this space? Like, what exactly is going on with that? I've been explaining to some of our investors who've asked this question that what we've been seeing in the market and hearing from customers is that the first wave of B2B SaaS apps that started in the 2000s and kind of are continuing to now, were really the non-real-time apps. The Salesforce being a great example, sort of out market leading there.
Those were much easier to adopt because they didn't need a lot of heavy lifting on the infrastructure side to make happen. It's a website. You move your CRM records, you got to deal with all the security issues and the trust issues and the remoteness issues and all those other things. It wasn't like you needed gigabits of bandwidth into your building in order to make that work. That was not true of the real-time apps. The real-time apps that are now getting traction and accelerating, like Zoom or RingCentral, those require much better internet connections for average businesses, and us too. If you're going to plumb us in, we need a solid and reliable internet connection to make that happen. In fact, part of our business is actually helping businesses get their networks up to snuff so that they can support Voice over IP.
In fact, we have the co-founder of the Voice over IP protocol working for us, Jonathan. We've done a lot to really help customers get there. What's happening is essentially globally, because just the internet is getting more and more plumbed in, and the infrastructure is getting more and more robust, it's helping our market. That's one thing that's driving this second wave of B2B SaaS apps, the real-time apps. The second tailwind is digital transformation. That's everything we talked about at the beginning. When you start to talk about digital transformation, what becomes obvious is customer service and customer experience is sort of one of the main, if not the main pillars that business owners think about when they think about digital transformations. How do I transform the experience that my customers get with me?
We have these really nice two tailwinds both driving our business. That's the market. Bad customer service is not an option anymore. Customer service professionals know that they have to transform. This represents a large and expanding market opportunity for Five9, and we are very well-positioned to capitalize on these tailwinds. That's the market. Now I want to move and talk about the Five9 execution engine, somewhat legendary in terms of its consistency and reliability. I'll just take you through a few points, and then I'm actually going to have some of my leaders come up and take you through even more detail and double-click on these. Obviously, on the proven leadership team, we have retained the leadership team who was at Five9 before. Actually, just texted with Mike Burkland, who is on a beach in Mexico. He sends his best. He's here in spirit.
He built an amazing leadership team. We've extended and added to that leadership team with enterprise experience, people that are experienced in the enterprise space and also deep technical expertise. We've really built the bench of this company to drive growth. Second is we have a differentiated strategy, which I'll actually touch on next, but we've been working very hard to create that differentiated strategy that led to one decision that we made this morning, which was the acquisition of Whendu. We'll share more about that coming up. We, of course, have a market-winning product. I shared that the proof is in the pudding on that front in terms of winning deals, but also being recognized by analysts. It's not like we're starting from scratch here.
We're building on an already market-leading product. We've been extending our robust cloud platform to deliver the kind of quality that is required in contact centers. It's unlike, by the way, just about any of the other real-time workflows. Contact centers are much more demanding than meetings and calling infrastructure in terms of their demands because the contact center, when it goes down, businesses are losing money, or they're getting customers complaining on Twitter. If a business loses their phone system, Voice over IP phone system, people can jump on their mobile phones. Knowledge workers, it's not that big of a deal. It's absolutely critical that you deliver that. We've got a cloud strategy that I've talked to a lot of you about that's extending from our existing infrastructure into the public cloud.
We're now moving into a blended world, and we'll actually demonstrate one of those products that's built on our new cloud architecture today. Finally, we've got an incredible, and I would say, world-class go-to-market machine that we are building and expanding upon. We'll share some new developments. Dan will share some new developments on that front around things that we've changed in the organization and new faces that we brought on board. I'll close out just by sharing with you the five pillars of our strategy, and then I'm going to turn it over to Anand. The first pillar is all about product. For those of you who know me, I am a product leader. It's really important in a market where innovation is a differentiator that you've got a great product. What you're going to see from us is investment in product.
You saw that last year. We began to invest more in R&D than we had previously, the results are paying off. We're going to be able to share those results with you and show you what the team's accomplished. The second, of course, is executing against the longer-term opportunity that we see with artificial intelligence. I'm a big believer that every tech company will become an AI company. That's nowhere more true probably than the contact center. That won't happen quickly, it'll happen slowly, it'll happen quickly. We're investing in it in advance of that, we want to be at the forefront of making that transition happen. Because we're a relatively small company, we're not Google, we have to be laser-focused, you're going to hear that from Jonathan. We need to expand our go-to-market engine.
As customers are, if they're increasingly making the decision to move to the cloud, we've got to have the coverage, the feet on the street, the right relationships, leveraging channels, expanding our strategy there. You're going to get a deep dive on that from Dan there. Then finally, investing in our employees and culture. When I talk about the market opportunity, it's a 10-year horizon of people moving off of the prem to cloud. We've been building a leadership team and focusing on getting great talent into the company and retaining that talent. We have world-class retention levels in this company. The attrition rates are very low. We do a lot to maintain that, and we've been doing a lot to attract even more world-class talent.
We're bringing that talent on and investing in the culture because this is a long-term growth opportunity for us, and we want to see this company flourish over the next decade and decades beyond. Those are the pillars of the strategy. Now what we'll do is we'll introduce the leaders who will take you through each of the pillars in more detail. With that, with no further ado, let me introduce and bring to the stage our head of products, Anand Chandrasekaran. Anand?
Thank you, Rowan. Hi. My name is Anand. I lead product at Five9. Great to meet you all. I'm going to start the deep dives from a bunch of leaders and walk through the product roadmap and product vision. By way of introduction, I wanted to talk about my journey to Five9. Just before Five9, I was at Facebook. Over there, I was leading our efforts on Messenger for Business, which was the business platform on top of the Messenger app. We grew from about 150 million people to business messages. People to business messages are basically just like you would message your friends on the friend graph, you would message a business and get a response back, right? We'd count two-way messages.
We went from about 150 million a week to about $8 billion a week on the back of sort of customer care, right. That was just one channel. I got to see firsthand the sort of explosive growth of consumers wanting to message businesses when that makes the most sense and get a response back. Before that, I built an e-commerce product called Snapdeal, where, among other things, I ran customer care for about 20 million users. One of the things I'm really excited about is this opportunity that Rowan talked about, where we can make customer service more human. I'm going to talk to you a little bit about how we're doing that. Let me start with a recap of what Rowan talked about. There's a lot in here, but the recap was Rowan talked about three big things.
He talked about the need to serve across the channels that customers cared about and want to engage on. He talked about unifying the data into a single interface, and he talked about modernizing this platform of engagement. This is where the customer side, Five9, and the business comes together to deliver these pieces, and that's being delivered in the product today. I'd like to walk through sort of what we're doing today as well as where we're going from here. What I wanted to start off by doing, though, is talk about how we're growing the investment in R&D. For those of you who've been following Five9's investment in R&D, we're going from about a 3% year-over-year growth in R&D investment in 2016, all the way to about 27% year-over-year growth in R&D, Q3 2019, the YTD year-to-date.
This is largely driven by our enterprise customers, which now comprise about 80% of our total revenue. A lot of what we're focused on in terms of our roadmap is building for this large enterprise demand as we're being pulled upmarket by that segment. To summarize where we're investing these R&D dollars, what we have done is focused on delivering a greater number of enterprise-grade capabilities as we're getting pulled, and we're seeing even more demand by large enterprise customers. Then I'm really excited to also share where we're going, which is accelerating this large enterprise's transition into the cloud that Rowan just talked about, which we've barely scratched the surface of. All right, you guys ready? Let me talk first about where we have covered a bunch of ground in 2019. We've been busy.
The engineering team has grown significantly, and Dave will talk a bit more about that. This team has been incredibly busy. We've launched, in a series of releases, 90+ new features for our large enterprise customers. One of the things that I got really excited about when I joined Five9 is that our NPS is at about 58. By comparison, enterprise software NPS is about half of that. It's in the early 30s on the high end, and our NPS is almost 2x that, and we're definitely feeling that love from customers in response to these releases and features that we're building on the back of a large market-leading product. We've been focusing on enterprise scale, voice modernization, and open platform, and partner integrations. I'm going to get into a bunch of those things in the coming slides.
If I can take a moment to focus on the enterprise scale, that's one of the areas that we very clearly focused on. We're now tested with a pretty robust system, thousands of concurrent agents on a single domain with multiple channels and with a full application stack. This is both the scalability features as well as the add-on applications like performance dashboards, gamification, QM, WFM, speech analytics, et cetera, and we're tested for thousands of concurrent agents at this point. We have several customers here who've deployed these in production. We recently got the validation from Gartner in the MQ that came out earlier this year, that we offer expanded scalability to support thousands of agents on a single domain. Really excited about that. Secondly, our customers are moving us into global deployments.
One thing we're seeing is both their customers are global, their consumers, I mean, that call them, as well as their agents are global. In this environment, it's really important. We have a global system, and we call it FGV, Five9 Global Voice, and this is one of the features that we got out recently. It's got a global set of points of presences all built on a modern architecture on top of AWS. This is sort of how we support both a follow-the-sun need as well as a global need that our customers have. Really excited about doing this for our enterprise customers. We're also really proud. We have Scott over here, and Scott's team works really hard on this. We're really proud of our platform availability. We're 99.991% uptime the last 12 trailing months.
For those of you sitting out there doing the math, that's less than four minutes of accumulated downtime a month. Really excited about that. This is one of the reasons that customers move to the cloud because they don't have to sacrifice the uptime that they may have with on-prem. Not just uptime, but also security. The data that is stored in the cloud, we take extremely seriously, and we keep that secure through a set of things that we do, certifications and attestations. This is everything from AICPA SOC compliance, PCI DSS Level 1 certification, et cetera. Regulatory compliance. We're HIPAA compliant as well as we're compliant with GDPR, which is required for most global deployments. Now there's also a California equivalent of GDPR, which is called CCPA, which is also sort of rolled out now and we're compliant with.
Then in terms of best practices, we're compliant with the CIS as well as CSA compliance for our data centers. We take this stuff very seriously. One thing I wanted to call out here is that this is important, especially the regulatory compliance, because we're becoming compliant as a platform, so each of our customers don't have to get compliant by themselves. This is extremely important for our customers. To summarize, we've covered a lot of ground in 2019 and in the past years. We're delivering an even greater set of enterprise-grade capabilities covering enterprise class scale, covering global deployments, and all of these features that we talked about, and we have an NPS of just under 60.
I'd like to transition now and talk a little bit about this effort that we're doing to accelerate the transition to the cloud of our large enterprise customers. We find from talking to a lot of our customers, so since I've been at Five9, we've been talking to a lot of our existing customers, a lot of the new customers that are coming on board. It really comes down to these four things in terms of what is driving their acceleration to the cloud. Number one, we need to enable them to be where their customers are. Rowan talked about that, and that's a very key part of this transition. Number two, we need to play well with the other tools and the other platforms that customers have deployed. We need to play well with others.
Number three, we need to elevate our platform flexibility even more, and we have some exciting announcements there. Then number four, Rowan talked about the expanding market opportunity in terms of AI and agent assist. My colleague, Jonathan Rosenberg, is going to get into a deep dive and a video demo of how this works. I'm hoping today is the day when there are no "Star Trek" references. Although I wouldn't hold my breath around that. Number one, let's jump into omnichannel. As our own consumer experience shows, we all live in an omnichannel world, right? These numbers reflect what we all experience personally from our consumer behavior. 54% say that they've used email as a primary customer service channel in the last 12 months. This is from Forrester.
50% of consumers say that providing a seamless experience across multiple channels is the most critical issue during customer care. This is from Incite Group. 40% of people say that when they talk to an agent, they have to talk to a real person if it's a complex situation or a complex interaction. This is from Amex. This reflects our own personal behavior. The other channels only go so far, and one channel alone, that just doesn't work for us, given how we work. This is only getting worse with a new demographic, which is mobile first, mobile only in most cases, coming online. We're reimagining what omnichannel is. Today, the way the industry thinks about it is you take the incident, you route it in an omnichannel way to the right agent, and that's it.
That's what's called omnichannel today, is routing the incident. We think that because of this pressure from consumers, our businesses are being driven to actually route the customer, and we need to build a much more comprehensive set of features. In fact, we're building over 50 features to actually support omnichannel, and this is a huge focus for us. Let me talk to you guys a little bit about what we're doing in this area. We're investing, as I said, in a pretty comprehensive set of features covering a bunch of channels, applications, features, and platforms, and it covers these 10 major areas. This is just a very quick summary of what we're delivering across these four areas. We support all of these channels to be primary channels for consumers. We're delivering on all of these applications.
If you imagine being inside a contact center, all of the omnichannel interactions need to be tracked by the supervisor. It needs to integrate into WFO. The joke that we hear all the time that Dan told me when I joined Five9 is that when something doesn't get tracked in Supervisor, they don't know the difference between an agent supporting a customer on omnichannel or they take a smoke break. You don't know because it's not connected with a Supervisor console. These applications are extremely important in a market-leading product, and that's part of our roadmap there. All these features that you talk about in terms of delivering integration into the systems that they integrate with, as well as encryption at rest, which is extremely important, all these platform features. That's number one. We're doing these incredible set of features to reimagine omnichannel.
Number two, we have a strong track record, as many of you here know, of playing well with others. This goes from our CRM partners, Salesforce, ServiceNow, Microsoft, Zendesk, et cetera, Oracle, and we have market-leading integrations with all of them. We recently announced a product expansion of our integration into Salesforce to support the Einstein Bot. We have the ability to escalate from an Einstein Bot integration to an agent or a supervisor on Five9. Really proud of where we are with our CRM integrations. Similarly, we have market-leading integrations into UC partners like Zoom and Microsoft, as well as WFO partners like Verint. This is across the board, and these are part of the prioritized 90-plus features that we released to actually deliver on not just CCaaS features, but also integration into these systems of record, if you will.
We're also making it really easy for our partners. Yesterday, a bunch of you may have noticed the launch of the App Marketplace. This is a step in that direction. We've had UC and CRM SDKs available, and we've done the integration into Salesforce and Oracle and a bunch of our strategic partners. We also have a CRM SDK, which the smaller CRM vendors have integrated into directly, like Pegasystems and Freshworks, et cetera, have been live with our CRM SDK. Now we've given them a single merchandising window to be discovered. This is extremely important. Dan will touch upon this a little bit more. This is important when customers are evaluating the solution.
They need to know what ISV partners are available, they need a single merchandising window to look at all of the solutions, and that's what you see on the right. We have over 55 partner solutions live today, and available to integrate into Five9. This, as you can imagine, provides a virtuous cycle for our partners to leverage the Five9 incumbency and actually integrate into the solution. That's number two. Number three, as we've been talking, we are seeing large enterprises deal with some challenges as they migrate to the cloud. We've been talking about that. Rowan talked about this acceleration and where it is today. It's only at 15%, and as we peel that onion and look at the layers, what we find is on-prem solutions that they might have deeply integrated workflow with their in-house information systems.
They all have some custom system, or they all have some system of record that are integrated into within their in-house systems. The current approaches give them sort of a false choice. On one end, you've got approaches which are low customizability, but they're easy to implement. On the other hand, you've got solutions that you can do whatever you want, but you got to call developers, and you got to get 20 people to come in and build that solution. Of course, we all know in most enterprises, save very rare exceptions, you don't have developers lying around waiting to do these projects. These are false choices where you only get one and you don't get the other, and that's one of the biggest reasons we feel enterprises find it challenging to move to the cloud.
We've been talking about all of these features that we've been building to help our large enterprises migrate to the cloud. Today, we're really excited to announce another big step in helping our customers do this. Today, we're excited. You probably saw the announcement this morning, and Rowan touched about it in his talk track. We signed a definitive agreement to acquire Whendu. Whendu is a small technology company based in California that's an iPaaS system that helps large enterprises accelerate their adoption of the cloud. Really excited about this. How does Five9 and Whendu work? I'd like to talk about that for a minute or two. We help solve two major problems. We just talked about how challenging it is. Number one, we help large enterprises migrate to the cloud. How do we do that?
Number one, there's over 50 out-of-the-box application connectors that large enterprises can use to maximize the cloud contact center investments by having these integrations. We'll get into how that works in a second. Number two, we make it easier to integrate Five9 into their existing environments. We just talked about how that's a huge part of on-prem deployments. Carry forward their customized workflows as they migrate to the cloud. Really excited about this. This is a huge step forward in easing the pain to migrate to the cloud. Number two, once these large enterprises get to the cloud, they're really excited to do more. We're seeing the rise of what we call citizen developers. These are basically someone who took some programming in high school or college. They're technical enough, they're not a developer.
They're not hired in the company as such, but they're usually the business analysts, they're the supervisors, they're technical people in the contact center who can actually write a bunch of workflow to make their integrations work and to make these disparate systems come together in real-time. We're building this integration between Five9 and Whendu to actually deliver a bunch of features for the citizen developers to connect information systems in real-time. Let me show you how this works. You're probably wondering what a no-code system is. Let's take the example that you need to follow up with a lead in moments. This is a very typical example that most large enterprises have to deal with between a CCaaS system and a system of record. In this case, this is the actual workflow on Whendu that you can create in a matter of minutes.
We took eight of the large systems that we've built out-of-the-box application connectors into, and you can actually just literally plug the logic. As you can probably intuitively guess, the logic is when something happens, do something. That's why you have Whendu. When something happens, which is a company's name is mentioned, categorize it, see if it's a support request or a sales inquiry, see where the sentiment is, create a case, tweet out a response on the same channel. If the sentiment gets even worse, and it's getting really negative, mark it as a high priority and call the customer ASAP. If you think about it, this previously would take millions of dollars and maybe 6-12 months to develop, and it would need actual developers.
With this no-code visual application workflow, it can be done in days, if not hours, and it can be done by a non-developer. This is just one example. You could actually go ahead and do all of these other integrations, like escalate omni-channel support, and that's the workflow for that. Monitor SLAs proactively and take action. If you miss SLA, add overflow agents during demand spurts and respond to mentions on social channels, et cetera. You probably got a little taste of this. This leverages out-of-the-box connectors into all of these well-known systems. You probably recognize a bunch of the logos. Really excited about the simplicity and the elegance that this brings as we ease the migration to cloud. This also really helps turbocharge the Five9 partner ecosystem. We just talked about playing nicely with others. There's already 55 partner solutions live.
Yesterday, we announced the App Marketplace. It's live today, is in production, and we talked about all of these strategic integrations. Whendu turbocharges that with 50 plus out-of-the-box application connectors. All of these logos that you see, many of whom I'm sure you recognize. Really excited about what this does, not just to the customer ecosystem, but to the partner ecosystem as well. All right. To summarize, in terms of where the ground that we've covered, we've been delivering a greater set of enterprise-ready capabilities, and where we're going is working on these four things that we talked about, omnichannel, integration into partners, platform flexibility, and now I'm excited to invite up Jonathan Rosenberg, our Head of AI, to talk about our agent assistant AI roadmap. Thank you.
All righty. Thank you, Anand. Thanks, everyone, for joining me today and for joining us here. I'm really excited doing Oops, wrong way. There we go. Let me tell you a little bit about my journey to Five9. It's actually pretty relevant to what I'm going to talk about. I came from Cisco, as you heard from Rowan, and I ran our AI initiatives there, as well as was the full technologist behind the Spark and Webex Teams technologies and cloudifying on-premise software. Before that, I was at Skype as the Chief Technology Strategist. My main claim to fame is that I've been working in the internet standards bodies, this thing called Is this not working?
Yeah, the webcast. For the webcast.
For the webcast. All right, double miked. I've been working for over 25 years in the internet standards community, where I'm known as the primary author and co-inventor of a technology called SIP, which is now the foundational telecommunications technology. You all use it every day. It's built into every cell phone on the planet and runs every single phone call pretty much that you probably make. I've written a lot of internet technologies. I'm the third most prolific author of internet standards. I went and I put it in a spreadsheet, 2,559 pages of documentation that I've written in my career on how to do things on the internet. I don't know whether to be ashamed. That's a lot of stuff. That's a long time to read. It's pretty boring, too.
I've worked in this area for a long time, and I've seen the waves and the changes in technology. When I came here, I was sort of really excited by the mission that the company was on, and you heard from Rowan, it's to make customer experience more human. The first question to ask is, well, what's our AI strategy? Well, sorry. It's exactly the same thing. Let's see what I'm doing here. It's to make customer service a more human experience. It's the same. It's not something different. How do we use AI to accomplish this task? The opportunity really from technology side is kind of exciting.
If you look at what we've been doing since the launch of Voice over IP, when we published SIP in 1999, we've mostly been looking at moving the voice bits from one phone to another phone. The network in the middle didn't really do anything with the bits. It just moved them around. What we're seeing now with the arrival of technologies like speech recognition and natural language processing in a cost-effective, scalable way is the software can now play a role in analyzing the content and take this dark data that we've never had visibility to and start to bring value to it. That's the technology opportunity. The product opportunity is the one Rowan already talked to you about, right? This is an area where a lot of people who use the software and use the products, they don't actually like it.
That just means there's opportunity for improvement. I think, in fact, to give you some examples of things, these will probably resonate with all of you, right? What are these pain points? Who here has called a contact center? I mean, come on, raise your hand, right? This guy over here, you've never called a contact center. Okay. Yes. Just looking at me dead on, no hand. All right. Why do I have to say who I am, like, three times? Why don't they know why I called? You just got a thing that says, "Hey, your bill is overdue," and you paid it. You call them up, it's like, "Thank you. Can we add a subscription to your service?" I mean, come on. Why don't you know? You sent me the notice, right? Why don't they know what steps you've already taken?
Probably my favorite example is that call to the cable company, and they tell you your cable box is broken. "Have you restarted it, unplugged it, re-plugged it in?" I mean, come on, I did that like three times already. Why don't they know that? Why do you get put on hold all the time? Why do you need to repeat everything? Why do you have to wait for the agent to type things? I call this the clickety-clack effect. The agent asks you a question like, "Hey, what's your name?" You say, "Bob." They type for 10 seconds, clickety-clack, clickety-clack. What are they typing? All I said was Bob. There's this long wait you have to take, and then you don't get a response quickly. It takes a while to resolve the issue. Sometimes the agent's not empathetic and friendly. Sometimes you have to call back.
These are the pain points, and there's so many of them that there's just lots of opportunity to make this better. Our strategy to capitalize this is to use AI. If you look at what people are doing with AI in the contact center, broadly speaking, you'll see the market and product splitting into two categories. The one you're probably most familiar with that gets probably most of the buzz is this one here, which is intelligent voice assistants, also known as IVAs or voice bots. This is sort of the next generation of your voice response systems, where as a customer, you're talking to a machine that's answering your questions and interacting with you in a natural way. The other is its sort of lesser well-known stepbrother called agent assistance. With agent assistance, it's different.
When you finally do connect to an agent, you still talk to an agent. It's a human-to-human conversation. What we do is we have the AI listening in on the call, processing it, and providing guidance and input to the agent. That's agent assistance. When we went to look at what our strategy is, we actually decided that we're going to focus on agent assistance. There's a bunch of reasons for that. I think probably the most important one gets back to the end user, the customer. There's a real risk, we think, that overuse of voice bots is going to lead to cases where people are frustrated. They call up, the voice bot doesn't understand, and everyone's going to just say, "Agent, agent, representative." Just like we do in IVR systems today, right? See, you laugh. I could tell you're one of those people.
With a voice bot, since there's a risk it gets worse, with agent assistance, it does not. We're still talking to a human. It's important to understand, especially for folks, artificial intelligence systems, machine learning systems always make mistakes. They are not 100% accurate. As a result, it's not a question of if, it's a question of when. When these systems do not do the right thing, what is the impact on the experience? With agent assistant, the answer is there's no negative impact. We kind of like that. It also, like voice bots, also saves money for the customers, which is critically important to deploy these technologies by reducing what's called handle time, how long it takes the agent to handle the call.
Most importantly, because of this blue line right here, there's a feedback loop from the agent back to our software that allows it to be self-learning. A feedback loop that is absent, interestingly, in the voice bot case. It leverages our strong services capability that you're going to hear more about from Dan. What we're not doing, because as you heard from Rowan Trollope, we have to be laser-focused, we're not building our own voice bot. We're going to partner with them and allow them to integrate into our system, and there is a market for that, no doubt about it, and it will find use cases. We actually think the much more exciting and larger opportunity ultimately is this one. We're also not building our own AI models.
I think there's this misperception that in order to do anything AI, the first thing you have to do is hire 30 PhDs from Stanford or MIT. The fact is, you do not anymore. What we've seen is that this technology has become commercialized by vendors like Google. Yes, I actually have a PhD from MIT, well, Columbia, but that's why they're laughing. Anyway, totally threw me off. Thank you for that. The point is, companies like Google and Amazon and IBM, they're taking this technology and they're making it available as a Software as a Service platform, so that you don't have to have these armies of experts. You can use their tools. Effectively, AI is becoming commoditized and democratized, and we're leveraging those platforms to reduce the engineering investment we require, yet taking advantage of all that they have, which is a lot.
That's what our strategy is. You can think about what we're trying to solve with a diagram like this. Here we have our agent in the middle, the customer on the left, and we have some software here. What happens is the customer will call up the agent and ultimately they have a question or an issue and they want an answer or resolution. That's the fundamental problem the agent has to solve. How does the agent solve it? Well, they ask lots of follow-up questions to get more information. They start to gather information from the user that's called up, and they combine that with their training. Of course, while exhibiting empathy, what they do is they enter the information in these back-end database systems and software applications, tons of them.
In fact, agents are notorious for having 10, 20 tabs open in their browsers as they take information from one website, put it in another, get something back, ask you as a customer a follow-up question, hear what you say, go into another back-end system, get the result, and that's all that clickety-clacketing. In fact, every time you go on hold, that's the agent doing something like that, unless they're just bored and are doing Facebook or whatever. Most of the time they're not doing that. They're doing this kind of data entry. In a lot of ways, sort of our market opportunity is at least equal to the amount of time people spend on hold in the contact center, because we can take that and we can automate our way through that.
What we're going to do is we're going to take all of this, which is ultimately taking this human voice from the customer and translating the back-end systems, and we're going to take that and we're going to turn that into the AI. The AI is going to interact with the agent, and then when the agent says okay, push all this information into these back-end systems and interact with the agent in the process. Let me give you an example of the call that most of us dread, that most of us hate, which is calling the cable company when something doesn't work. I find a lot of people resonate with this problem, right? I personally have this problem. What you're seeing here is a screenshot of the software that we're developing. Let me explain to you what this is.
This is what the agent will see. This is not what you as the consumer would see when you call the cable company. The agent is looking at. Here we have on the left is a live transcript. As the agent and the customer are talking to each other, this shows up over here. That's actually not the important part. The important part is our natural language processing technology is analyzing that conversation, and it is figuring out information to present to the agent. The agent clicks these suggestions and it brings up what we call push cards. These cards contain important information, guidance, and actions, more importantly, for the agent.
In this case, the customer calls up and says, "Hey, I'm having trouble setting up my new DVR." In this particular case, when the software hears setting up trouble and DVR, our natural language processing system says, "Oh, I think I know what's going on." Through automation, it looks up the customer record in the database. It goes to the IoT and device management platform that the cable operator has, looks at the status of the DVR, sees that it was just rebooted 10 minutes ago. Relevant information. Provides this information right to the agent, so they don't have to tell you to do something you've already done. Then it runs a troubleshooting script which goes through the steps that the agent would otherwise normally go through in a manual way. The next thing it checks is the firmware out of date?
It turns out that the firmware is out of date. It surfaces that information to the agent and then tells the agent, "This is what you do. Just press that button," the automation will then go and update the firmware in the system. The conversation goes like this. The customer would say, "Hey, I'm having trouble setting up my new DVR." The agent says, "Oh, I see you've just rebooted it. Your firmware is out of date. Can I fix that for you right now?" As an end user, it'd be like magic, right? I mean, you didn't even finish your sentence, the agent already knew what you did and has your fix ready to go with your confirmation. The agent hits reboot now.
The system also knows that, in this case, the best time to upsell to a customer value services like HBO or something is while they're waiting for these processes to complete. In this case, it analyzes the customer propensity to buy and what have you, and it recommends to the agent to ask the customer if they would be interested in HBO. This is a really exciting capability that we've been working on. To make it even more real, what I'd like to do is to bring up on stage one of our customers who's working with us on this program, and that is Ed Albrektsen, Director of Voice Solutions and Audit. Ed, would you please join me on stage? Give a warm welcome for Ed, please. Thanks, Ed. Thanks for joining me here, and thanks for coming and being such a great customer.
I don't know that everyone is familiar with McKesson, although, gosh, they probably should be. Tell us a little bit about McKesson and your business?
Okay, great. Thanks. Good morning. I'm excited actually to go through this exercise with you. Most of you know McKesson as the Fortune 788-year-old company that distributes pharmaceuticals and medical and surgical supplies. What we're going to talk about here today is the division of McKesson that focuses in on the patient, specifically on the patient and the interaction between the patient and our agent. This really is the representation of the pharmaceutical manufacturer. That's really, they think they're calling the pharmaceutical manufacturer when they reach our agent. In this particular case here, the interactions between the patient and the customer are being essentially done with our agents and handled by our agents. With what you're going to see here shortly, this is really a game changer for our business.
In today's business, we are required to collect and record and report PAEs, and PAEs are potential adverse events. You generally would know them as side effects. So these are a requirement. Keeping in mind that the average duration of our calls are about 20 minutes, and in that 20 minutes, there's a tremendous amount of information that goes between the patient and the agent. The agent is required to document that and whether they're typing, as you had said before, and all the clickety and clacketing in there, whether we know the patient and we have the patient information in advance or not, these are all things that vary on the program and the duration of the program.
I would say that, in looking at that information that is actually transcribed into or the agent has to type in, we're really looking at this opportunity here to, A, collect that information so that we're in compliance on it, and then also to speed up the call. It's a matter of the accuracy and then being able to document all that information is extremely important to us.
Great. Thank you. It's been great to work with you and excited to share with you where we've come. Ed, thank you so much.
Yes. Great.
Much appreciated.
Thank you.
Thank you, Ed.
Yeah. This is fabulous, and thank you very much.
Oh, thank you.
It's just been a pleasure. Just a few more comments.
Oh, sure thing.
We've been a customer of Five9 since actually 2013. We transitioned our data center over in 2014. We are in eight different locations across the U.S., but the most important thing is that we've grown with Five9. I know I'll get a little chance to talk a little bit later on there, but I wanted just to add that in there, that we are one of those early adopters of the cloud in an enterprise setting. I think that we were way ahead of our time at that time, in 2014, for an enterprise business like McKesson. I'm happy that we've been able to progress not only with AI, but all the other services that we'll be able to talk about later. Thank you.
Thank you. Much appreciated. All right. Without further ado, let's actually see it in action. Now, I'm not going to do a live demo. We record for time purposes. I've got a recording that I've broken up into a few snippets. We're going to start, and hopefully everything works.
Good morning. Thank you for calling the Ketoprofen Support Program. How can I help you?
All right. What you, again, just you saw now the start of this, the patient called up the agent, and they started to speak, and you saw how it was transcribing nice and fast in real time. What happened is, because this was the Ketoprofen Helpline, and we knew this information about this customer, we already, just as they started to talk, figured out that this is a case where we need to surface these key pieces of information to the agent. Namely, collect their reaction, talk about the drug that they're taking, and general customer information. This is what we call the Reaction Push Card, and that has all the information there that either exists or the agent needs to collect for the purposes of this call.
Now, what you're going to see next is this critical moment, all right, where the patient is going to talk about their side effect. Today, the agent would have to clickety-clack through that whole thing, type it up, and delay the conversation, especially when it's a lengthy side effect. See what happens here. This is where the agent would otherwise type this. Watch what happens.
Hi, my name is Taylor Smith. I've been taking Ketoprofen now for a while. I just developed a rash underneath my arm. I'm supposed to take my next dose this morning, but I'm not sure if I should.
I'm sorry to hear about your rash, Mrs. Smith.
There you go. Automatically pulled that out of the transcript and put it in the right place. The agent didn't have to do anything except just look and make sure it's correct, as it is in this case. That's an example of a form fill. Again, we just saved a bunch of time. The next piece I want to share with you is confirming information, right? In this case, we have to check the email address and the phone number and all that kind of information. Our automation has also gone and pulled that information out of the CRM. There's tons of information in that customer relationship management database. It's surfaced just what the agent needs to see in one little click. We'll see that portion next.
I am required to let you know the information you provide today regarding your experience will be reported to our safety department. A Ketoprofen representative might want to reach out to you by mail. Do you prefer to be contacted by email or physical mail?
Oh, sure. I prefer email.
Is your email still taylorsmith@five9.com?
Correct.
Again, there you see just a simple confirmation, nice and fast. The information was all there at the agent's disposal. The last part I'll share with you, again, is automating the collection of data. In this case, the agent is missing a critical piece of information, which is the dosage. We're going to hear the patient speak the dosage, and it's going to automatically put that in the form, too.
Mind if I ask the reason you take the medication and the dosage that you take?
Sure. I take it for my rheumatoid arthritis, and I take a 25 mg tablet three times a day.
Thank you. In this case, we aren't equipped to make specific recommendations to you. Your doctor who.
I just cut it off, so we're done. This was not staged. It looks like someone was just typing, Mechanical Turking it, absolutely not the case. I just did a video recording of it. The last step of this would actually be the real awesome part, which is the big green button of happiness. The agent just hits that button, and under the hood, it goes, and it pushes that information to the FDA databases, to the McKesson internal databases that have to collect and process all this information today, a process that's extremely time-consuming and tedious. What we've done with this thing is we've just saved the agent a ton of time, which saves McKesson a lot of money. We've made the patient experience much better, and at no time were we putting anyone under risk of incorrect information.
The agent, if something was wrong, this is what I mean about what happens if it messes up. If this had said 35 mg here, the agent can still correct it. Just imagine in other cases, if there was no human being confirming, we could have serious problems. That's why we're really excited about it, and at its core, what we're doing is we're taking three totally different capabilities that exist in other parts of the contact center industry. I've shown you some AI, speech recognition, and natural language processing. We're not the first to do speech reco and natural language processing in a contact center, but we've taken that bit out of other products that do that. Then what we've done is we've looked at automation. There's lots of products that do automation and workflow, but we've taken some automation capabilities here, and then scripting.
This is a feature of most contact center systems. Our current product has it too, of course, that allows the presenting forms and guidance to the agent, and that's existed for a while, but it never had anything to do with the live voice content of the conversation. That's what's new here. By taking all these three things together and making them work in real-time, we're actually creating something entirely new. We're really excited about this. I'm excited about it. I view this as the next stage in really the revolution that has been modern telecommunications. It's going to start in the contact center and expand from there. Where are we? We locked down this strategy many months ago now. We validate it with both industry analysts and with customers in our advisory board meeting back in the spring.
We hired a team, and we've been ramping up. It's a small team. This is not an army. We signed up a bunch of alpha customers, of which you've heard just from one of them, which is McKesson. We're currently going through this alpha trial process. We're actually building this product in conjunction with these customers to make sure it absolutely meets their needs. We'll continue to develop it and work on it, and we're really excited about the opportunity it brings. Of course, a big part of how we're able to do this is because we're using a modern and agile engineering engine that I'd like to bring my colleague, David Pickering, up on stage to talk to us about. Dave Pickering.
I don't know if I need this or not. Man behind the curtain?
Yeah.
Okay. My name's David Pickering. I've been at Five9 for about 11 months. I'm the EVP of Engineering. Here's my journey to Five9. I did not come from Cisco. I came from Intuit where I ran QuickBooks Engineering for about six years through a period of really strong growth in that sort of core SaaS business. We grew about 10x to 2.6 million customers. While doing that, we also went through the transition to get into the public cloud, to get Dockerized and containerized, to get automated, to push our releases. We would release about 300 times a month, just my team. Before that, I worked at Tapjoy, which was a smaller company that was large scale, sort of targeted marketing, alternative payments processing. Before that, Oracle, where I started their interaction center platform way back, I don't know, 20 years ago or so.
My key objectives when I got to Five9 were, first, Five9 has a great track record of delivering software, delivering it on time, delivering on quality, uptime, all of those kind of things. Number 1, keep the trains running. Don't mess anything up. Number 2 was grow the engineering team. As Anand talked about a little bit, we're looking to expand the engineering team, expand our capabilities, grow that team, get some world-class people in to add to our already existing talent. Number 3, those two both were sort of referred to by others. Number 3 was accelerate the speed of delivery and update our engineering platform. That's the one that I'm going to focus on most in the next few slides. I am maybe coming at this from a different angle than most people. I'm an engine guy. I like building teams.
I like building development capability. I like horsepower. I like cornering. Those kind of things. What you want to build with it, I'm interested, but that's not my primary focus. My primary focus is how to get horsepower, how to get cornering. This is the basic answer to how you get horsepower and how you get cornering in a modern SaaS platform. Public cloud, get onto public cloud. You need a public cloud strategy. Maybe you have your own cloud. We do. We have that for good reason, but you need public cloud as well. Containers, this is Docker and Kubernetes. It's how you deploy architecture in a modern system. Decompositions, you can release individual parts of the product, roll them back, roll them forward, those kind of things. Data, that feeds all the AI strategy. It weaves through everything, particularly when you're a cloud company.
Interfaces and APIs and automation. These are the things that allow us to accelerate, to go faster, to corner. Critically, it's what turns the traditional release train that you wait for, and it comes once a year, into more of a release Uber, where when that individual piece of software is ready to go, you call it, you ship that piece of software. If there's any sort of a problem, you can ship it again, you can iterate, you can solve your customer much faster. Why does this matter? Okay, just deliver more customer benefit. You get more output when you do this stuff. You deliver more stuff. It's collaborative. Again, the release Uber model. When we're working with McKesson on something for AI, we can release that independently.
We don't have to wait for 11 months or four months or two months or any other months for a release vehicle. That's important to us. Be humble. You need multiple shots on goal. You do not get one shot on goal and actually nail it in the real world. You need multiple attempts at it. You need to work with customers. You need deep expertise. The modern innovation engine, which I'm going to go to a little bit more into the next slide. This is a platform as a service over here where it says service mesh. This is effectively all the stuff that we don't write. That's basically what modern platform as a service is about, is I don't want to write release, I don't want to write Docker, I don't want to write Kubernetes.
I don't want to write a pub/sub bus. I don't want to write any of that. I want Google to write that stuff. I want to sit on top of Google. I want to sit on top of open source. That's what all this service mesh is. That's effectively the stuff we don't write. Over here where it says Five9 core IP, notice that's in the Five9 cloud and that's really important. Public cloud, Five9 cloud, all that functionality is accessible to me when I'm building a new component. When I build a new component, oh, great, agent assist, Jonathan just talked about it. I'm sitting on top of both, right? I'm building it using pub/sub and using Kubernetes and using Docker and using containers, using all of these things that I'm not using my engineers to write.
I can have my engineers do our differentiated software, the things that we do better than anyone else in the world. It sits on top of this core IP, the deep roots that we've been building up for 15 years. It sits on top of Google and AWS, all of these reusable components. That allows us to go much, much faster and use our software and use our people much more efficiently. Here's the result, right? Which you just saw from Jonathan. In three months, we went from a customer advisory board where we're having a meeting with McKesson. Jonathan is presenting this. This is a mock-up of what we later made. Who's interested? Does this fit anybody's bill? Anybody want to work with us on it? McKesson says, "Yes, we do." Fantastic. Some other people did also.
We're able to build this, take it from concept, delivered in three months. Why is that? There's no way I would've been able to do that except I'm sitting on top of Google. I don't have to invent all the stack. We've got all of these core capabilities that are in the cloud from Five9, from 15 years of doing this stuff, that I can leverage, right? That's the key. That's an amazing efficiency that all comes about from modern SaaS, and that's a really big deal, and that's why I'm here at Five9. Dan Burkland.
Thank you, David. There you go. I've been informed that this will suffice. Very glad to be with all of you this morning. Lots of familiar faces. As most of you know, I've been at Five9 for almost 10 years. It'll be 10 years next month. Very excited to share with you this morning about our go-to-market and service delivery. Let's get right to it. First of all, why are we confident? Why are we so excited about our future? I'm going to take you through this in really two different ways. One is to look at it from a customer perspective. What do our customers go through in their life cycle in working with Five9?
Then we're going to talk about how we can scale this go-to-market engine to really capitalize on the opportunity that Rowan shared with you earlier, and the tremendous upside that we have in the years to come. We think we're well-poised and positioned better than anyone to take advantage of that. First, let's jump in on our real customer life cycle. For those of you who are familiar with us, you've heard me talk for years about our ability to continue to capture and land larger and larger enterprise customers. That hasn't always been the case. Our industry was challenged years ago with, when is this going to take off? When are we getting A lot of you asked those questions. When are you going to be successful in really working into the large enterprise space?
We'll talk about what those challenges were and what we've done to overcome them and be successful today and set ourselves up for the future to go even larger. We'll also talk about the adoption phase. Really, what do we do to enable customers to really take advantage and get the value out of the partnership and the technology that we're delivering to them? That includes not only the technology stack itself, but really what we do with it and how we tailor it to the customer's needs. We'll also take you through the expand phase. We talk about land and expand, a term everybody throws around. Every SaaS company talks about the importance of doing that. It's not done automatically. It's done with a very concerted, focused effort, and we believe we have a formula for success with our clients that's unparalleled in the industry.
It's been lots of years of trial and error and figuring out what is the right recipe, what's the right formula to help us succeed with our customers. It results in really that retention and the retaining of our customers. Again, that's not done automatically. It's not done with technology. It's done in many parts with people, and very experienced people that can come in and consult and work with the organizations. Let's jump into that piece of it and talk about first the land. I talked about it hasn't always been the case where we could just go in and land large enterprise customers.
We talked with Gartner and other analysts early on, five to 10 years ago, and said, "What is it going to take for us to be successful at market?" They helped us formulate really these five key points that are so important. Scale means a lot of things to a lot of people. It doesn't just mean building a capacity in the platform that meets the needs of that client. Because remember, large enterprises, many of them are going to go through M&A, they're going to go through organic growth, and they've got to be protected for, in many cases, not only 2x growth, but 3x, 4x, and so on. We've got to build a highly scalable platform, which Anand shared with you earlier, that we've done. We also need to develop and deliver platform reliability.
Scott Welch's organization and the Cloud Operations is making sure that we have an architecture that is always on, an achievement that is very difficult for enterprises to accomplish on their own. Being able to bring the redundancies, the failovers, and the ability to make sure that for a mission-critical application like contact center, that we're allowing customers to keep that platform always on and always available. Security and data protection, whether it's compliant reasons for what Anand talked about with PCI or HIPAA compliance, or just making sure we encrypt and store data in a safe method that gives customers the ability to trust that we have their data and that we're going to take care of it as well, if not better than they can. Okay. Then feature parity.
No one's going to move to the cloud if they lose or compromise capabilities that they've had historically in their existing premises-based solutions. We had to emulate all the feature sets from all the leading providers historically in the space. Only then can you give them a platform that will allow them to then enhance that and innovate beyond what they've done. These were kind of prerequisites that we knew we had to accomplish before we could go up market. The good news is we've checked all these boxes, or in this case, circles. We've checked all those, we're there, and we're executing flawlessly in that area. Very excited about that. Let's talk about the adoption phase.
Once somebody signs on and says, "We want to do business with Five9, and we love what you guys are doing," we do things a little differently. When I first came to the company, a lot of businesses were saying, "Oh, it's in the cloud. There's nothing to install. You don't need to go on site." I insisted, and Barry knows this very well, insisted for years, we're going to go on site for every customer over 50 seats. It's going to cost us more upfront, and we know that. It's going to affect our margins on the PS side, we know that. What it's going to do is it's going to set us up for success. That's due to having been in this industry for almost, well, a little over 30 years now, knowing that enterprises expect that, but they also see the value in that.
They want to know that you're going to come in, work with their business leaders, understand what they're trying to accomplish and the experience that they're delivering to their customers. That's so important to understand that and know how they want to deliver it. No two companies want to deliver that experience the same way. Everybody has their individual nuances. So we've got to be in there working with them, co-designing and developing and really setting them up with the right design and workflows, because those are so unique in every environment.
Our 100-plus professionals that do this on a regular basis have a proven methodology, and they get certified at Five9 to be able to deliver on that promise and be able to go through that six-step process where they're doing discovery, they're doing joint design, they're there for UAT testing, go live, the launch, and then they optimize the system to see what's working and what's not. That's a very unique hands-on approach that Gartner's given us credit for that not all of our competitors are doing. It also allows us to do back-office integrations and customizations with a lot of their systems that are unique so that the agents can be empowered to deliver a better experience.
Anand talked about a great new tool, the platform, the Whendu platform, that's going to allow us to do that much quicker and more efficiently, both ourselves, as well as our partners that do take on services for us. Okay. Let's talk about expanding. One way is we come into an enterprise, especially when we go up to larger enterprises that have, in some cases, they'll land with us in one department or one location. What we do is we make them successful, then we oftentimes will use that success to reference sell into other departments or just take organic growth within that one department and take us up to provide additional seats. We do project-based optimization reviews, these are key because we can bring professional services experts into a business and do an audit.
A system, if you kind of turn your back, it can get stale. It may have been set up initially and designed very specifically for what they needed. Over time, as their business changes, if you don't change the platform to adapt to that change, then you may lose some of that efficiency. We come in and do projects for our customers, and oftentimes the result of those projects are we see new applications, we see new partner and ISV applications that we can deliver to them, and we see new services that we can deliver. It's very important for us to, again, have that engagement with our customers on an ongoing basis. The retention. It really comes down to any SaaS business, right? It's about expanding and retaining those customers.
Every one of our customers has a CSM, a customer support manager, customer success manager. That's great. Most competitors and everyone else will have those as well. What's unique here is that we also offer, as an opt-in service, the ability for them to purchase a technical account manager through our premium support offering. That technical account manager is an expert who can log into the system and assist with whatever the customer's trying to do. These TAMs work with a handful of accounts. They know them on a first-name basis, they talk to them on a weekly basis, and they're helping them build custom reports, analyze certain data elements, understand why things are happening within the IVR, and what they can do to fine-tune and optimize that on an ongoing basis.
Very powerful in the retention strategy to have those individuals responsible for taking care of that account. That's really looking at it, and what does this result in? Anand mentioned earlier our net promoter scores. Nearly double that of the industry benchmarks and the other SaaS companies. I think this is done through our effort and through our people, and when we hire individuals into the organization, we're looking for people that stand for our company values. Not only have some expertise and some knowledge within contact center, but really we look for people that are customer first. They're going to be looking towards, what do I want to do to help this customer be successful? That has to be genuine. That has to come from within. That's not something that we just want them to do their job and move on.
It's about helping customers succeed in accomplishing their business goals, and that's extremely rewarding. If it's not rewarding, it's probably not the right person for our business. That's very key. The other thing that it's done is it's created this, if you take nothing else away from today from my presentation, take this away. It's the powerful slide that if you look back at our IPO five and a half years ago, we had three customers that were generating over $1 million or providing Five9 with over $1 million annually. Okay, three. Okay. Today, we have 49 that are $1 million ARR and up. You see a tremendous amount of adoption suddenly within large enterprise that we didn't see before. Again, it was because those prerequisites and those hurdles that we had to overcome in order to gain the trust.
We now have that. We've gone from three to 49 customers in the million-dollar ARR bucket, and it's gone from $6 million of ARR to $88 million of ARR. Now nearly 25% or roughly 25% of our revenue coming from this segment. The largest growing segment and also the segment that we're focusing the most on when I talk about our go-to-market model. Let's talk about that. If you look at the go-to-market engine, we're looking at how do we accelerate to capitalize and take advantage of the opportunity that's in front of us. As Rowan mentioned, the percentage penetration in large enterprise is single digits. Think about the opportunity and the TAM that's available to us just for this. Going upmarket is key. How do we get to them? We know what to do with them once we land them.
We've got a proven recipe for success with our customers. Okay? How do we get involved in more opportunities and get involved in making sure that we can capitalize on all the business that's going on out there in the world? It's done a variety of ways. One is we continue to scale our go-to-market direct sales organization. We talk about that a lot on the calls that we're on with all of you about, oh, gee, every year we're growing and expanding in correlation with our growth rate, our direct sales organization. Well, 2019, we took the opportunity to make some very key investments, we decided to invest into areas that would allow us to broaden our reach and increase our market share. What those key areas are is really expanding the technology ecosystem is really big.
We're now seeing. We used to knock on doors and say, "Hello, we can enhance to." We might have a video application, or we might have a scripting application or something unique, and we'd knock on doors of companies that would want to add to it. Now they're knocking on our doors. In fact, they're lined up ready to go. They want to work with our customers and add their solutions to what we do. We also looked at deepening and investing in strengthening our partnerships with the global SIs. Okay? We'll talk about that and the ones that we're replicating our work that we've done with Deloitte over the years and replicating that across others. We'll talk about scaling the channel in general, and how we're going to really take advantage of that worldwide, and increasing our international footprint and expanding internationally.
That's something that we've been cautious to do because of the LTV, the CAC rate, returned on enterprise business here in the U.S. is much higher and much greater than you can achieve in going into a brand-new market, because you got to get started, and it's very costly. First, let's talk about our go-to-market engine. You've all heard us talk about our direct sales being two buckets. We have our inside team or our commercial team that works really at 50 seats and below, and then we talked about our enterprise team. That's great. Well, enterprise was great when it was 50 to 200, 300, 500, but now it's well over 1,000 for many companies. Well, that's a different sales motion, right?
The individual that can go out and sell to a 100-seat contact center requires a very different skill set, and that's more transactional compared to the 1,000-plus seat global enterprise. Within the enterprise field team, we've created a strategic accounts organization that we're building out right now to focus exclusively on those large global companies, because we know that takes a different skill set, and we want to make sure we have experienced individuals. One of the things we've noticed this year is people can go upmarket and get big eyes and go, "Hey, I want to work on 1,000-seat deal," and neglect 100-seat deal. We've got to create swim lanes and keep people in their business segment that they're successful with and that they have the skill set for. Okay?
That goes with not only pre-sales, but in the post-sales model and the ongoing support model. Very important. Now that we're hitting on all cylinders in all those areas, we're looking at how do we have to be nimble and work with different constituencies. If you think about this for a moment, the stakeholders that are making decisions within contact center aren't just turning to a software vendor like us necessarily, okay. They're not just turning to a VAR like they may have traditionally, but they're also talking to consultants. They're talking to SIs to help them come in and build out their digital transformation strategy, and then work with companies like us to help implement and deploy that and be successful with it.
The first thing we've done in 2019 is really taken on new management, as Rowan referred to earlier, that had been there, done that. I got here. We grew the business from $20 million- $300 million. Now it's important that we have leadership that can help take us from $300 million to a $1 billion plus. The individuals that we're bringing in have all had the experience of what it takes, and they've run businesses that have over $1 billion of revenue in this space. That's very important that we have that. Brought in Andy Dignan a little over a year ago. He's running our global channels and services business. He was able to help us attract Jake Butterbaugh, who ran the go-to-market channel business for Cisco, another Cisco person.
He really had that whole route to market for the entire collaboration business unit in North America. The key there is between him and when I talk about Brian Atkinson, who is our VP of international, based in London, okay? He took Cisco's contact center business in four years from zero to $150 million, almost exclusively through channels. Between Jake and Brian, the experience and the relationships that they had, in many cases, decades-long trusted relationships with, have allowed us to open those doors into new channels and open those doors very quickly because the channels have worked with them for many years. The channels are starting to realize that their customers are asking for cloud, their partners are asking for cloud, and all of a sudden, rather than us knocking on channels' doors saying, "Hey, will you carry Five9?
Will you go put us in at your customer?" They say, "Yeah, we'll sign up. We might need it if the customer insists on it," they put it in their back pocket. That's how it's been in years past. We've been primarily a direct organization. Okay? What they bring to us now is these channels that are leaning in. The real sign of a channel leaning in and wanting to lead with cloud, they start investing their own dollars and putting skin in the game and sending their people to get certified on our platform. That's what's happening today. We have a full room every week with people coming in and getting certified and trained on our platform, not only for pre-sales but for implementation and support. Very, very important there.
Andy Zazzera came to us from Genesys, spent 17 years there running their PS organization. The last four years running global PS, the 13 years before that running PS for the Americas. Again, having the experience of a proven leadership team that's been there, done that, sets us up for great success as we move forward. For them, it was stepping down to the $300 million company and helping lift us. That's great. We're looking forward to that. Expanding technology ecosystem, this is continuing with the App Marketplace being announced and launched yesterday, as was referred to earlier. We not only have our CRM partners, vast array of partners, but also an immediate adjacency that's oftentimes bundled with our solution is the WFO piece.
That includes the workforce management, which is the scheduling and forecasting software that you need to figure out how many agents you're going to need to have online at 3:00 P.M. next Tuesday, as well as the QM and speech analytics and some of the more automated solutions that are so important to us. We've partnered very well with CSI and Verint and others in that space. If you look at the ISV partners, that community just continues to grow, especially as we publish easy-to-use APIs and SDKs for them to write to and be able to integrate their products to ours. We have a certification process, and we charge for them to have an integration with us, which is a great program.
We have our UC partners, as mentioned before, Zoom and Microsoft Teams, as well as others, for those customers who want to integrate to their back-office systems and have seamless integration between front office, back office. Okay? The deepening partnerships with SIs. Most of you have heard me talk, if you've been on our calls together, about Deloitte and what we've done with Deloitte for about four years now and really learning how to help them with really their Salesforce practice. They needed to have a contact center or CCaaS offer with that and bundled with it. If you go to any Deloitte Digital customer experience center around the world and you say contact center, they're going to demo Salesforce with Five9 for you. They've even OEM'd some of our solutions for things like healthcare with a patient connect offering that they have.
They're leaning in, and they've invested, and now we're turning to Accenture, EY, and several others, Slalom, IBM, and replicating what we did with Deloitte in those environments, and we're seeing great benefits from that already. The time is now where that pendulum has swung over, and they're being called on to help customers make that transition. The traditional channels, just looking at resellers, referral partners, and master agents. Again, a short time ago, they represented in 2015 only 9% of our bookings, overall bookings. Today, it's over a third of our bookings. Again, we're getting great traction, great mileage, and great pull out of these folks. On the international piece, we're growing our footprint. We've had great focus here in North America, in the U.S. and Canada. We've got a small team for Latin America.
As you know, we built out our data centers in Western Europe to handle Western Europe and the Nordics, okay? We're continuing to expand the capabilities, build in multilingual localization of the product, provide things such as local currencies as they're needed as we move forward. There's investments to be made there to deliver in those countries, but it gives us the ability to expand our global footprint much more effectively. We believe you can go into some of those markets prematurely and end up spending a whole lot more. We believe, and we're feeling the case that the timing is perfect right now to be in those markets and to expand.
That really gives you a feel for the customer. We believe we've created an environment that can scale and bring customers on board. Our formula or recipe for success is something that we've worked through the years to refine. We believe it's the best in the industry. It's reflected in our dollar-based retention rates. We also believe we have a go-to-market ability to reach and get to more opportunities through our partner ecosystem and our investments we're making in those different routes to market. With that, I'm going to pause. We're going to take a 15-minute intermission. When we get back, we're going to have our distinguished panel members. Ed, you met earlier, and we have two other customers that'll be up here.
I'll ask them a few questions, then we'll open it up to all of you to ask them questions as well. Thank you. For those of you who are tuned in remotely, this is Dan Burkland. I'm going to continue now with a customer panel. We've got three distinguished customers of ours that have been kind enough to come spend this morning with us. Before I bring them up on stage, by way of introduction, I wanted to make sure that you all had an opportunity to understand a little bit more about their backgrounds and their businesses, and then we'll bring them on stage. The first guest who's going to come up, and I'll do the introductions and then bring all three of you up at once so you're not sitting in front of the video.
You met Ed Albrektsen earlier from McKesson, and they've been a long-time customer and grown up to several thousand seats on our platform and replaced a lot of those legacy platforms. Once I ask a few questions, we'll open it up for all of you to ask questions of them as well and ask them about their journey and experience with Five9. One thing we want to do is I'm going to play a short video that's going to introduce you to Gina Wiley from DoorDash, and then also to introduce via video, Teddy Liaw, CEO of NexRep. Without further ado, please fix your eyes on the video. Which I have to click one more time. There we go.
DoorDash is a logistics technology platform that allows customers to download our mobile app or go online and order from their favorite local merchants. We're really passionate about allowing our customers to have more eatable moments and moments at home with their family so that they don't have to waste time cooking or cleaning. They can spend more time on doing the things that they love. What we're seeing more and more is that customers really want to be able to choose the platform and the way that they engage with any support team. They don't really have time to chat in or email in. They really want to be able to call in. The contact center is really at the heart of a customer experience.
There are always things that might be out of anyone's control, such as weather or traffic, things that might get into the way of a perfect delivery. Really, the contact center is there to help a customer, a Dasher, or a merchant feel like DoorDash is there for them at every step of the way and resolve the issue as fast as possible. With the previous vendor, stability was really a big issue, especially when it's lunch peak time or dinner peak time. We switched to Five9, we were really able to get that stability and have just that assurance that we would be able to develop and accommodate the growth that we're seeing in the delivery food space. We needed to implement Five9 within a matter of 30 days.
Five9 was able to accommodate our needs and our changes and hit the deadline and have a very smooth cut-over process so that our agents didn't feel impacted, that our customers, Dashers, and merchants didn't feel impacted. At DoorDash, we are really all about listening to our customers and hearing their feedback. In working with Five9, we've really been able to see that exact same interaction. Whenever I have any question or any feedback, they make sure they bring in all the right people to address that feedback, which really is a testament to the great culture and alignment that we have as companies. In working with Five9, we're helping our agents help our customers. We're helping our sales individuals make our merchants more money, and we're helping our internal teams get the data that they need to do their everyday jobs.
One more.
NexRep is not a technology company, we're a workforce company. The NexRep contact center is different from other contact centers in that 100% of our agents are virtualized at home. What we are able to do is take people that are stay-at-home moms or people that are veterans or disabled folks, people that are working in all parts of America, and we're able to route contacts, whether it be voice, emails, chats, or SMS, or even social media, through the cloud directly to the homes of our agents. NexRep's culture is all about service. We differentiate ourselves by looking at how we treat our people. They're people. They're human beings. They're not just a number. We're here to help them do a great job. In keeping them happy, we know that ultimately we're going to deliver better customer experiences to the consumers that we're serving.
Everything is consumer-driven, and where consumers want to go is where technology and services and solutions have to go. What consumers want is they want to be able to access a plethora of information at their fingertips. They want to sometimes talk to an agent. Sometimes they don't want to talk. They want to have that choice. Oftentimes, with more challenging interactions, they do want a human being, somebody who actually understands what they want and can deliver what they need. The world of contact centers is absolutely at an inflection point. It's more exciting than ever before. Having vendors that play well together is incredibly important to the future. Beyond that, everything is moving towards a data-driven world. We, as providers, need to be able to understand our consumers. We need to understand the agents.
We need to make sure that the right agents are interacting with the right consumers. We need to make sure that consumer dynamics are being understood. I'm so excited about the future because finally, at a point in history, we've got the explosion of the gig economy. We've got 50% of millennials that are already living in the gig economy. When we talk about the future of the work, well, it's actually here, it's the present. We sleep well knowing that we've got Five9 to power our technology to be there when we need, to make sure that as we embark on doing things that have never been done before, that Five9's there lockstep with us to give us everything we need. We see them as partners, not just vendors.
Excellent. Please welcome to the stage, Ed Albrektsen, Gina Wiley, and Teddy Liaw. Thank you. Thanks, Teddy. Thank you .
Thanks.
I'll get us started with a couple of questions. Thank you again for doing those videos. Those videos are on our website if you want to see them again or refer folks to them. Thank you for doing that. Thank you, Ed, for your time in here with a proof of concept for our AI technologies in addition to all the rest of the stuff we've done over the years. I thought it'd be a great opportunity, perhaps I'll start with you, Gina, to spend a moment and talk You mentioned eatable moments in your video. Tell us a little bit more about how Five9 helps your different constituents, both the Dashers, the consumers, and your merchants all working together and what you've done over the time we've been together to help scale that business.
Yeah, sure. Hi, everyone. Thanks for having me here. For DoorDash specifically, we see a lot of peaks and valleys in our volume that we get. Lunch is just extreme high peak times. Dinner is also an extreme high peak time. What we really need is a customer contact center that enables us to grow and flex along with that. Especially with food delivery, when you have your order outstanding, if there's an issue, if you have a question, if there's a problem, time is really of the essence in that. The nature of people and what we're seeing is that a lot of people do want to still call in.
They do want to reach a representative right away because when it comes to getting your food in 30 minutes, 40 minutes or less, getting it hot and making sure it's the right order, it's really important, and a lot of people do rely on this meal to help serve the rest of their day or their kids or family and make sure that they are connecting with the ones they love most.
I remember when you first reached out to us, you were on a different cloud platform that was serving much smaller companies, and you felt you had outgrown it. The call was, "Hey, Five9, can you help us? How quickly can you get us up and running?" You referenced in your video the 30 days online. At that time, we had about 600, 650 agents that we deployed immediately. I know your business is growing. Where are you at today with that scale?
Yeah, we have been growing. Just in general, the space has been growing so much, and our volume has grown. We've grown to more cities. We're now international. For our particular use case, we were having a lot of stability issues before. These stability issues, they would come during the peak times. They come on the weekends. They come during nighttime. When you're someone like me, you don't want to be on call at all times over the weekend, every single moment of the day. Ensuring that stability was up and running at all times was really important to us. Ensuring that there was also a platform that allowed our agents to connect with our Dashers, customers, and merchants in an empathetic way was really important. We've been able to handle all of those peak concurrent calls.
Five9 has really been able to flex and grow with our scale and help define some of the roadmap items that do help customers like us.
Great. Thank you, Gina. Teddy, you've taken an industry that really is used to filling rooms somewhat like this, large office buildings with hundreds of agents sitting with PCs and headsets and working at that. You've taken that approach and kind of turned it on its head and really taken a very unique approach to delivering customer service for your clients. Tell us more about that and how you've done that.
Yeah, super exciting for us because we believe that over the last several years, there was this movement that was all cost-driven, right? Everybody's moving to India and the Philippines and offshore or nearshore because of cost. There was this evolution of, well, the quality is terrible. Okay, people don't want to talk, especially American consumers do not want to talk to people overseas. They tried nearshore and it's like, the quality is still not there. It's a little bit better. How do we actually find a solution where you can get high agent performance and a marriage of cost? We have found through our model that we're able to deliver that.
We can have an entire ecosystem of contact center agents where we don't have to buy land or lease land or erect facilities and drop wiring, cabling and do all these old-fashioned things. We can have thousands of agents from their home, and we can empower people. The other secondary benefits of that are you have higher loyalty, you have lower attrition, you have higher commitment, and all that results in better performance. Now consumers are getting what they want, which is they want a domestic agent, they want performance, they want someone who understands their culture. You have cost savings passed to the companies and to the brands, and then you have ultimately agents that are just thriving in ways that otherwise would not be possible without technology like Five9.
Right. One thing we do with Teddy's organization because he has an elastic workforce, agents can log in and take calls when they want or when they're asked to. They can flex up and down very easily and pay their agents by the hour. One thing Teddy contacted us, I think it was about four years ago to help with something that's very unique, and that is and for a great cause. The Stand Up To Cancer Telethon is a one evening event. It raises $125 million last year. We have a scalable, flexible, elastic platform that can scale up to thousands of ports or thousands of lines, if you will, in a very quick period of time. He can bring on his talent to volunteer for that one evening to help raise money for a great cause.
For the last four years in a row, we've been the technology, and he's been the people behind that Stand Up To Cancer Telethon that you see on TV every year. It's a great way for us to help them raise such a great amount of money and do it with something that can flex up in one day and flex down the next, and is really a great model for what we can do together.
Yeah.
Gina, you're in the thousands of agents at your peaks. You've done thousands of agents both for the telethon and for your clients. Ed, coming to you with McKesson, I think we started with 40 seats back in 2013 with McKesson. Again, talk about land and expand. We came in with a small operation that needed a specific application to do some outbound dialing, and the next thing we knew it was, "Hey, what else can you do for us?" We grew it from there, and you're up to near 5,000 seats with us today, right?
Correct, yeah.
Yeah, great.
We peak out during our seasonal time at about 5,000 seats.
Okay
Possibly pushing a few more this seasonal.
Share that journey, if you could, with your legacy platforms and specialty health, part of life sciences, and then how that's blossomed into other areas. If you can share that.
Yeah, sure. I think, in the case that you had put together here, where you're looking at these large enterprise operations that have premise-based equipment. That was our previous situation, and we really had actually more than one platform. We were on Aspect and we had Avaya, and we had Cisco. We were using all three of those platforms, and mostly because the Aspect hardware piece of it was just brutally expensive, and the licensing on that is a fixed license on there, and we had some seasonality to our business. Keep in mind also that we were somewhat of a dynamic business with these pharmaceutical programs, and there's constantly new drug programs being brought in. The FDA mandates changes to existing programs. One of the issues that we saw was our speed that we could deliver those changes.
In those cases where the FDA mandated that we make these changes, it became an urgent situation with our telecom IT groups. So, in looking at the overall situation, we decided first that the cloud was the right place to move to and then did an evaluation. You're right. Actually it was 20 seats that we did, and it was strictly an outbound program that we launched on there. Then the people and the service and how easy it was for us to work on that platform, really opened our eyes up to bigger opportunities on it, and then expanding and growing from there. Over the course of the last five, six years on there, we've, of course, acquired multiple companies, and we've been rolling them in to these platforms and just growing and expanding the business.
Great. Excellent. Now I'd like to open it up to the audience to ask questions. We've got several microphones that are floating around. For the remote folks, it's important to use the microphone so they can hear your question as well. Thank you. Yes.
Yeah. Thank you for giving us your insight and perspective. I guess a two-part question. Just as you look into 2020, what kind of expansion do you see with Five9? Secondly, with the acquisition they announced earlier today, is that applicable to your business? Thank you.
Yeah, I can go first.
Sure.
Yeah, I think our space is growing. For me, as a business applications person, it's really important, someone was talking to me earlier, is that we're able to find vendors that can grow alongside us. Definitely in 2020, I think growth for DoorDash specifically is in the books as we expand to more markets. Secondly, as far as the acquisition, today was the first time that I saw any of those flows. Really, I started out as the first and only administrator of the system, along with all of our other applications, so I wasn't really a call center person. I've never done that before, with the flows of being able to build all of those things, Five9 does already have a flow builder.
Seeing that one that's able to integrate with other applications such as Salesforce and things like that is really important for us. Even now, if you ask me to build a phone line for a new partnership, I'm able to do it probably in five minutes, which is really great. I think that flow will be really important because what's really becoming increasingly important in the industry is that all of our applications talk to each other. I'm really excited for it.
Great.
Yeah. We're seeing a tremendous growth. I think from a macro level as well as I'll go into the macro specifically next up. From a macro level, what we've observed is there's this mass migration in the BPO space moving back onshore, number 1. Number 2, we see a mass migration of moving at home. A lot of publicly traded companies, a lot of consumer electronics companies, including the largest fruit company in the world, has made a migration, you know who I'm talking about, has made a migration to U.S.-based at home. Number 1, that's great. Number 2, for us, our business is growing tremendously. We have had tremendous growth in the last several years. 100% of that is with the cloud and through Five9 and virtual solutions because we have no facility.
I think the other thing that's a little bit different from us versus these guys is that we're a BPO. We service an array of different clients that all have their own technologies, their own solutions. The ability to be nimble and fast and integrate with my clients and then their various numerous amounts of technologies is super important because I need a client to come in, and I need to be able to turn them up because it helps me, it helps you, it helps them.
Right. Yeah, your question, they've just seen this today for the first time as well, obviously.
That's true.
it's confidential. We're really excited to come out, Teddy, to your point, and share with you how we can help you integrate to your clients' back-end systems much more easily.
I can't speak to any growth that we have or any acquisitions. However, I can tell you that what we're really looking at is to optimize what we have today, and this is really focused in on the AI that you saw today. How can we make our agents more efficient? How can we provide higher quality services that, again, it's that customer experience. In our case, it's the patients, the patients that are calling in and sometimes doctors and pharmacists and healthcare professionals in general. How can we make that better? My focus into 2020 would be more the optimization and getting better and optimizing WFO and how can we work, specifically the AI is what we're really excited about.
Next question. Yeah, Scott.
Hi, good morning. Scott Berg with Needham & Company. The question is on scalability. If you look at Gartner's recent Magic Quadrant, the knock, if you want to call it that, on Five9, is its ability to scale relative to the other cloud vendor or well-known cloud vendor in the space. Can you all talk about your experience, because you're with larger, obviously, organizations, on the platform's ability to scale and then maybe compare that against your prior technology experiences?
Yeah. Let me start off with that because that was one of the primary reasons why we moved to the cloud is because of that and as I'd mentioned before, having premise-based hardware. We typically, in our seasonal, will add anywhere from 1,500 to 2,000 agents, and we normally launch that in the month of December, and we run that through March. We have the ability of scaling that up in the various domains, building those campaigns. Then, as the season starts to roll down, we start to roll agents off and again, it's subscription-based. It saves us millions of dollars by the ability to be able to add and delete licenses.
Yeah, I would just to say, to add on to that, it's the same thing for us as far as the peaks and valleys of what we experience in any day. As we have promotions or things where there is more interest in something, I think for us, the ability to just serve all those customers and add agents or use up the licenses that we only use is really important.
The story that Dan alluded to illustrates that perfectly. If you don't know what Stand Up To Cancer is, it is a telethon where 50-plus TV stations all call truce. For one night, they completely stop their programming, and then they show the same exact telethon to raise money for cancer research. We're talking about a TV show that Bradley Cooper hosted last time, and at 7:59, there are literally zero phone calls. The 800 number is not public. Nobody has anything. At 8:00 o'clock, the television goes on. There's music. Bradley Cooper goes on, and by 8:02, there's 5,000 ports completely full. To go from zero to 5,000 ports completely full is absolutely absurd, and I don't think it's really anything that happens anywhere else.
Anybody who's watching TV that night is probably watching that show other than ESPN because they're jerks, and they care more about showing their programming. Every other show, every other channel is broadcasting this. It's super powerful and that this only happens with Five9. I will say an interesting thing. Sir Ken Robinson, who is the single most viewed TED Talk person ever in the history of the platform, and he talks about these things called unintended consequences. He talks about Galileo and especially in technology where all he wanted was a telescope so he could observe the moon. It turned into humankind completely debunking what we think about in terms of our role in our solar system, in the entire universe, and we end up realizing that we're not the middle of everything.
We talk about Steve Jobs in 2007 when he made the iPhone and 2010, the iPad. There's no way that he knew that there would be apps like Uber or DoorDash sitting on top of that platform. He sure as heck didn't know that people would be spending eight hours a day on the tablet and the phone. There's no way. Unintended consequences. When Five9 came out, I think they looked to tackle the contact center space. I don't think they realized that they would be a part of raising $125 million in one night for cancer research. I don't think that was the intention, but it sure as heck was the unintended consequence. I think moving forward and looking to the future technology, especially with how we're doing things, it's so exciting with Five9 because we're going to unleash business practices that have never been fathomable before.
What I mean by that is for an organization, nonprofit like Stand Up To Cancer say, "You know what? In one night, I want to have thousands of agents." That wasn't possible before. I think businesses are going to change their behavior. They're going to be able to do promotions. Literally 40% of our country or our population is going to be affected by snowstorms in November. Florida's getting snow? Like WTF. This is crazy stuff. There's so many things out there and how businesses are allowed to flex their behaviors and interactions with customers is going to be so transformative beyond just these slides up here and the cloud powers that.
Wow.
I could go on forever about technology, Dan.
I was going to say. I thought only I had energy and passion about this, right? Go on. Next question. Yeah.
Just curious. Hey, you're obviously sharing some super insights here about Five9. There's other cloud peers out there who say they can scale, they can migrate you to the cloud, obviously. What do you say makes Five9 stand out from its cloud peers? You've kind of touched on it, kind of integration with Salesforce and obviously ability to scale. Again, other people say they can integrate with Salesforce and scale too. Just curious in your opinion.
Yeah. I'll start off with that one there, and I'll let you guys add onto it on there. I would say, from the very beginning and just to preface what you had said here, we actually did a very formal RFP, we looked at the entire landscape, the big three that you see in the Magic Quadrant there are ones that we looked at. We found something very unique about Five9 that we did not find with any other supplier that was out there. Basically, it's their culture, and it's their ability to listen. From our perspective, we really needed something that was custom-built and unique, they made us feel that way.
I know the platform is adaptable and there's many other uses of it throughout the Five9 enterprise today. I think that each customer, and you can poll them all individually, probably feels that, "Well, Five9 built this just for me. This is mine." That's the way I feel. I think that if you were to look at how we use the product and how we interact with the people, and in our particular case, we actually have on-sites, badged people from Five9 that are on our site that help us administer the system, help us optimize the system. They know us. They know our business practices. It's a very special situation that we have and a unique relationship that is very personal to me, and I would say not just Ed, but the people that actually run the contact center.
Yeah, for me, I'm definitely a technology agnostic person. I'm a DoorDash person. I'm always looking for what's the best tool that we can use. I've seen a lot of the demos even, continuously evaluating all of these things, and just based off of the criteria that we have as far as requirements, Five9 is really the only one that's been able to satisfy every single one of those requirements. If there is a customer pain point that I bring to them, there is immediate action on that or a game plan that's set out, "Hey, these are some resources.
This is what we can do. We do use the technical account manager, that has been extremely useful for our entire team to have someone who knows the system so intimately that they're able to give us recommendations and be there at any time of the day. I even text my TAM. I text my account manager when I have questions or if things are going off. It's been really valuable to have a human to talk to and really have the partnership there too.
To Rowan's dismay, because I am a BPO, I am forced sometimes to use other vendors. Oftentimes, I wish I could port all my business over, and Dan would be happy, and Rowan would be happy, and Barry surely would be happy. I can't. I'm stuck using a plethora of different platforms out there. You would think that all these cloud platforms are the same and they're just delivering a cloud-based system, and they're not. When they display the full suite of services that they have, it's super important to understand the layers of this. At a bare basic level, they've got to have reliability and stability, and a lot of these other guys just don't have the same level of reliability and stability. Fact.
You could go ask everybody about what their uptime is, and that's a really important stat that they put up earlier. The second thing is the ability to have everything in a one-stop shop for the agents, have reliable reporting. Reporting is so bad with so many other places, to be able to export out data, be able to pull in data, especially when you're working and interfacing with so many other platforms. The play nicely comment that Dan talked about earlier, is so important, and it goes to playing well nicely with other vendors. Really with us, I think what keeps us coming back when there are issues, and there sometimes are, is really their commitment to client success and services. The TAM program is completely defining and different. I hope that they continue to keep it, as it's been historically.
The people that they have that work on our setup to making sure that we're using it properly, we're deploying it fully, and to making sure that it's reliable is just unparalleled out there.
Yeah. Meta.
Just a quick question. Are you primarily using it for telephony-based I know you guys all gave telephony-based agent answers, but telephony-based agents, and then are you using them in omni-channel, where you would be able to see if somebody emailed and called in as well?
That would be yes for McKesson. It's very unique to the actual pharmaceutical program, though.
We use Salesforce as we're a big Salesforce shop. We use Salesforce's omni-channel. Five9 plugs into that.
Yes, as well. We actually use them, and they don't know. We use this for time tracking also. Sometimes we have really crappy other solutions that we have to use, so we actually use them for payroll and, because we rely on them, so we actually use it for virtual time tracking.
Okay.
Thanks. In the back. D.J. Hynes from Canaccord. Maybe for Ed, since you seem to be furthest along with the AI strategy, I'm curious how you see AI evolving in your contact center. We heard Five9 talk about today that they're going down a path of agent assist versus IVA. How do you see that playing out? We also heard stats that 16% of seats could be automated away over the next 10 years. How do you see that playing out in your contact center, and then what do you think are the right mechanisms to pay for that type of technology?
Yeah, that's a great question because it's more than just the agent assist on the front end because if you look at a call. If an agent without AI today makes a mistake on a call, when we go back to the QA, which I'm going to address here in just a minute for you, we basically have to put together, when our quality management system gets in and starts to look at it, and we QA the call, and we find that we missed a PAE, we've had to put together a campaign and it becomes a CAPA for us, and we have to log that CAPA and investigate, call the patient back, get more information.
If we can make that call experience and capture everything up front to make it an A call on the very beginning, there's really no need for us to do quality on the back end. Now think about that. If you know contact center and the way that our contact center works today is, believe it or not, we do four QAs per agent per month. That is an insignificant, it's a small sampling, but the cost to do a higher percent is just prohibitive and the customer won't pay for it. If we can QA that call while it's happening and make that experience perfect from the very beginning, this is where our cost saving comes in on the back end. We don't need to have 189 QA people in our contact center. We have that experience up front.
We capture all of the PAEs up front. We address that. The patient experience is taken care of on the first call. First call resolution, if you're in the contact center business, is a big deal. You've got to get it right the first time. AI moves us closer to that nirvana when it comes to that experience. Did that answer your question?
Yes.
One other point on that, I'll add to it, is we're really not looking at the contact center from a standpoint of reducing headcount. We're looking at really bringing on more business. That's the way that we're looking at it is from that perspective, is how much more business can we do, how much more efficient can we be, and then be a differentiator because we do have competitors out there. Thank you, Jonathan. We believe that the AI will make us that much better than our competitors.
One last very quick question, that Lisa indicated we have time for one more. I think somebody over here had one. Is that right? No? Okay. If you do have questions, they can take them afterwards privately. They'll be here for lunch as well. Thank you all very much for your business and for coming here today and sharing your experiences with our audience. This was wonderful. Again, thank you.
Thanks, Dan. Our pleasure, by the way.
Without further ado, the one, the only, Barry Zwarenstein. Take us through the numbers, Barry.
Thanks, Teddy. Thanks, Dan. Thank you.
On to it, Barry.
Good afternoon, everybody. Just want to add my welcome to what Rowan said earlier today, and just to say how heartening it is to see the growth in this audience over the last three years since we had our last Analyst Day. Thank you for your interest. We're going to start by talking about our financial commitments. Those of you who know us even a little bit know that this is a management team that takes its financial commitments very seriously. What are they? First of all, sustained 30s enterprise subscription revenue growth. Second of all, a continuity of the disciplined execution. Rowan, at the outset, talked about the fact that we're riding two waves, the re-platforming of America and the rest of the world to the cloud, and second of all, digital transformation with customer experience prominent in that.
Doing that with the frugality and the pragmatism that we customarily exhibit. Finally, doing it with a balance of the top and the bottom line. We still subscribe to this quaint notion that profits and cash flow matter, we strive for that balance. Taking a retrospective back to the second quarter 2014 we went public to now, we've maintained, in fact slightly improved our revenue growth from 27% to 28% LTM. We've had a massive expansion in the bottom line, 44 percentage points. Almost all of that is reflected in our operating cash flow, given our working capital efficiency. If Teddy will forgive me, with all due respect, this is an intended consequence. What's driving it? Well, it's the enterprise business. That enterprise business is now 80%, as we've talked about before. When we went public, it was 60% of the total.
It's immensely profitable. On a truncated five-year basis, the unit economics, the LTV to CAC are 6 to 1. That is something that I'm told even many people would be happy with a 3 to 1. It's a really solid business. This here is the entire enterprise business. Remember the 3 components, the subscription, the usage, and the professional services. Drilling down particularly into the subscription part of it, we've demonstrated consistent growth in the enterprise subscription in the 30s. It should be as no surprise when you're growing in the 30s and the corporation as a whole are growing in the 20s, eventually, and we've now reached that stage, the enterprise subscription is now more than half of our revenue. When we went public, it was 34% of the total.
Dan was very proud, justifiably, for the work that him and his team have done to take us from $3 million ARR customers up to 49. By the way, no customer concentration. All of our customers are important, but none of them account for more than 3% of our revenue in any given quarter. What we did is we looked at those 49 and did some additional cuts. We're talking about truly massive corporations. McKesson's prominent, but there are others. The average revenue, $10 billion, and the employees, 12,000. A question we frequently ask as a management team is, we understand that you get a lot of your revenue from new logos, majority of it. What about your install base? Well, for three reasons, that is growing in importance. First of all, the base is getting bigger.
Second of all, the metrics show that the amount that they can afford, the bigger customers, they buy the extra SKUs, is higher than the ones from the lower end of the scale. Thirdly, the MRR per seat is going up due to this new product team, Jonathan, Dave and Anand. They just inhabit a different universe from what we used to have, and they're being properly funded, and the additional SKUs and features are going to be showing up increasingly in our bottom line. What you're going to see now is to help you get an impression of how successful we've been in terms of expanding in the install base. On the left-hand side over here are those 49 customers that we're talking about. On the right-hand side is the CAGR in billing since they joined the platform.
Now you're going to see a sort of Minecraft-type situation where the green is the initial booking or subsequent booking, and here goes. Oh, no. There we go. Obviously you want to have a high persistent rate of green to white, which is what is amply demonstrated over here, and it just keeps on going and just keeps on going. When you average out all of those CAGRs, it turns out to be a very respectable 67%. We're growing very nicely in the install base. There's a disclaimer in the financial community that past results are not a good indication or proper commitment for future results, but they do have some relevance. We're very proud to show our Mona Lisa slide over here. Blue is the revenue growth by quarter. The red is the improvement.
You'll see that we went from a negative six under the rule of 40. We peaked in the fourth quarter of last year at 53, so an improvement of 59 percentage points. At that point, we decided, given the strength that we were seeing in the market, that we wanted to meaningfully start increasing our investment in both go-to-market and in R&D, as both Anand and Dan talked about. I'm extremely happy that we did that. We took a step down and now we're starting a new staircase, but from a much higher starting point. Operating cash flow. I alluded to this earlier on.
We've now had 15 consecutive quarters of positive operating cash flow driven by the EBITDA, driven by the fact that we only have 31 days of sales outstanding, and driven by the fact that with over $206 million in federal NOLs, we're not going to be paying any cash taxes anytime soon. The long-term model. The model you see over here is a model that we have in our current investor deck, we're reaffirming it at the moment. We have us going from the current 18% up to 27% over the next five years. Where we draw a lot of confidence is that the two biggest drivers for the upcoming improvement are ones that don't require a whole lot of creativity or new ideas or anything like that. It's a continuation of what's happened in the past, namely on gross margins and G&A.
In terms of gross margins, there are three reasons why our gross margins improve. The biggest one, though, which I'll talk about now, is a very simple fact that our subscription revenue grows in the 20s, a blend between the 30s that you saw for enterprise and the single digits for commercial, while the cost of that subscription revenue grows typically in the teens. It's slow, but very steady. In terms of the G&A, where most of the rest of it, the increase will come from, we've built this big fat pipe with the excellent staff and systems, and we can run a lot more through there without increasing that to any meaningful extent. With that, I'd like to turn it back over to Rowan to wrap up and to handle the Q&A.
Thank you, Barry. Thank you, Barry. In summary, just take it back to where we began. We have a huge market opportunity in front of us, which we see as expanding, given some of the technologies that you saw today and some of the interest that we've had from our customer base. We see this very large opportunity being consistent and also growing over time. We're confident in our execution engine, as we hope we demonstrated today with the new leadership that we've added and some of the plans and the progress we've already made over the last year. Finally, as Barry alluded to, we're going to continue to grow the business over the next decade and more, profitably, and continue to drive the growth as you've seen us do.
With that, I think it'd be great if we could just open it up to Q&A for my leadership team. I think we're going to bring a few stools onto the stage here. We won't bring everyone up because there's not enough room, but all of my leaders are in the room, you can ask questions of anyone or the group as you like.
Thank you, Chris. How much time do we have for Q&A? 15. Okay.
All right. We have mic runners in the audience again. Yes, open to your Q&A.
We're going to start with Sterling.
There you go.
Thanks. Sterling Auty with JPMorgan. Thanks for having us. Rowan, on the comments that you made in your prepared remarks, and also the Q&A came up during the customer panel about the shift that AI is bringing along. We think about chatbot support, conversational commerce, et cetera. Curious from your perspective, how you see the portfolio of products today being positioned to handle that? What's left to fill in, and how does that change the go-to-market motion with your customers moving forward?
Yeah, I'll start it and then maybe turn it to Anand or Jonathan, who can comment here. There's a bunch in the question that you just asked. There are several new technologies that have been coming about, like chatbots. On that specifically, just yesterday or the day before, soon, recently, we announced support and an integration with the Einstein chatbot technology, and that's an automated chatbot. It's purely, you're just talking to a robot or a computer that's trying to give you automated answers. What we announced in that new technology was the ability to seamlessly move from a human-to-computer conversation into a human-to-human conversation. We want to bring that human into the loop, which is where Five9 becomes required to come into that picture. Now, we're not off building our own chatbots.
We are very laser-focused on making human agents more efficient, and we think that's the best approach in terms of where our expertise is, because the one thing we have access to that no one else in the ecosystem has access to as a CCaaS vendor is the real-time voice. We're looking at that as a moat that we can continue to deepen and widen and leverage that real-time voice. Any things to add there?
Yeah. Again, this point of moat, there's several aspects of this that we feel also make the product capability highly differentiated. I spoke a little bit about AI and how we're leveraging tools from Google and what have you. You still have to provide them with training data. What products like this allow us to do is to take our recordings, our voice, turn them into training data, and over time, build up catalogs of this stuff that serves across customers. He who has the most customers, has the most recordings, has the most data, and that helps us drive greater and greater strength. On another piece of your question, too, I'll mention, and I'll turn it to Anand. On the go-to-market, interestingly, this is also a product that works really well with channel.
Like contact center in general, there's a lot of work to do to set up and implement and make it work for the customer. As you saw with this product we showed you, it was fairly customized for McKesson. You saw a UI that was custom-built for them. We see a great opportunity for channel partners to provide value add and to help us scale this up, and that's a really great combination as well to expand our business.
Yeah. I'll just add, too, in terms of capabilities. One, as Rowan mentioned, we're integrating with wherever customers want to start. You heard Gina talk about they made the decision to have one system. We need to play well with that system, so that's a key design principle for us. Even for the system of the application of record, if you will, if they have a CRM system, they could work with that. If they don't, we have our user applications, right? We have this design principle in production already with hundreds of our customers. We have the same design principle, which is we have an end-to-end system that will work if that's what the customer wants. If they've made choices, we'll work with that stack. The other piece I'll say is, we're also building the ability to Jonathan briefly touched on that in his demo.
We're building the ability to take the transcription and do a bunch of interesting stuff with it, which is a very important requirement that we're getting from customers, which is it could be I literally want the transcription in my system of record as the starting point, but it will also be the starting point for new ways of doing WFO use cases. Now that I have this system, with an omnichannel interaction between a client and an agent or a client and a bot, if you will, take that and still see if there's adherence, to still see QM on top of that, et cetera. We talked about Whendu, that allows for those integrations to happen as well.
A lot of the building blocks are there, but we still let the customer sort of decide what their configuration is, and you heard all the customers here talk about that earlier.
Thanks, Charlie. Yeah. Got one here.
I didn't introduce myself earlier. It is Terry Tillman from SunTrust Bank. Maybe the first question, Dan, for you, or the multipart question as I always do, is the strategic sales team, maybe some perspective on how big that team is. Also in one of the slides earlier, the commentary, I think, maybe it was from Barry, I may have forgotten, but the idea of new logo growth still very much fuels the growth, but there will be an increasing evolution around installed base. I am just curious on the go-to-market side, is that a more meaningful shift next year as it relates to harvesting more from the installed base? That was the second part. Thank you.
Yeah. Great questions, Terry. The first one, if you look at our strategic group, it's actually just taking our enterprise sales team as it exists today and just really dividing that and making sure that we keep people focused on the opportunities that they're most suited for and skilled for. Some folks want to do multiple transactions and really have the month-in, month-out singles and doubles, and there's folks that are really good at that and want to do that. There's others that are willing to go take a risk and work with the large opportunities that may be year-long sales cycles and may be much larger, but they're putting more eggs in fewer baskets, there's more risk and reward there. It really is up to the individual.
As we build out and scale, when we talk about scaling our sales teams 30%-40% year in, year out, that includes the channels and the SI focus. It also includes taking our enterprise team and adding to that. We're adding some experienced sales folks from outside, and we're making sure that we just allocate folks to stay in their swim lanes. It's not really this, "Oh, we're building it from scratch." We're just saying, take experienced people that have been successful selling those 49 accounts that you saw that are over $1 million. It's amazing. You look across, and you see who's selling into those accounts, and it's a handful of folks that are very experienced and have been there, done that. They're the first folks that are making up that team. Second part of your question. Boy, short-term memory loss.
Selling into the base.
Installed base relevance.
Oh, installed base relevance. Yeah. The installed base, yeah, as that base gets larger and larger, ever increasing to make sure that we're doing the things to not only retain them but expand them, grow our technology footprint. I say that technology footprint because it's one thing to just expand and get all of their seats. Once we've got all their seats, how do we also make sure that we retain them and expand them? The way you expand them in that case is new applications. The R&D investment that Barry referred to, that we're putting into R&D to build more applications, more SKUs, the acquisition of Whendu to be able to extend the platform to do a lot more.
We're looking at being able to take our ARPU and increase that within our existing base on a per-seat basis so that it doesn't require organic growth, it can require application growth. I'll add to that a little bit, which is looking at the sphere of the applications that we deliver today and expanding upon that. There's some immediate things and adjacencies that we resell that we can put on our paper to increase that, we're always looking to do that where we can. Okay.
Great. Scott.
Pardon me. Scott Berg again. Rowan, I guess this is a question on your TAM slide that you had up there earlier. You had the smaller circle down below that was a $58 billion TAM. That included that $24 billion of the TAM that we're all familiar with, but that other $34 billion was of tools, automation tools, et cetera.
Yeah.
You didn't explicitly say this, but I took your commentary around all your AI tools and some of the other product innovations really kind of focused on that part of the spend-
Yes
in the market.
Yes.
I guess the two-part question is, one, how much of your innovation efforts over maybe the next two to three years are really focused on that area versus the core market? Two, now that you have some alpha customers on at least some of the products, what does pricing start to look like there?
Yeah. I would say we don't know yet on the scale of the mix of the investment on sort of new versus existing. We'll have a better sense of that next year as we begin to commercialize. Jonathan's goal is to have this live and selling in the market by 2020. I think we'll get a sense for that. In terms of pricing, I think a big part of that depends on us talking to our customers. We don't know yet. We've had that question a bunch. It really depends on how much money we can help save these businesses and how much of that we can capture for ourselves. What I would tell you is there are products that are similar to this in the market today, some from startups.
There's little bits and pieces that are out there, so you can get a sense for how some of the pricing is, and it's sort of all over the map. Some of them do skin in the game pricing, how much they save you, that sort of You'll get a piece of it. Some of them do an agent-based model upwards of $100 per agent per month. I would say, look, if you do the math, $100 a month would be roughly 5% of your labor. You have to think about can you automate 5% of the use cases or make an agent 5% more efficient to be able to afford $100 of extra savings to be able to pay for software.
You wouldn't want to capture all of it probably, or I doubt businesses would give you all of it, but you should be able to get some portion of that. At a 5% level for the average business being about $100 per agent per month, that's an interesting number, we think. Jonathan. Oh, sorry. In the back.
Hey, guys. Mike Latimore, Northland Capital.
Hey, Mark. Hi, Mike.
On the Agent Assist, the early kind of data, what kind of savings are you seeing in terms of just agent savings? I guess the second question would be probably for Anand. As you look at Five9's position to the broader conversational commerce messaging market, how do you see Five9 kind of evolving into that space?
Yeah. On the sort of total time save, it's dramatically dependent on use case, right? We, I think, need a lot more scale and usage across even a larger base of customers to get a good handle on that, and I think it'll wildly vary. Too early to say is sort of the answer. Again, I think there's cause for optimism. If you look top-down, like, well, what fraction of the time are calls on hold? That's one piece that's a lot of time. Those are probably areas that we could be able to tap over time with this. I think at a top level, that one looks pretty good, but we don't have hard numbers.
The interesting thing about the contact center space in general is there's a tremendous amount of data out there already for us to mine. In addition to all the calls and the transcripts and so on, the dispositions of those calls and what were they about, and there's training data that's been built up. There's already a very heavy process focus in the contact center. The thesis here, anyway, is that we can leverage all of that to identify. I think if you go to most of our customers today, they'll tell you instantly what their top call driver is. What is the average handle time? What is possible if you have a trained agent with a year of experience? What are the use cases that they're working to automate or deflect away from agents? We're not coming at this cold.
This process has been worked for 20 years, or more like forever, essentially. What's new is access to these automatic triggers using AI and automation technologies that are sort of now just becoming at the scale where we can put them to work. Conversational.
On conversational, not just commerce, but conversational AI and interactions in general. I've worked with more customers than I'd care to admit about the early integrations of conversational interfaces. The two big things that we realized, which I'm so glad we're positioned so well to do a much better job on, is, I think customers don't want these swim lanes of interaction, is a big thing that we realized. They would start with Messenger because they're in a meeting and they can't take a call and they want to start some interaction, or it is an interaction that is the kind where I say, "What's my balance?" I get an answer back, and I'm happy, right? Maybe I'm not happy with something there, and I want to talk to someone, right?
That's when one channel stops scaling because it reaches its limits in terms of strengths and weaknesses. The fact that we're allowing this interaction across channels, I've been on multiple calls with customers where they'll say a customer will call them to say that they just emailed them, because they're so annoyed about something. They'll be even more annoyed that the agent doesn't know that they just emailed them, right? The multi-channel interactions really matter, far less kind of routing the incident and more kind of a customer view, which we talked about in terms of the omni-channel focus. The other piece is a lot of times when they reach out the second time, they're like, "I told you this the last time," right?
Don't you have the transcript from the last time I spoke with you?" To Jonathan's point, like, "Why are you asking me for this again? I told you last time what plan I'm on. I told you last time sort of what my preference is. Why are you asking me if I'd like a window seat or an aisle seat? I told you what my preference is last time." This ability to store a set of preferences for the customer in one back end, whether it's in the Five9 system or in their CRM system of record and bring it back out, I think those are two big things. The last thing I would say is it really has to interface with the other apps that Dan talked about. We talked about the supervisor example, the same thing with WFO.
If they're seeing if they're in adherence or their QM is going okay, that has to include all the channels, right? If 30% of their agents are on one channel, but WFO doesn't work for that or QM doesn't work for that, the supervisor interface doesn't work for that just won't work. The last thing I'll say is, for most, if not all of our large customers, all the data at rest has to be encrypted. We're reaching a place where the early experimentation on conversational is done, and we're down to the serious implementations now.
Jonathan, then [Meta].
Great event. Very informative. Thanks again for having it, and having us. Jonathan, thank you for some insights. You've talked about investing in enterprise. I want to ask about the flip side with commercial. In the past, you mentioned you're not going to invest in commercial. The LTV to CAC is not as strong there. I guess, are you at least looking to optimize the cost there, bring that down, improve that ratio? Are your expectations still for mid-single digits in terms of growth rate for that?
Yeah, sure. On the first part of the question, if we can improve that, we do have various programs in there. It takes a fundamental change in the market to be able to get that 2x versus a 6x up meaningfully. Why do I say that? I say that because the biggest single driver over there is the lower retention rate or higher churn, which is just endemic to that type of business. The more experimental business models, they go out of business, whatever, they don't have enough funding. In terms of the growth rate, the facts are well known. Last year, 2018, we just had a spectacular year under a new leadership with a whole new selling motion that Dan can elaborate on if you're interested. This year we are going against that with a difficult compare with single-digit type growth.
Frankly, I would not put into the model anything above single digit for commercial. It's just the nature of that business.
Great. If I can follow up, Barry, since you're answering this question. You have the long-term model. You have sales and marketing. You are already there in terms of the range.
Right.
You're actually on the low end. I guess with all the initiatives that you're launching, with channel and all that stuff, is it fair to say that there's some room for expansion in terms of the percentage of revenues? Thanks.
Absolutely. The things that Dan talked about, the channel, the SIs, the international, all come with a price tag. That's why we have it at that range, and you should see it drift up.
Okay, one last. We'll take the last question now. Yes.
Yeah. Thank you. Nandan Amladi from Guggenheim. In the long-term model, Barry, you've shown R&D at 8%-10%. Given how fast the space is evolving and the investments you're having to make in AI and all these new technologies, is that a reasonable investment level to stay ahead of the competition?
Yeah. There may be some rejigging over the years between the different categories. What's important to us is the 27%. We believe that that's something that we should strive for over time. It won't be linear. I do remind you though, that it's not always the biggest R&D budget that's going to do that. It's how focused and how disciplined you are. Secondly, we do have a portion of our revenue. If you look at our recurring revenue, which is 91% of the total, a little under 30% of that is usage, the resale of the minutes. Frankly, there's not a lot of R&D that's involved in that. If you normalize for that, you're somewhere in the mid-teens, which is not unreasonable.
Okay.
There's a healthy debate about that.
Thank you.
Thanks for that question. You can imagine, again, the thing to focus is the bottom line number at 27. There's some puts and takes in the middle, based on how we feel the business needs to be run, we would make those changes. Great. Thank you, Barry. I think that was our last question. Thanks to all the E-staff members and thanks to all of you. I think we're at the end of our agenda, I'd just like to wrap it up. Hopefully you can get out of here in case it starts snowing. We really do appreciate you spending the whole morning with us. If there are any follow-up questions, we will be here. I think we're going to spend lunch together.
We'll be spread around at all the tables, so find a place and definitely engage with us over the lunchtime period. Again, thanks all very much. Thank you.
Thank you.
Full speed warp 10.