Good afternoon, everybody. I'm Scott Devitt, Wedbush's consumer internet analyst. Very happy to have Isabelle Winkels, the CFO, and Chris Ferris, the head of IR for Braze. I think a good way to start, Isabelle, maybe give your background and give a little bit of a background on the company as well.
Sure. Thanks, Scott. Thanks for having us. I'm Isabelle Winkels, Chief Financial Officer here at Braze. I joined the company in January of 2020. A t the time, I was the company's first CFO, and at the time, the last member of the executive team added, really as the company eyed an IPO to come about, about two years later. W e IPO'd the company in November of 2021. Quick background on me, I actually did the bulk of my career at Morgan Stanley, so actually, Scott and I go way back.
I started life as an investment banker in the ever-so-sexy electric utilities sector. Ended up doing banking and then research, and then actually migrating over into the finance organization at Braze. S ort of figured out pretty early in my career that I wanted to be a CFO, spent a bunch of time working under, then Ruth Porat, who was CFO, of Morgan Stanley, starting in 2010.
A fter she left to go to Google, or Alphabet rather, worked under, John Pruzan. Left to go work under the CFO of Cognizant Technology Solutions to really kind see how all the pieces fit together. Did that in 2018 for about two years, and then in 2020, about six weeks to the day ahead of, the sort of global lockdowns in March of 2020.
I started at Braze, and I was super excited to, you know, be at an up-and-coming startup, really excited place to be, especially after doing what I did for Morgan Stanley, which was a lot of bankruptcy planning in the post-financial crisis regulatory environment, you know, six weeks after I started, we went into lockdowns and suddenly realized,
"Hey, it's possible our customers aren't gonna pay us," and there was all these sort of, you know, negative sentiments out there before things settled out, and I turned to our CEO, and I said, "You hired exactly the right CFO, because downside scenario planning is exactly what I know how to do." A nyway, it all turned out pretty great. Ended up IPO'ing the company, as I mentioned, in November of 2021.
I'm happy to give a little bit of background on Braze. Braze was founded by two technical co-founders, Jon Hyman and, Bill Magnuson, who are still currently with the company today as, Bill is CEO, and Jon is CTO. We were founded in 2011, with really the deep-seated conviction by both of these individuals that actually mobile was really going to change the way businesses were built, grown, and how businesses were going to engage with their customer base.
The advent of mobile really changed computing and digital customer engagement from kind of, you know, batch and blast sort of email capability to this always-on, this ability for brands to actually be in constant contact, and for individuals to engage with brands pretty much anywhere, anytime.
What it did is it created a host of net-new channels across which you could start to engage with users. It created the ability to engage with them in real time, in an event-driven way, in any point in time. I t created the ability for brands to actually become much better listeners and really understand, on an individual personal level, what each individual needed and wanted from every respective brand, and how to engage with them in the highest ROI type of way.
Braze was built with a couple foundational points that distinguish it from the existing competitor set. One is we are a fully integrated stack, which starts with a software development kit embedded straight into the digital native applications of our brand.
That might be a website, it might be an application, it might be an app on a smart TV, it might be a connected fitness device. W hat that does is that allows Braze to live in the moment. We like to say we experience the world as it happens, and that enables brands to hear in the moment how are individuals engaging directly and digitally with the offering that they have.
Number two, we're built on a stream processing architecture, a fully integrated stream processor that essentially mimics a high-frequency trading system, where information comes in. That would be the engagement by the individual with the brand. Information comes in, it is automatically classified. T here is a message that might go out. It's personalized and sort of orchestrated through the process, we are.
Last piece of key differentiator is we are omni-channel, and we were born omni-channel. A lot of our competitive set were largely born single channel. W e started life offering four channels, now offering upwards of a dozen channels. E nabling that, a sort of feedback loop that goes from an action taken by an individual through a digital interface that might trigger a message all the way down the stack, and then that message might trigger another reaction by the individual.
You create this sort of ongoing feedback mechanism and this feedback loop, where brands can be consistently and constantly aware of and have a better understanding of how their consumers, how their customers are experiencing their own brand.
This can enable them to have better and more personalized and more relevant messaging, that ultimately, the end goal is to create more evangelical users, create higher value out of those users, and make sure to retain and upsell and cross-sell those users across all of the brand's variety of offerings.
Who makes up the competitive set? Like, who are the companies that you compete with in displacing?
I would describe the competitive set, which has been around the vast majority of our entire existence. T he competitive set is not new to us, and it hasn't. It's evolved over the years. I t broadly distributes itself between a legacy marketing cloud set, so that would broadly be Salesforce Marketing Cloud, and Adobe.
These are two organizations that, you know, they've been around for a long time. Back in 2013, they each bought an email service provider, and essentially, the vast majority of their customers who are leveraging them are, by and large, only sending email through them. They are largely built on batch processing technology, as opposed to our stream processing architecture.
They struggle a little bit more with in-product use cases, the ability to actually engage with and message a user in the moment, while you are actually engaging with a digital application. T hey really haven't. If you think about what's happened to your inbox, your email inbox over the last 25 years, it really hasn't changed that much, and that broadly mimics the R&D investment that you might have seen go into a Salesforce Marketing Cloud or an Adobe.
That's fine. There is a, you know, there are a set of use cases that exist for that batch and blast email process. A s brands, again, have become more desiring more to engage one-to-one in a personalized, real-time, event-driven way, those companies cannot service those use cases.
At the other end of the spectrum, leveraging potentially more modern digital architecture are a number of point solutions, who came onto the stage largely by saying, "Hey, we are gonna be the best at pick your favorite channel." T hey largely each started single channel. Maybe their channel was mobile, maybe their channel was SMS, maybe their channel was web-based. Some of them, their channel was actually email.
What we've noticed is those organizations have been trying to broaden their channel offering over time. The problem is, when you start with an architecture that is equipped and built to service a single channel, adding additional channels can create a bit of a clunky and laggy experience, as you're actually trying to create campaigns that leverage and require a more seamless cross-channel messaging strategy.
I view Braze as being very well-positioned to compete and win against both of these sets of competitors. T hat's largely the way I would describe the competitive set today, and like I said, it's been this way for the better part of our 13-year existence.
When you think about growth going forward, given that this is a subscription business, how do you think about growth through the acquisition of new clients, layering in new services, versus possible pricing power that you may have with the existing clients?
There's a couple of growth vectors that we certainly lean on and have been successful for us over the years, and then some of which, you know, we're seeing challenges, certainly in the current macroeconomic environment, which is not dissimilar from a number of other software companies out there, so a lot of times, we will land with a net new customer because the customer is actually looking.
Perhaps they already have a subscription with Salesforce, or Adobe, or even one of the point solutions, and they recognize that actually a net new use case that they are looking to service. They actually don't have a technology solution that can handle it.
Maybe they're looking to do more in product, maybe they're looking to do more real time, maybe they're looking to collect more first-party and zero-party data from their users, and their existing technology stack and ecosystem can't service that. T here are a great landing way for us with a new customer is to service a brand-new use case.
Sometimes we will actually do a more fulsome rip and replace, and sometimes you'll have customers who are actually looking to do consolidation, vendor consolidation, because they have tried to cobble together over the years the ability to mix and match different channel offerings that they have. T hat they wanna message their customers through a variety of different and evolving and increasing channels, and they haven't found the vendor that can satisfy all of them.
Over time, they've been purchasing software solutions across a number of vendors. When they learn about Braze, and they recognize that, "Oh, got it, I can actually do all of these use cases across just one vendor," vendor consolidation is another way we might get in with a customer. Once we're in with a customer, the growth vectors are varied.
Now, it's important to understand how we define the customer. We define a customer at what's called the ultimate parent level, which means that expansion across new geographies or new brands within an existing parent company, for us, those count as, w e look at those as upsells. W e have a number of different examples of expansion vectors.
Sometimes we will land with a brand in a particular country, and then marketing organizations around the world of that same brand have decided they also wanna be leveraging Braze's technology, they will come to us and actually buy net new contracts, they will be new order forms within the existing parent company,
within our existing customer base, we definitely see volume increases across existing channels, so a customer might buy email, they might decide to buy more email, or they might decide to expand their channel use cases, they'll buy email, then they'll buy WhatsApp, then they'll buy SMS, and so we have the capability of broadening out the SKU, number of SKUs that customers are actually purchasing.
Along the same vein of geographic expansion is brand expansion, you can think of a large multimedia entertainment company where we might service theme parks, and then we might service hotels within the organization.
We might service cruise lines. We might service their streaming service. These will also represent net new growth vectors that we have within an organization. W hat's great, given the diversity of these different growth vectors, is we're able to expand with our customers in whatever way makes the most sense for them.
I would say, in the current environment, with the macro being as it is, when customers are doing slightly less net new investment in the ad space, specifically, where they're looking to acquire net new customers, and they're potentially investing less in initiating brand-new use cases, that will provide a bit of a headwind to us as we are, you know, looking to enable better existing customer or customer engagement.
That we sort of look at this as a moment in time, and the winds of change, of focusing on first-party data, real-time engagement or event-driven engagement, and the ability to to act in that sort of personalized, relevant way. These are broader secular changes that we believe to be really a one-way door. And we think that, you know,
We're under-penetrated across both the very large customers, where we have $10 million+ accounts, and even across sort of smaller customers. There's still so much running room for us to go that, you know, we're really excited about the prospects for our future.
You mentioned, I think, kind of macro factors that do determine how quickly that you can attain new customers or gain share within the customers. W hat are the other sources of friction that, you know, that sales tells you that make it difficult to convince customers to switch or add services? J ust 'cause it seemed like your solution is best-in-class that would come fast. I'm just curious the things that do hold possible clients back.
In some cases, for example, there are companies or industries whose business model is fairly well-protected, either by high capital costs, barriers to entry, or high regulatory frameworks. W e, for example, service a number of challenger banks and payment processors in the financial services sector, we don't have any too-big-to-fail banks as of yet.
You know, until the too-big-to-fail banks recognize that, "Oh, wait a minute, we actually have to migrate and improve our digital first-party based stream processing architecture, digital engagement for our end users," they may find that a batch and blast, you know, email and more of a legacy system satisfies their, their needs. Similar in the healthcare industry, we have a lot of health and wellness organizations, but no major hospital chains yet and no major insurance carriers.
There is, for some parts of the industries that we service, there are barriers that currently are sort of there, that don't create the impetus or the catalyst yet for that migration to a more best-in-class service. The macro headwinds that we see. Back when the macro really started to deteriorate, we experienced quite a bit of disruption in the sales cycles, and so you would have plans for growth initiatives and new customer acquisition initiatives that would be tabled.
A s those were tabled, and as those investment dollars were taken away, the need for Braze to come in for that new use case or that new initiative would fall away. F or some of these things, it's not a question of if, it's a question of when.
It's really just the macro kind of slowing down the pace. It is work for an organization to say, "Hey, I'm on this legacy technology solution. I really wanna migrate to something that's more sophisticated, and forward-thinking, and capable of elevating the capabilities of my brand, vis-à-vis my end users." I f they have a bundled set of cloud offerings from Salesforce, it can be harder to convince them in an environment where budget dollars tend to be tighter, to go ahead and make that change.
Now, usually, again, once they've recognized that the use cases that they're trying to execute on are just not. They're not capable of being executed on other services, generally, that's when we end up winning the business.
Your average contract tenure? Are they annual contracts or are they multi-year?
All of our contracts, just talking about the kind of nuts and bolts of the business model. A ll of our contracts are minimum term of one year, dollar-weighted average of two years. Our entitlements are sold on a volume commit basis, and you can trade up during the contract term.
If you need more monthly active users, or you need more volume-based messaging, or you wanna buy a new channel, all of that is possible within the confines of the annual contract, or a two-year contract, or a three-year contract. Upsells are fine. ownsells during the contract are not permitted. We do not sell seats, and so that vector does not, we're not sensitive to, you know, growth or shrink of the actual marketing organization.
We recognize revenue ratably, so because all of the volumes and dollars are committed at the start of the contract, the actual usage of the entitlements does not define the evolution of our revenue model. O ur revenue model is a ratable revenue recognition over the contract term. T he only thing that is sensitive to usage is actually our gross margin because our cost structure will evolve as the usage changes.
T he costs are sensitive to usage, the revenue is not. So Q4, for example, for us, tends to have a quarter that's a quarter where there's quite a bit of activity, both from a messaging perspective and just general activity and data gathering on platforms because customers are engaging with these applications to a high degree, both in the retail sector, but across the board. W e tend to see more gross margin pressures in Q4.
That's helpful. Thank you, and maybe switching gears to everyone's favorite topic, AI. What does it mean to Braze? You know, and the companies like yourselves, that AI's been part of the business model probably since the beginning in different ways, right?
How you think of AI, you know, the way it's been used for the past 13 years of the business relative to, you know, this new fun toy that we're all talking about?
It's interesting how you know, a little tongue-in-cheek with it being a, a fun toy. We try to-
I don't mean that. Yeah, it is a big deal, but, you know, it gets a lot of attention, a lot of this has been embedded in companies. Yeah.
It's a very big deal, but what's interesting to us is sort of in the last two years, when it sort of had a bit of its coming out party, we at Braze kind of looked around and we're like: "Yeah, but wasn't everybody using, you know, AI before? Wasn't it part of other people's roadmaps before?" It's been part of our roadmap since the beginning, AI and machine learning has been a part of our capabilities as you think about more sophisticated, maybe campaign A/B testing.
Our system, for quite some time, has been capable of enabling brands to run multiple variants of their campaigns with their existing user base, and having the system learn which ones are actually performing better in real time, and then automatically adapting to the better outcome, choices.
You know, that has been around for a while, the ability to for the system to learn which end users are best messaged by which through which channels. This is why being natively cross-channel and having the cross-channel communication be so seamless is so important, because you want the system to actually be able to abstract away the complexity of which channel are you, the marketer, going to use?
No, let the system decide, you know, which channel is actually best for this type of campaign, for this type of end user, and for the situation-specific dynamics at play. So we've had kind of that intelligent path and intelligent architecture from the beginning.
Now, as the AI has improved, and in particular things like transformer technology, which is the T in GPT, and LLM models, we've certainly embedded those capabilities into our tooling in a variety of different ways. And so the most basic, we have integrations with ChatGPT and DALL-E, so that marketers, as they think about, "Oh, well, I'd like to, you know, create five different variants, six different variants that I'm going to test, and I want them to each have these different characteristics."
You can, with natural language, ask the system through a prompt to develop headlines, taglines, you know, text, images, and then we always. You know, those are meant to be just a starting point for a human to then review and then execute the actual campaign strategy. And so we do have those integrations and have had those.
We have had those integrations since before those kind of names became buzzwords in the industry. So as soon as those capabilities started to exist, we've had those integrations.
Then leveraging, as I mentioned, that more sophisticated transformer technology, we have more sophisticated item recommendation engines that are sold as a premium SKU. Now, when people hear things like item and recommendation, they think, "Oh, well, that's a retail thing." It doesn't have to be. It's actually applicable across any type of organization that has the concept of a catalog of offerings. Those can be any number of SKUs.
They can be products. They can be services. It can be matching, you know, just supply and demand across a variety of different vectors. And then you have a body of takers or end users for whom you want to figure out: "Okay, well, what is the next SKU that I should recommend to someone?
What is the next product or service that I should recommend to someone?" And that is leveraging more sophisticated Transformer technology, which again, is the T in GPT, and that is sold at a more premium SKU level. Because that is an always on, one at a time, in real time capability, which, if you think about it, has some of the highest ROI because you're leveraging the system to provide the most intelligent and most likely kind of next action, or next purchase, or next engagement, that the individual might have.
Is there anything an investor should pay attention to in this segment that the large language models that you know as we go through this kind of transformational change that can differentiate companies in ways that weren't possible before?
Y ou know, I think when I think about the tooling that AI and the capabilities that AI can bring to bear, I think about the relative value add that it can provide to a software platform. I like to say that, you know, AI is just a...
It's a very, very rapid way of assimilating a whole lot of information, and then taking an action on something. Y ou want the AI to be as smart as possible, and to be as relevant as possible, and the only way to do that is to actually enable the AI, to feed the AI as much real-time, and relevant, and current information as possible.
A lot of people will talk about, "Oh, it's important to us to have the 360-degree view of the customer," which those things are important, and we can certainly house and have that data in Braze. I think one thing that differentiates Braze is the ability to live in the moment. What you want when an AI is actually engaging in that next action that is going to be the next highest value action towards an individual, it's important to understand, well, what is the context in which that action is being executed?
I think the combination of our SDK living right in the native digital application of the brand, combined with the stream processing architecture that can process all of this information in milliseconds. I think that leads to a better AI outcome.
If you think about, you know, how a comedian sort of tells a joke, timing is everything, and I liken, you know, AI capabilities to a similar style of conversation, and I'm not talking about chatbots in this situation. But if you are kind of laggy with how a punchline is delivered, and that's because, you know, there's a batch process that depends on some, you know, time-designed event that happens every...
Even if it's happening every minute, think about the patience of an individual in the context of a digital interaction to wait a minute for the right piece of information to come through, or the relevance, how the relevance might change over the course of 60 seconds or two minutes.
You'll have companies out there that tell you they're real time, even when they're running batches that are 30 seconds, or a minute, or even five minutes. That can be defined as real time to some people, which is why we like to use the word event-driven, because when you are event-driven, that means you are responsive, and you are by your very nature relevant, and so we like to think of our software as actually enabling the smartest use of AI technology, because we are feeding the AI the most relevant and current information.
Good. Very helpful. Last question. Switching gears and just, 'cause we're all financially minded, even though this is more of an education session, I'd be interested, to the extent... Have you shared, or are able to discuss kind of longer term, how you think of, the growth band for the company over an extended period of time and margin profile?
We haven't provided any kind of sort of long-term top-line CAGR or anything like that.
What we did have an analyst day a few weeks ago at our customer conference out in Las Vegas, and what we did is we tried to elucidate a bit of a framework for how we might return to becoming, once again, a Rule of 40 company.
We have historically been a Rule of 40 company, as we have sort of continued to grow, but growth rates have come down as the macro has kind of, you know, had its impact on us and a number of other companies out there, we've pulled away from that. T hat's, you know, very obvious in the numbers. W hat's important to us, growth at all cost has never been our mantra, ever.
That was true before I got here, it was true as we were leading into the IPO, and it's been true since then. C ertainly as the growth rates are going to evolve in a variety of different ways, and we've provided a framework that has a range, our plan is to add a relevant, you know, and a certain amount of incremental operating income margin annually in the context of whatever that, those ranges of revenue growth might be.
The idea, and this is a three-year, i t's a framework that should be applicable over the next three years. Y ou know, we want that framework to be robust against a variety of different market and macroeconomic conditions.
Okay, perfect. Well, really appreciate the time. It's good to see you again, and appreciate the, you know, the intro to the business and the education. Thanks a lot.
Have a good afternoon.
Thank you.
See you later.