I know, that's kind of crazy. I didn't hear that earlier when I was here. All right, anyways, thanks everyone for joining us today. My name is Scott Berg. I lead the enterprise software and SaaS research efforts here at Needham for those that are not familiar with myself. Today we actually have my top pick for 2026. I'm excited about this one. Woo! We have Braze with us today. We have the company's CFO, Isabelle Winkles, with us. Isabelle, thank you so much for joining us.
Yeah, thanks for having me.
Hopefully the day's going well so far. I guess, how about the 30-second elevator speech on what Braze does for the two people in here that might not know?
Yeah, so Braze is the reference architecture for the modern customer engagement stack. So a little bit more than 30 seconds, but if we go back to our founding in 2011, I'll do the quick version. The advent of mobile technology, brands and large enterprises started to use more and more of these platforms, digital platforms, to communicate with their end users. Proliferation of different messaging channels, digital messaging channels, really created the opportunity and the imperative for brands to start collecting and processing and leveraging high-quality first-party and zero-party data. And so when we think about Braze, natively built cross-channel, built on stream processing architecture, our SDK lives right in the native digital properties of the brands that buy our software. And speaking of data, we are existing at a level of data scale and processing trillions of data points and billions of messages.
And the Braze Data Platform is really enabling large, large quantities of data syncing in through things like Cloud Data Ingestion. We're syncing trillions of lines of data into the Braze platform. Our objective is to enable brands to have their data in there completely, quickly, and at a low cost of ownership. And when you add together the real-time data that's being generated by the platform and all of this additional data that can be streamed into the platform, and then add on top of that AI technology, you're really injecting intelligence into a very high-quality data set, all built on stream processing architecture, able to meet consumers exactly where they are on their own unique personal journey with a brand, increasing the ROI of that relationship between the brand and the end user.
That was pretty close to 30 seconds, I think.
There you go.
If I had Bill here, that might be closer to 30 minutes if I gave him the time, which is great. All right, let's talk about the industry a little bit, because I don't know, there's this agentic theme that's kind of sweeping through the industry a little bit. I guess the broader commerce industry is. I forgot where I was going with this question. All right, anyways, Braze is at the center of all sorts of things going on right now. It's been a long day, sorry. I guess, what new and early-stage commerce trends are you all starting to see that the typical investor might not be aware of, outside of the big high-level agentic items?
Yeah, so the themes that we're seeing, and some of this is kind of continuations of things that we've seen for a long time, but the customer journey is continuing to become more and more complex. With agents and these aggregators, it is becoming harder and harder to command the attention, the time of an end user, the dollars of an end user. And so the complexities around this are really leading to a need for brands to have more and more high-quality first-party data and to be able to engage directly with their end user in that direct-to-consumer fashion. And so I think these are wins that are at the back for Braze, given this is exactly what our platform enables consumers to do, brands to do, rather.
So agentic commerce is a popular investor focal point today. Might have been an industry note that I published today as well. But how do you view the impact of agentic commerce on the broader customer engagement or marketing software space versus what I think most are focused on, or at least a lot of investors are focused on, the e-commerce vendors?
Yeah, so it's important to understand, first of all, Braze is about 80% not e-commerce. So we're retail and e-commerce, 80% not retail and e-commerce. So highly diversified. But when we think about the agentic space and the impact across any industry, really it's about being able to siphon away the value of the end user. And so the question is, are these platforms going to, and these agents, going to behave in more of a closed ecosystem or more of an open ecosystem? And either way, Braze has a role to play. If the ecosystems end up being more closed, that will foster the need for more direct-to-consumer engagement and kind of the value that Braze brings through the other first-party properties that the brand is managing.
If, on the other hand, they end up more open, then it just becomes a new platform for us to enable that data capture through those platforms, such as the SDK that we've built for ChatGPT.
So I have a theory in this space. If I go back to the first conversation yourself, Bill, and I had when I first met you guys, I came away thinking, wow, the ability to hyper-personalize, because you guys have been able to do this for a while. This isn't brand new to Braze. But anyways, the ability to hyper-personalize with the Braze platform seems amazing, because I hadn't really heard this from another vendor yet at that point in time. But if I fast forward a couple of years, the entire marketing industry is now talking about hyper-personalization at scale. Is the market, with all this new kind of Gen AI talk, really just catching up and tilting towards what Braze has already done, or do you feel like you're playing catch-up to where everyone else thinks the market's going?
Yeah, yeah, so look, I think we've been ahead of the curve the whole time. And since our founding in 2011 and since the early days of the evolution of our tool, we have had AI incorporated into the tool from the beginning and have continued to be on that roadmap. So I do believe that we are seeing new and exciting tooling and capabilities out there. But really, the value of those tools is in the ability to marry the capability of the AI, the intelligence of the AI, on high-quality data that's able to live in the moment in a highly personalized way. And so our stream processing architecture, natively cross-channel, we are there in the moment with the customer.
The ability to inject AI into that level of quality in terms of the interaction between a brand and an end user is really where the value gets created. I would say that the market is actually broadly trying to catch up to where we have been the whole time. These new AI capabilities are really helping us to move our platform forward, but really grounded in capabilities that we've been working on since the beginning.
So obviously, this is a space that's had a lot of AI conversations the last couple of years in general. Do you think the customer kind of exploration of what these technologies is, has that been a headwind to the business around new customer growth, or has it maybe started to be maybe a tailwind? Your Q3 customer metrics were probably about as good as I've seen in several years.
Yeah, so look, I think what you're seeing is the results of kind of relentless focus on execution, and so I think the AI story is certainly helpful in the context of really wanting to leverage, again, that highest quality first-party data in a real-time and action-oriented, relevant way, but I think we've been really pleased with our results, and I think what you're really seeing is just relentless focus, and also, a couple of our competitors, especially in the legacy cloud, really, I think, dialing down some of their investment in the space, and so I think it's an exciting time for us to be participating in the space and taking on our leadership position and continuing to secure that, and so we're really excited about the value of the execution that we've been able to deliver.
So if we think about the agentic kind of general commerce or more agentic marketing space, obviously, lots of changes recently. The one product that you brought to market that I thought was super interesting was the SDK for the ChatGPT platform there. I guess, why does the embedded app opportunity look more like a platform for Braze than a channel that's more like maybe WhatsApp or Signal or something?
Yeah, yeah, so look, I think it's still pretty early with this, and we're not exactly sure directionally where ChatGPT will ultimately go with this. We obviously built the SDK in pretty record time. I think it took about two weeks end to end, but a good part of that was spent trying to get access to the dev environment. And so that speaks really highly to the flywheel that we've created from an R&D perspective. And so we are there ready for however that platform will behave as sort of a place where more of a platform where the customer and the brand can engage with each other. There can be an exchange of goods and services versus just a messaging channel where there's just kind of a directional flow of kind of information or data capture.
What I thought was impressive was, one, the speed that you mentioned that it came out, but of the seven, we'll call it brands or merchants that were on there first, two of them were already Braze customers, Spotify and Canva. They're public, so they're publicly known at least, so that's interesting, but on my Q3 callback, when I was having this conversation with Bill, he mentioned that there's 10 more brands that are ready to go live on it. I assume probably Q1-ish here, but seven of those brands are also already Braze customers, without naming who they are necessarily. Whether it's the first two or any of the initial seven, do you have any learnings or feedback or anything that would be interesting to share on how they're using that in terms of the value that it brings?
Yeah, I think it's still pretty early. Like I said, I think the fact that we have the SDK ready to go and our customers are free to implement it when they're ready, I think they're still figuring out exactly how they're going to leverage the app or the new platform. But the reality is we're all over it, and we're really excited to be able to support them however we can through that new platform.
A company also acquired a company called OfferFit mid-2025. It's been since rebranded as Decision Studio. Customer reaction to this product at your customer conference in September was amazingly strong. The Yum! Brands customer you had on stage had a great kind of use case, and you talked really well about the ROI there, and of course, Kevin spoke highly about it on the bus tour that it had in June. I guess, why is that so compelling? When someone like, whether it's Yum or others talk about it, because of the other customers just couldn't wait to adopt it, it was really kind of interesting, but what does that value bring that's unique that they can't get today?
Yeah, so it's a reinforcement learning engine. And we sell it based on optimization of a business use case. And so it moves you from a place where you're kind of engaging in sophisticated cross-journey, cross-channel customer engagement journeys, and trying to optimize for each individual customer to a place where you're actually selling performance and where our customers, the brands, are able to see measurable uplift because you're optimizing for a particular business outcome. And so another customer example beyond the Yum example is Capital One uses the tool in order to convert business account holders, trying to get them to get the credit card. And that is a very high-leverage use case where even small uplifts in that trajectory have meaningful economic impacts to the brand. And they've seen meaningful positive flow from that.
And so it's great to see that reinforcement learning engine really be able to take high-leverage use cases at scale, provide value to the brands. And we're excited for the growth prospects there.
Hadn't heard the Capital One example yet. The Yum example, your customer talked about a one-point improvement in conversion rates, which is massive in that area.
Yeah, exactly.
Good start. I guess with the addition of Decision Studio and, of course, other innovations on the platform recently, are you seeing customers fundamentally use Braze differently than in past years, or are the additions seen maybe more as evolutionary than something that we should think of as revolutionary?
Yeah, I mean, look, I think they're evolutionary, but it's really exciting to see some of the ways in which our customers have been leveraging, say, things like Agent Console. And so Agent Console is in private beta. It'll launch in GA a little bit later this year in Q2. And we're seeing some pretty creative and unique use cases where our customers are, with the use of natural language, just building agents to go out and accomplish tasks and build campaigns and optimize and be in kind of this continuous sort of always-on mode of high-quality customer engagement, but built through agent technology. And so a beauty provider app store, for example, when a customer books an appointment, there is a lot of contextual data that's available about the individual, the appointment, what they're getting, what they've gotten in the past, who they are.
And then a message is immediately sent out to upsell or cross-sell additional services. We have an e-commerce platform, a marketplace that uses Connected Content to figure out what is being requested by the buyer sphere and then feed that information to potential sellers to give them high-quality information about what is in demand. We have a spirituality app that is using multiple agents to go out and write copy for the various spiritual leaders and provide daily prayers to their end users. And so this is content generation. It's personalization. It's relevance. It's an always-on motion. And so it's been really exciting to see some of the use cases that have been developed by our customer base. Now, they are evolutionary in so far as we've always been talking about personalization. We've always been talking about messaging at scale. We've always been talking about real-time and relevant contextualized data.
But now it's getting done more efficiently through these agents and with higher levels of context. And so it's just exciting to see the direction of travel that our own customers are pushing the platform towards.
I know this is going to surprise you, but I have no knowledge of the process in a beauty store. I was born with this look. I have not been to one yet. So I guess kind of last question, I guess, around that is I noticed this at your conference. I had a bunch of customers talk about actually wanting to procure budgets around some of these functionalities, which was kind of like the first time I had heard that or whatnot. Are you seeing that in any of your buying processes, that customers are finding the dollars in more apps to try to buy some of the new functionality, or is it really still on the core Braze set that we're all used to?
Yeah, I mean, most of the dollars right now are flowing to the core Braze set. And things like the AI Operator, we don't charge for. So there's some features and functionality that are just embedded into the core platform. But we are really excited about where we go with things like Agent Console, which will be incorporated into our credits model when it goes GA in Q2. And so that will be the monetization approach for that feature. And then, of course, Decision Studio, which sells for about $250,000-$300,000 per use case. And so we're excited to see these mix in more over time, starting from small dollars. And so things like monthly active users and the overall messaging dollars and messaging credits will continue to take on center stage in terms of the contribution to the top line.
But excited about these new AI capabilities that are starting from a small base but are gaining very rapid traction with customers.
Funny, that's a great segue to my next question, which happens to be on Flex Credits.
There you go.
You previously announced, I'll call it a new Flex Credits pricing strategy. It's not new today, but you've been talking about it for roughly a year now, I think it is, and recently, you've talked about how WhatsApp and Content Cards and how they've become your fastest-growing channels recently. I guess, is there a correlation here between all that?
Yeah, so I think generally with Message Credits, we try to do a couple of things. So we've sort of lowered the friction in the selling cycle. So it used to be that customers had to kind of pick the volumes by channel that they were going to actually send. And it's frictional. It's complicated and just takes a lot of time. So here, it's one set of entitlements. They're purchased upfront. The revenue recognition model is identical. But now there's fungibility. And so customers are automatically enabled for the totality of our set of available channels. And what that means is we've lowered the activation energy for them to do some experimentation. Because if they want to try a new channel, it's not a new procurement cycle and new dollars and a new signing of a contract. It's just experimentation within the tool.
What we have seen with the Flex Credits is the ability for customers to actually try out these new channels and, one, go more cross-channel, which is great. Customers with more channels end up being stickier with Braze. And they're able to experiment more. And they're able to utilize their credits more fully. Because in a world where you committed to a certain entitlement in a certain country, if you weren't able to use that fully, you can't really move it over to a different country. Here, it's fully fungible. Now, what that means is customers are, at the time of renewal, there isn't as much pressure on the downsell, right? So because if they overpurchase, there are opportunities for them to use them across any channel. So the utilization rates of the purchased messaging entitlement are becoming closer to the purchases.
So that's great from a customer retention, customer satisfaction perspective. There is a margin impact here that does have to be understood. In a world where the quantum of breakage at the end of a contract, so there isn't as much downsell needed, there isn't as much unused entitlements, and the consumption is greater, that is going to mix into the cost structure. And we have been seeing very strong adoption of things like WhatsApp and things like SMS. And so from a cost structure perspective, that is similarly mixing into the cost structure. So from a gross margin perspective, while we're really excited about the value of credits, the ability for our customers to use their entitlements, there is a direction of travel and impact on margins that we have to be aware of as the consumption of the purchased entitlement is approaching. The gap is closing there.
We are seeing strong adoption of some of these more premium channels. Great to see the proliferation, but have to take into consideration the margin considerations.
My guess is if you see that those usage trends continue, it's probably a pricing element to kick in to help offset what some of that looks like.
Yeah, and then there's upsells, right? If a customer is consuming up to their purchased entitlement, then maybe they bought a little closer to the pin in the beginning of the contract, not exactly sure how much they were going to need to use. Channels are working for them. They're able to try new channels. There's upsell opportunity and then less risk for downsell at the time of renewal.
Okay. You mentioned some of the new AI functionality that's exposed to the Flex Credits. Is that kind of the pricing strategy you think for some of the new innovation coming forward on a consistent basis, or do we see it more dedicated to a product-by-product like historically?
No. So for right now, so obviously, Decision Studio sold kind of separately as a single SKU. The AI product that we are putting into the Flex Credits model is largely around the Agent Console. And right now, in the private beta, we're actually using the private beta to figure out number of invocations and volumes of invocations and cost of invocations of the various LLM models, whether it's our LLM or the customer's own sort of bring your own LLM. And there will be a portion of a credit assigned to the consumption of triggering an LLM step. So that will be how the credits get used is at the invocation of an LLM step. And we're really excited to get that into the Flex Credits model in the next couple of months.
Okay. I guess last on product for me. I believe Bill recently said the company processed 100 billion messages during Cyber Week. That's a lot of messages. But my math says that's about 13 or 14 messages that week for every person on the planet. It's just what the rough numbers look like, at least as the math. It's kind of mind-blowing at that scale. But do you think that scale is well understood by customers that you are trying to attract? Because when I look at other vendors in the space, I don't hear of anyone actually delivering messages at the scale across all the channels you do.
Yeah. And when we see our strength in the enterprise, we know our customers really understand the capabilities of the platform. And when we see what happened over Cyber Monday, Black Friday, and actually simultaneously in another part of the world, the Korean Grammys were taking place. And that triggered a 3x increase in the amount of data that was running through one of our servers at the same time that I think it was happening on the morning of Black Friday. So we have not only the capability of processing enormous amounts of data at scale, but we can do it almost on a surprise basis and able to scale up very, very quickly. So yeah, I do think that's fairly well understood by our existing customer base and prospects when we go out and market.
Okay. Moving to go-to-market a little bit. Longtime Chief Customer or Commercial Officer and President, Myles, departed the company after about 10 years this past summer. I like Myles.
Beyond Myles.
Had a lot of good conversations. Yeah, he was great. You brought in Ed McDonnell as CRO, who has 10-plus years in the marketing space. He's well entrenched in this space. But what does he bring that may be different than Myles, or what is he already doing differently that we should pay attention to in fiscal 2027? Because it's his first real year to make a true impact from start to finish.
Yeah. So look, I think the benefit of what Ed brings is some of his experience actually from outside of Braze. So he spent a long time at some of our competitors. And he has just a vast amount of experience at the enterprise, but also kind of at other levels as well. And I think he's going to be really helpful at furthering a number of strategies that we have already been putting in motion. So I think it's a little bit less about doing things differently and really about taking some of our efforts around verticalization, internationalization, going out and finding additional customers that are heavy email users. There's a whole host of things that are deepening our relationships with the partner ecosystem. Things where he's not only seen the movie, but he's seen the remake and helping us kind of continue to further along those lines.
And broadly speaking, just helping us sort of move faster and also increase our overall product moat with our customers. Now, the thing to keep in mind is he joined in the summer, and some of those leadership changes can be a little bit disruptive. And we did see a little bit of elevated departures among some of the sales folks across the organization. Nothing that surprised us. And we are investing to actually increase some of that capacity. But it's important to note that with some of that, the capacity will come online more fully in the second half of next year. And so I would look for that to not be a Q1 or Q2 event, but more in the back half of the year from a ramped sales equivalent growth perspective.
Okay. The company has also been talking a lot about pricing and packaging changes, even outside the flexible kind of credit discussion that we already talked about. But I guess, what other changes have been made? What type of impact are you seeing on customer adoption or maybe even some of the expansion opportunities?
Yeah. So there's kind of two real main areas where we've made some changes. And really just one, to better align how we sell with how customers want to buy, and two, to dispel some of the FUD that gets thrown around in the marketplace and just make it easier to sell to customers. The first one around making it easier is we've eliminated data point caps from our monthly active users. And so that just makes it easier for customers to understand and recognize that they can leverage Braze, use it, and not feel like they're being capped by some data point threshold. We price monthly active users such that we don't end up upside down on that. But we are very comfortable with enabling customers to feel like they can just use the platform to the largest extent of its capability and usefulness for a customer.
The other changes are more on the customer success side and finding ways to actually better align the services that customers are willing and wanting to pay for rather than just bundling services into the platform, actually selling those uniquely to customers, and so you might see as a result of that a little bit of an uptick in the professional services revenue as a percent of total, mostly because we're simply finding ways to better monetize and create more efficiencies with those humans and those assets.
Okay. Two more questions for me, and I'm happy to take questions from the audience. The question I easily get most often over the last six months, which I'm sure you guys hear often as well, is, "Are you through the zero-interest-rate downside yet?
Yeah. So look, I think the language we've been using is I think we're sort of through the belly of the beast, right? So the attenuation of the downsell that we've been able to realize over the last several quarters has been meaningful. We've been able to stabilize our in-quarter dollar-based net retention. And that's been really exciting to see. And so I think we're through a lot of it. We do have a big available renewal dollar quarter in Q1. So we've got eyes on just making sure we're managing that. But I think we're through a lot of it at this point. And so we're very happy with some of the stabilization that we've seen in our metrics.
Okay. Kind of to that point, Net Revenue Retention has turned positive, mildly positive as of late. Now that you seem to be seeing maybe a little bit healthier demand environment, what should expansion rate be over, I don't know, kind of the intermediate term? I don't recall that you've actually kind of phrased what your expectations are now that we've maybe bounced a little bit off the bottom.
Yeah. Yeah. So rather than give any particular guidance on what the NRR sort of should be, what I like to do is to provide a framework for how people can think about that in the context of broader top-line growth. So our top line is broadly comprised of approximately half upsell and half net new. So the net new just mixes in at whatever proportion because there's nothing that nets against it. But the upsell has obviously the downsell and any modest amounts of churn that bake into it. And so using kind of even numbers, if you're growing at 20% and about half of that is related to upsell or half of that is related to net new, you'd get the 10% benefit from that. And then on the upsell, that would be 110% if you had no downsell.
But then, of course, downsell would net slightly against that. And so that's just an illustrative mathematical example. But the thing that's important to understand is that we broadly look at our bookings quarter in and quarter out. It's about 50/50 upsell net new. It ebbs and flows 60/40, 40/60, but it's kind of in that general zip code.
Okay. With that, happy to take any questions from the audience if there are any.
Sorry. Just to be clear on that, you're saying that the upsell piece has to overcome the churn to then grow so it's growing faster. The gross dollars are growing faster than the new, but there's some churn.
No, I'd say it's the other way around. The upsell dollars and the net new tend to be about the same, but then the upsell dollars have the churn that nets against them.
Okay.
Any other questions?
Can you honestly just explain in very simple terms what is the moat? Is it the screening? What is it? Is it the connections to all these different messaging channels? What is it when somebody comes to you? And then who is your competition?
Yeah.
How do they compete? What are they?
Yeah. Yeah, so I'll start with the competitive set. So we have, at one end, the marketing clouds, the legacy marketing clouds, at one end of the spectrum. And they largely service customers that are mostly email only. And they've tried to kind of cobble together other channels. There's some latency, some interoperability issues that exist with them. It's not sort of a fully integrated stack the way we are fully integrated. Also, they're largely based on batch processing technology, whereas we are natively stream processing architecture. So from the legacy clouds, the legacy replacement cycle will usually look like a customer that is trying to engage in more cross-channel, more real-time, more contextualized. And when you bring all of those things together, they're trying to experiment more with different use cases. That is when Braze is kind of ripe to take over a contract such as that.
At the other end of the spectrum, there's a series of kind of subscale point solutions that can be sort of a good entry point for an organization that is just thinking about, "Okay, I need to now start thinking about customer engagement." But similarly, I think they very much try to compete on price, but in some cases lack the full feature functionality of the fully integrated stack, the SDK that sits right in the native applications of the brands, the stream processing architecture, and again, the natively cross-channel. A lot of those, we used to call them point solutions, a lot of them started single-channel and then realized that actually to fully add value and to not become just a commoditized action layer, they needed to go multi-channel.
And so because they started single-channel, it's a very different sort of architecture problem to actually iterate from being single-channel to going multi-channel. We started multi-channel right from the beginning, which included both in-application messaging capabilities as well as out-of-product messaging capabilities. And so we did the hard work upfront to enable brands to continuously add more and more channels, and we can add more and more channels to our suite. And our new channels are generally automatically featured with all of what's available in the stack. And when we add new feature and functionality in the stack, it's generally made available to all of the channels that we have. And that's by virtue of kind of how we were fundamentally integrated right from the start.
Probably have time for one more question if there's one. All right. With that, I'll give everyone an extra minute to go fight the elevators. Isabelle, thank you for your time so much. Appreciate it.