All right. Thanks, everyone, for joining us at the Barclays TMT Conference. I'm Ryan MacWilliams, mid-cap software house covering DevOps, Cloud Comms, and MarTech. This is my triumphant return to the Barclays TMT Conference after I missed last year due to the birth of my son.
Congratulations.
I didn't have a good excuse for this year.
No, no.
With us today from Braze is CEO Bill Magnuson. And for the folks in the room, we won't be taking questions directly from the room or from the webcast. So if you do have questions, please email me at Ryan.MacWilliams@Barclays.com, and we can get those questions in. So Bill, just to start, not going to do the normal Braze background question, but I thought the Project Catalyst announcement at your investor day and how you talked about earnings was interesting. So we'd love to hear details around what that can mean for Braze and how to think about getting that in customers' hands next year.
Yeah, absolutely. So I think when I was asked recently, earlier this week, actually, how long have you been working on Project Catalyst? And it's an interesting question because there's an answer which is like, well, 13 years, right? Because a lot of the foundations and a lot of what surrounds it that's going to make it successful have obviously been part of our product strategy the whole time. Probably a better answer goes back about seven or eight years ago, which is when we first launched Canvas. And so Project Catalyst, for those that don't know, is just the next stage of our artificial intelligence machine learning investment around the kind of Braze AI umbrella. And what it'll be is effectively experiment optimization that lives within the context of a Braze Canvas.
And so for people that don't know, Braze customer engagement platform , what Braze Canvas is, is our visual development environment where primarily marketers can build the kind of companionship to their customer journeys. And I use that distinction because I think that modern customer engagement is really embracing that role for marketing and messaging to be there, collecting first-party data, understanding the context of the customer as it's evolving over time, communicating flexibly across different channels, and kind of being there to help detect when consumers are straying from the well-beaten or the positive path, as well as encouraging them to make purchases, become subscribers, become advocates, stay more engaged, and really build stronger relationships over the long term.
That is a lot that we've been very excited about building into Canvas over the years, including a lot of great artificial intelligence and machine learning-driven optimization features, including things like Intelligent Path, which uses the same type of personalization that you'll see in an Instagram or a TikTok feed that kind of learns about your preferences over time by doing things like vector mapping a user profile that we've built up and then seeing how other people with similar mappings to you have responded to different variant options in the past. Within Canvas, you're actually able to integrate that directly at any point in a customer journey or a flow that's being created.
If you go back further in time, simpler things like optimizing the timing that you would deliver a message, choosing which channels, being able to do other forms of experiment optimization, or also just control flow, which is like during a complicated multi-touch journey where a customer might be going from an app to a website to interacting with your email loyalty program or maybe a connected TV product or a connected fitness product, right? The modern digital journey is one where consumers are constantly moving around across all these different platforms and touch points, and being able to kind of keep track of that real-time context and then engage them with the right relevance and timing and delivering in the right channels is obviously a pretty non-trivial problem.
And what we've been trying to do with Canvas over time is be able to kind of encapsulate the complexity of really trying to attach to and optimize those journeys into more and more places where you can just kind of trust Canvas to do it for you. And so timing, channel selection, experiment optimization, different messaging cadences, right? Different content approaches, whether or not are you going to do an appeal to FOMO? Are you going to try to be dry and informational? Are you going to try to be snarky and kind of play with a new brand voice? As you're launching into a new market or a new geo or a new socioeconomic category, how do you localize and adapt those as you expand your product offering? Usually, you're bringing in new customer personas over time because those features appeal to people in different ways.
And so there's a lot of reasons why there's kind of a moving target around what the highest level of relevance and the best engagement is from a brand over time. And to be able to allow for the programs to automatically adapt themselves to that is obviously super powerful. That's a force multiplier on marketer productivity. And we talk a lot about the importance of Gen AI being able to help a marketer write faster copy, be able to generate images more quickly in other sorts of multimedia, to be more engaging. But there's also that same concept applies at the high-level strategy level. And so Project Catalyst, one of the other amazing things I think about it is that when Canvas is executing, it's executing in real time, it's making these automated decisions, it's helping enhance relevance, and doing all those great things.
But it's also kind of built in a paradigm where we get to a point in a customer journey and we say, hey, I'm about to send this message. Am I allowed to send this message? And it's kind of a greedy use of that marketing pressure capital that you have, right? We work in the first-party world. And so you've gotten this permission to communicate with a customer in a way where you have a low marginal cost, which is great, far better than being in the advertising ecosystem and needing to kind of pay to rent someone's eyeballs every single time, right?
That's really a core thesis of Braze's focus area in first-party data is that the ROI that a marketer gets out of being able to kind of earn the right to put their message in front of people and earn the right to be able to engage with them in a low marginal cost way.
But there's a different way of looking at that, which is like, hey, I have the opportunity to engage with someone. What's the most valuable strategy that I could currently employ now that I have the opportunity to message someone? Or I'm at this point in the journey and I've gone through my normal onboarding flow. I've kind of said all the table stakes that I need to say.
And at this point, I really want to make a play for them to become a premium subscriber, for them to move to family, for them to try out these other features. And I've got a global varied customer base, right? I've got people who have been with me all different tenures. They've tried out all these different parts of the product. They speak these languages. They come from these cultures. They have this amount of purchasing power, et cetera. And when you consider just the sheer optionality that exists there and the number of strategies that probably would be required to have an ideal match, even if you have the capability like the TikTok or the Instagram feed to be able to match those people to the right paths, historically, you still needed to generate those paths as well.
And so Project Catalyst is kind of taking together all these concepts that we've been building toward all this time and starting to also then multiply that productivity by saying, hey, you now need to just give us the goal. And you're going to define the guardrails around that goal, right? We still are not kind of fully trusting, especially in these first-party universes, the generative content to just kind of go with it, right? So we want to put some constraints around it. And so Project Catalyst will give marketers that opportunity to define the constraints, observe how the system is running over time, and then be able to obviously define what their goals are.
We can use all these building blocks that we've been building over time with Braze AI and with the Braze Data Platform in order to run that optimization on their behalf and really start to amplify results that may have been theoretically possible before but would have required a huge amount of production to build all those variants. In reality, I think actually it was never really possible for a human to do this either, just because the sheer amount of complexity that exists in a modern B2C consumer base requires that you have the variant production that you can really only get from artificial intelligence.
So with that hyper-personalized real-time messaging at scale, right, now with AI, that's kind of been the dream we were working towards with Braze. And some of your gaming customers, right, or the leading edge are going to be able to get value out of this very quickly. What about for a customer that maybe is less willing to take chances or isn't as robust in their marketing methods today? What's a good first step for them with Project Catalyst?
Yeah, so I think that one of the great things about the Braze customer base is how diversified it is and how much different parts of the Braze ecosystem can actually inspire each other and, in fact, in non-competitive ways. We just had our big annual customer conference called Forge in Las Vegas a couple of months ago. And one of the amazing things is sitting down at a table, and this literally happened where I've got a gaming customer, someone from the dating online or mobile dating service, and then a car dealership from the Southern United States. And they're all sitting at the table together sharing ideas about how to push forward on their marketing programs in more data-driven ways.
And so first of all, I would say that no matter where you come from, there's a lot of great inspiration about how to really push ahead on the sophistication of strategies and also the creativity of the engagement that you're using. And some of the gaming has generally been on the leading edge of a lot of these things. But also places like, if you want to look at loyalty efforts, like airlines and hotels have been like, it's basically been their primary product and profit drivers for a really long time. If you want to look at mutually reinforcing product universes, you can look at Disney as an amazing example of that. He's been doing it for generations, right? So there's a lot of, I think, great inspiration to be able to drive these strategies from.
And one of the great things about Braze, and this has been a product mantra forever, is that we want to enable sophistication without complexity. And as I spoke about the evolution of Canvas, the visual development environment, one of the key things that we continue to push ahead on is that we look at what our most sophisticated customers are accomplishing, but sometimes through a lot of effort, right? Like a Canvas, some of our more sophisticated customers, you can open up one of their canvases and it'll literally have thousands of nodes and there's arrows all over the place. And it's very complex. But what we've done with development over time is we'll look and say, like, OK, what are they conceptually trying to do down here? And it's like, oh, well, this is trying to test different mixes of cadences and channel types.
We're like, all right, well, what if we just turn that into an experiment step so that you don't need to be this sophisticated in order to try out that strategy? We've been systematically doing that for years now. Those customers that were on the leading edge, they have the resources and the know-how to be able to go and then push the envelope even further, right? One of the great things about the way that we've kind of inspired Braze and continued to kind of work with the frontier of our most advanced users is that we're basically. I've always kind of visualized it as like they're off like foraging in the wilderness, finding out, finding new things and working really hard to do it. We're paving the path behind them so that everyone else can kind of follow.
And so with Project Catalyst, I think that the accessibility for our whole customer base will be there right out of the gate because it'll help you kind of use generative tools to kind of build content and design the guardrails around it. There are places where the human side of this, I think, comes into play. And there's a lot of interesting questions around how, for instance, even brand evolves in a world where you can do this much adaptation. So there certainly will be some brands where the way that their brand promise presents itself to customers is still highly static all around the world, right? And that's an important part of how they've built their brand strategy. But now that we live in a world where your brand could actually show up in different ways for Gen Xers in the U.K. versus Gen Z in Brazil, right?
Both of whom might be a really big and important part of your customer base. And it probably makes sense for your brand voice to not speak to those two groups in the same way, right? Those are the kinds of capabilities that Project Catalyst will make very easy to actually implement. But there is also this kind of brand apparatus of what is our marketing strategy. And I think a lot of CMOs and a lot of marketing organizations are going to have to have that conversation in the coming years. It's going to be pretty interesting to see how that evolves.
That's a good point. You kind of abstract the complexity of actually sending the messages and orchestrating the campaigns. But actually having more strategy around how you're going to do this is kind of going to be a lot more complex.
Yeah.
Just on your side, as you think about, Braze has done a great job of let's get the tools in the marketer's hands and then we'll think about charging later. Let's get utilization first. But as you think about how you charge for Project Catalyst or how monetization happens, I mean, not only any numbers behind it, but just how do you think that evolves over time?
Yeah. Well, so first of all, right out of the gate, everything that promotes usage of Canvas is already automatically monetized because we do sell the message volumes, right? So to the extent that Project Catalyst is delivering emails or it's delivering SMS or WhatsApp or what have you, we're already charging for that. Similarly, I think as we look further into the future, we shifted actually how we report this last year. If you look at Braze's annual message sending metrics that we shared last year, which number in the it's like we have over 10 trillion data points were processed. And then we have single-digit trillions of message actions. But actually within those message actions are also webhooks, which are not. That's not an email or an SMS. It's actually calling a web service.
And it might be doing a job that's sending data from one place to another. It might be working with a Braze partner and it's delivering things like direct mail. Or it might be interacting with the paid ecosystem and updating ad audiences in various ways. It might be going back to your own system and delivering vouchers, right? It's this very flexible concept of an action.
I think as we continue to develop Canvas and as we also see things with non-zero marginal cost, right, which would be like the invocation of LLMs to be able to produce content or make judgment calls, there's a lot of conditional logic that exists when you design a journey where today we're limited in that, not just because of the difficulty of building it, but because like an if-then-else kind of a concept needs to be very specific and kind of a binary decision that's being made. Whereas the logic and reasoning capabilities of an LLM actually can really help make these decisions in a more fluid way with consumers. But that obviously has a non-zero marginal cost as well. And so we launched our Flexible Credits model earlier this year in anticipation of this.
And the way that works is that you used to need to commit to individual messaging channels with Braze for the contract term, which is a minimum of a year, but could be as much as two, three, or four years, and do so on a line-by-line basis of like, I'm going to send this many SMS to the U.K. I'm going to send this many WhatsApp to Brazil. I'm going to send this many emails, et cetera. And that's created a lot of sales friction. It was kind of required just given the way that our prior pricing worked. But we're out of that world now. And all new Braze customers that purchase those premium channels do so with flexible credits.
One of the benefits of that is that for invocation of things like Project Catalyst or invoking LLMs in a message flow or doing other sorts of work within a canvas that's not just sending a message, we have this very flexible way now to kind of sell that compute time or sell that time. And so that's another approach that will be immediately available to us as we're experimenting with how we want to monetize.
And then the third thing, which we've already done for a number of Braze AI features, and this includes both the predictive targeting that we do as well as the AI item recommendations that we do on top of our product catalogs, is that we actually sell those in a freemium fashion, which is that the basic kind of general model that's used across all customers is available for customers to use with certain volume constraints around it. And then if they want us to start training bespoke models on top of their specific data or use more advanced kind of model features, that becomes a premium SKU that gets added on top. And so you have those three different dimensions.
Depending on kind of what we see in terms of early customer usage and how we want to monetize it and where the costs end up, I'm betting out we'll make decisions around that. Project Catalyst, I think also, if you go back to that concept of it being this place where the marketer will increasingly be able to just say, go accomplish this goal for me, and we figure it out from there, right? That's going to have a lot of flavors and dimensions to it over time. I wouldn't think of Project Catalyst as like one thing that we're going to launch, but rather as the beginning of a long journey that really kind of turns this programming concept on its head and is able to continue to abstract away more of these details.
And so I would anticipate that the eventual full monetization of this family of products will be attacking that in multiple ways.
So I'd love to hear about what does that long journey take Braze to in the sense that you log into Braze and you send a command line, or you send a problem you have or a goal. But also, would you log in and you would submit playbooks and it's like, from this campaign, this will be the cost, this is the ROI? I mean, where do you think all this ends up with now real-time personalization at scale?
Yeah, I mean, there's a lot of complicated dimensions to this that we haven't even touched on, but that even become things that are operational. You touched on budgets, and obviously, not all message channels are created equally from a cost perspective. There's also interesting considerations like deliverability, and deliverability is becoming more and more of a challenge across all channels. I think all of us have probably seen the evolution of our SMS inboxes over the last couple of years that they now have promotions and spam tabs and things like that. It's going exactly the way of email. Same thing with push notifications where the operating systems are increasingly prioritizing, sorting, and now even collapsing the notifications that are going in.
All of those changes to the messaging landscape are actually highly beneficial for Braze because as the most sophisticated choice in the market, the harder it is to simply get your message in front of people's eyes, the more sophistication becomes table stakes as opposed to nice to have, and that comparatively benefits us quite a bit, right? The harder that it is, the harder of a challenge that deliverability is, the more a brand that might just be like, well, you know what we've been doing is working for us so far. We're just going to keep doing it. It's like, well, no, actually, because the landscape continues to get harder.
Like one email a week doesn't work?
Yeah. And one email where like every single one looks the same for everyone, right? And we send it all at the same time, right? These are strategies that are not just standing still in time. They're actually deteriorating in quality. And so there's a lot of additional layers to this that are going to have to get optimized and worked on over time. And so those, I think, are really interesting implementation details that we've got a lot of experience in that are going to be really important to bake into these as they evolve. The other thing that I'm focused on as well, and we've seen this over time in the marketing space where there's been a lot of companies that have tried to build these black boxes infused with AI and machine learning. And the promise is like, just trust it.
It'll handle all your marketing, right? And all is kind of, there's different layers of encapsulation. But I think the logic flaw in kind of all of those approaches is that it assumes that you're getting to kind of the right answer for every problem in that one black box and that then the problem space will be solved forever, right? And we have all these examples where we intuitively know that this probably isn't correct, right? Like I provided some examples earlier about how your product expands and you've got new personas coming in, or you expand to a new geography or a new socioeconomic reality for the new cohorts of customers that you're targeting for your acquisition efforts or new languages and cultures or what have you.
And those are all reasons why you should expect either that you've got a moving target for how you're going to adapt and who you're trying to talk to and what's resonating. All your priors are constantly changing, right? And it also probably implies that when we just look at kind of models that you're probably going to want different ones, and then you're going to need a supervisor that's going to be able to know which ones to invoke in different ways and at which times. And similarly, the frontier of this space is moving so quickly that the idea that you would want to kind of snapshot one generation and have that be your strategy is also a little ridiculous. And I think we intuitively know that.
But the implication of that actually is that the environment that you deploy them in ends up becoming probably more important than the capabilities of those today, right, or in the near future. Because that ability to govern them, to regulate them, to monitor them, and to challenge them head to head against each other, I think, is a really interesting one. We would never expect a customer to just turn on Project Catalyst and just let it run forever without looking at it, right? And similarly, wouldn't expect people to put 100% of all their traffic in it right out of the gate either. In most cases, they'll already have a solution that they've programmed, right, in Canvas to be able to cover that part of the customer journey.
I entirely expect that they should use a Braze experiment step or a Braze Intelligent Path to test those head to head against each other. As we roll out new solutions or even some of these examples we talked about, which is like, hey, let's run these against each other. This one's going to have a premium budget. This is going to have a moderate budget. To be able to then have a higher order logic chain, which is like, all right, let's split these people into run this propensity model. The propensity model will tell us their likelihood to be a high LTV customer over the next two years versus a medium one. Then to be able to run those into different flavors of Catalyst, this one gets unlimited budget, right? This one is more constrained.
It's not allowed to use WhatsApp or SMS. It's only allowed to use push notifications and in-product messaging because we don't think that this customer is necessarily going to maybe their buying power just overall isn't as good. They're not as engaged, whatever, right? So there's a lot of additional, I think, work that will need to be deployed over time. And yes, of course, as the kind of generative reasoning gets better, we'll be able to trust them to make more of those decisions in a single block in a canvas. We'll just be able to encapsulate what was previously complicated into one thing. And we can say, go ahead and trust it, just like we did with timing and channel selection before, and we do with cadence testing and content testing now.
And as we'll do with the combination of those with the first versions of Project Catalyst, we'll continue to push ahead on those things. But I really think the role of the marketing professional that is managing and experimenting and building and governing and monitoring, reporting on, optimizing all of these things, and then also adapting the strategies as their own company strategy evolves around it and their product expands and their customer base expands, that's a very active process that's still going to involve human inputs. And it's still going to require an environment that provides agility, visibility, and enhanced productivity to the people that are running that.
It's always surprising to me how proactive you have to be around campaigns. If you have an email campaign and you wait six months and try to send it again, your delivery rates will go down like 30%-40%. So it sounds like there's always different gears and levers you can pull. Kind of switching gears, that was a really great breakdown of where Braze could be. Kind of in your most recent results for Cyber Week, you saw a strong reacceleration of messaging growth. We're trying to figure out here at this Barclays conference, since the election, have we seen any changes in demand? For many vendors, it seemed like Black Friday, Cyber Monday was really strong for messaging. Do you think people try to activate post-election around this opportunity, or do you think that there's something separate going on?
Yeah, I mean, I think that a lot of Braze's use cases are always on, right? You're growing. You're expanding your business. You want to be able to cultivate relationships with people over time. And so Black Friday, Cyber Monday is a really important opportunity to acquire new customers in many cases or reinvigorate old customer relationships. I'm sure that a lot of us got a lot of messages from things that we bought years ago. And we're like, oh, wow. Chris actually mentioned that he got an email promoting blood testing, like at-home blood testing. And it's like, I wasn't really thinking about that on Cyber Monday, but they clearly saw that as an opportunity. People have got their wallets out. Let's go for it.
They're going off blood testing.
Yeah. And then Giving Tuesday obviously comes up the next day. And so a lot of message volume around that as well for our nonprofit customers and a really important part of the year. But I think that when you look at things like the election or the NFL season and such, there's a lot of seasonal things that happen in the advertising world where it's like, were marketers waiting for the election? It's like, yes, but not necessarily because of the election result, but because all the political campaigns were bidding up all the prices leading into it, right? And you see the same thing with a lot of the advertising world around the NFL. We work with a lot of sports betting applications. And obviously, those are really big days for them as well as anything in sports news.
But if you're a mobile game, you turn off a lot of your advertising on those days because you're getting outbid by the gambling applications because they monetize better than you do, right? And so I would say when you're trying to read tea leaves like that, it's important to understand that actually there's a lot going on under the surface in a lot of these kind of performance marketing acquisition worlds. And then as a reminder, Braze lives downstream of acquisition, right? We were there in the early days of all new opportunities to kind of activate a customer, make a great first impression, get them off to the races as they're trying out your product and service. Ideally, you get your new customers to kind of integrate your brand into their life in a way that they value so that there will be stronger engagement over time.
And then you try to optimize that value over the long term. And that's an always-on responsibility. And the holiday season is a crucial time for new user acquisition. Back when we were much more predominantly mobile, today, obviously, mobile has been an important part of our history throughout, but we're very diversified across different kind of company types and different platforms now. But Christmas actually used to be our biggest day because people got new phones for Christmas. And so that was one of the most important days from an activation standpoint because when people would get a new phone, they would go and download more apps on that first day than they will the whole rest of the year, right? And so that's your major opportunity to try to engage and activate people.
Now, the way that people buy phones now, that's spread out throughout the year a little bit more. But you've got to kind of think about these important activation moments, acquisition moments, making the most of making a good first impression for those new cohorts. But yeah, there's a large aspect to it that's always on as well.
Sure. And better messaging is a good thing. But like you said, people are planning all year or planning multiple months in advance. So you have a good understanding of where people are trying to get at for Cyber Week.
Yep.
But just in general, are you seeing when people are talking about their performance marketing budgets for next year, is it a healthier conversation? Does it feel better or any differences?
Yeah, I think one of the biggest differences is actually the stability of the strategies and the teams, which is separate but related to the budgets that they have. One of the things that we saw two years ago is just a lot of CMO turnover, a lot of marketing teams getting disrupted, a lot of them getting downsized, or a lot of their hiring plans getting canceled or their agency support getting pulled, a lot of the ancillary budgets that supported their programming. And not only were these declines happening, but everyone was in a little bit of an existential panic. Like, am I even going to have a job next week? Is this project going to get canceled, et cetera? So that was two years ago. Last year, we started to see some stability in that.
People were no longer worried that their team was going to get cut, but it was a lot of new people, right? A lot of new CMOs, a lot of new marketing leaders, et cetera, and that lost a lot of the momentum and the clout that they have to be able to kind of push strategy and such. And obviously, the vast majority of the corporate world was focused on cost cutting and not on growth. And marketing tends to be a growth engine, right, and so I think where we are right now is that the teams are stabilized. We don't have any pipeline. We haven't had any pipeline that I remember recently where the signatory got laid off two days before the deal was supposed to close. That was happening all over the place two years ago, so these teams are stable.
They're working through their annual planning. We are seeing in some pockets, there's companies starting to move back toward that growth posture. Even without necessarily changing the budgets, at least deciding like, hey, we're going to expand our product this year. We might expand into some new geographies or we're going to, and it's not necessarily that the acquisition budgets are through the roof, but they're at least on an expansionary footing again as a company where they actually value value creation, right, as opposed to valuing cost cutting. And I think that when I look at, and it's one of the reasons why I use the phrase demand environment when I kind of talk about what most people would call the macro, because broadly, the macro is OK, right? But the demand environment for growth initiatives has been very challenged.
And I think that's the key thing that we're looking for is at what point are our companies going to start to turn the dial back over and start to value value creation more than they were and start to be on a more expansionary footing. And we're seeing, I think, signs of exhaustion on the cost cutting side, which is great. CFOs that are like, all right, I'm done looking at every single deal. Let's get back to a growth posture again. We remember how to do that, right? And for the sake of competitiveness, I hope all companies figure out how to find that muscle again. But that's the key thing that we're looking for.
Me as well. So look, that was a really fun insight into where we are today in marketing and also the future of marketing. If anyone has any questions for the Braze team, we can get those over. But Bill, thanks again for being here.
Yeah, absolutely. Thank you. Cheers.