All right, we're good to go. Perfect. All right, thanks everyone for joining us today. My name is Scott Berg. I lead our enterprise software and SaaS research efforts here at Needham. Today we have with us Braze who's been public, I guess, just over three years, three years in November.
Yeah.
I remember the time and the dates. And with us, we have the company's CFO, Isabelle Winkles. Thank you so much for joining us today.
Yeah, thanks for having me. Great to be here.
Yeah, it's been a fun four-plus years knowing the company. Love the product, as you guys know. But why don't we start off by just having you tell us a little bit about the company for those that are less familiar with Braze?
Yeah, so founded in 2011, we are a cross-channel customer engagement marketing software platform, so enabling brands of all shapes and sizes across the globe and across many, many different industries to engage their end users in ways that are in real time, respectful, actionable, and in ways that really kind of help amplify the brand messaging and get the highest ROI out of our platform, enabling the highest level of kind of monetization of their end user base.
Awesome. Well, one of the things you mentioned is Braze is a cross-channel marketing platform in a sea of dozens and dozens of effectively single-channel solutions out there. What's occurring in the marketplace today that's really driving customers towards this multi or cross-channel type of platform?
Yeah, I think back in the day, marketing teams used to be kind of set up as a fairly siloed across. You'd have an email marketing team, and maybe you'd have a mobile marketing team as mobile started to become a thing. And as we, as users, as end users of global brands generally, we are demanding that brands meet us wherever we are on our own personalized journey with the various brands. And what that means is that you got to know where I am, what I'm doing, what my last action was, how I like to be communicated to, what means of communication is going to be the most relevant to me. And as the number of platforms of communication continue to increase and proliferate, there's just an increased need for that personalization.
And so what might be effective with one end user sending SMS might be not really effective with another end user. And it might be much better to send an email or to work in product with that individual. And so you really need the ability to be, I don't even use the word omnichannel, but really cross-channel to understand how the individual is interacting with the brand across a number of different platforms. That might be a connected fitness device, a smart TV, a mobile device, an email, sorry, a web page, and then be able to message the individual in the way that is the most relevant and actionable for that individual at that moment in time. And again, as the number of platforms has evolved and the number of messaging channels has increased, it's really heightened the need to be a cross-channel platform.
Actually, one of the advantages of Braze is we actually started life as a cross-channel platform. We went commercial with four channels right from the get-go, including both in product and out- of- product channels. We did the hard work upfront in order to enable a more seamless cross-channel experience as opposed to some of the other competitors in the marketplace, which really started single channel and then had to sort of bolt on net new channels in a way that can lead to limited interoperability and kind of more lag in that customer experience on the user side.
I know we're going to move to product. It's kind of my favorite area to talk about, at least one of the company's key points of differentiation is around your real-time data streaming and the ability to personalize messages and that connectivity with those consumers. This AI trend over the last couple of years has really kind of heightened, at least maybe not expectations, but interest in this personalization space. How much of that's been a tailwind, do you think, to your business here today?
Yeah, so we've had AI embedded in the tool in some way, shape, or form across different dimensions basically since inception. So AI has been part of our roadmap from the very, very early days. I think generally speaking, AI would tend to be a tailwind to us. And I think you mentioned kind of the real-time nature and stream processing architecture, zero-party, first-party data that we're living in. And when I think of AI as a tool, AI becomes smart as you feed it information and data and events. And the more relevant and real-time the data is that you can feed it, the more it can make actionable decisions and work for you and have better outcomes in the near term, medium term, long term. And so it's exciting to me to see the evolution of our own AI investments.
And I know we'll get more into that with the coming questions. But to me, when I think about garbage in, garbage out, if you have a tool that maybe doesn't work in real time or isn't event-driven and where there's a lot of interoperability and lagginess in kind of the flow of information and you're not a fully integrated technology stack, you're more likely to get that garbage in, garbage out on the use of AI. So it's exciting for me to see kind of the evolution of our own AI in our product roadmap and its ability to actually provide really high-value results to our customers because of the strength of the underlying technology of our platform.
Yeah, that AI theme is going to be peppered in here just a little bit today for some strange reason. I think most of you are probably familiar with email marketing. If we talk about the marketing software space over the last 15, 20 years, the legacy vendors of ExactTarget, Eloqua, Responsys, Constant Contact, Mailchimp, the list goes on, would probably come up. I don't think everyone knows that Braze's heritage kind of started within the mobile. I know that was a big part of the thought process of when it started. But how do you think about kind of channel development today? And where is the growth coming from? Because you all had a stat in your Analyst Day in September that talked about, I think I got the number right, 29% of your customers have used five channels or more.
Five channels, yep.
I think the average person here thinks about like three channels.
Correct, yeah, yeah.
but there's five, let alone ten. Where's that growth coming from?
Yeah, yeah, yeah. So just to go back to an early part of the start of your question, while I think it was the advent of mobile and the explosion of mobile that gave our two technical co-founders conviction that there was a real investment opportunity and that Braze sort of had a future, we went live with email. So we had email from day one. So the, sorry, I just lost my train of thought.
We're talking about the number of channels.
Oh, yeah, yeah, so number of channels. So when you think about the number of channels, so we went live with email and then both in-product and other out-of-product channels. So when we think about the plethora of channels that are available to us now, there are some that I'll rattle off and they'll make perfect sense to you. Channels that we monetize, so email, SMS, global and U.S.-based channel there. WhatsApp, very global, also available in the U.S. We have a proprietary in-product messaging called Content Cards that's also monetized. So those are kind of the main monetized channels. LINE, which is a primarily Japanese channel that we've just brought online recently. And so that's the grouping of primarily priced premium message channels. In addition to that, though, there are a number of channels, so you're already up to five, right?
Even just that, you're already up to between four and five channels. Then SMS is going to evolve as RCS grows. And then you've got the in-product. So that'll be like a push notification, an in-product message, an in-app notification, a web push. So then those are currently not monetized and just sort of generally available with the platform. And so you can see you quickly get up to 10 channels , 11 channels , 12 channels. And our most sophisticated users are actually leveraging almost all of those in some way, shape, or form for the right type, whether it's in-product or out-of-product and whether it's sort of a push or a pull notification. Sometimes you need sort of two-way messaging ability, which is also supported. And so that is all part of the ever-growing number of channels.
I think that that channel list is just going to continue to grow, particularly as we expand into other parts of the world where there are kind of more niche channels that are relevant for those parts of the world.
So last time I checked, 12 is probably a bigger number than the three or four channels that I think most would know.
Yeah.
But as you start layering in additional channels, and I know this is part of the Braze theme, is I guess, first of all, how easy is it to layer on another channel?
Yeah.
As you think about the competition, how easy is it for them to layer on another channel?
Yeah. So like I said, we did the hard work right up front to start as a multi-channel, cross-channel platform. And so actually, we have a fairly efficient R&D flywheel as it relates to the way we've built the platform. And so it's relatively straightforward for us to add net new channels into the system and to do so in a way that does not add complexity to the users at the brand so they can easily access those net new channels. In addition, because of the way the fully vertically integrated stack was built, those channels are generally automatically featured with the relevant features that are available through the orchestration and personalization aspects of the technology stack. In addition, when we make R&D investments to add in net new feature and functionality, where relevant, those become automatically available to the various relevant channels where they might be useful.
The other aspect, and we'll probably talk a little bit more about this, but the evolution of our credits model is enabling us to essentially make multiple channels automatically available to customers the moment they buy the platform. So if a customer purchases no additional channels, no channels, they are automatically featured with all of the mobile and internet-based push notifications. So a customer could purchase the monthly active user entitlements that they need and not buy any extra premium channels and be automatically featured or automatically enabled with cross-channel messaging. With our flexible credits model, we are enabling customers to be automatically set up with a combination of SMS and WhatsApp, which dramatically lowers the activation energy for customers to actually engage in more experimentation.
One of the benefits of having 29% of our customers using five or more channels is at that level, the number of competitors that you have that can help you run a sophisticated cross-channel engagement marketing messaging campaign is virtually zero. So we're really excited to kind of continue to see that growth and the proliferation of channel usage because it really increases the overall stickiness of Braze as a platform.
Let's just skip to the credits question because it's a good follow-up. You announced the change around credits, I think it was May last year. I'm not looking at the calendar, but I think that was the right time frame.
That's right.
Kind of to your point, it allows a customer to say instead of buying, I don't know, 10 credits around this channel, maybe in this geographic region, you can use those 10 credits across whatever the channel is based on the pricing kind of how that channel works.
Correct, yep.
In SMS in the U.S. might be $0.01 and it might be $0.03 in Europe, but you buy $0.10 worth and you can do it wherever you want.
Correct.
Making up the example, obviously. But what have you seen so far? Has that vision of getting customers to try to experiment and try new channels, I guess, A, is that working? And then B, are you seeing their purchase velocity or purchase motion actually increase because of that?
Yeah. So we're not necessarily seeing right out of the gate that the volume of purchases is differentiated from what it might be if they were just buying the individual channels. I think a lot of times a brand will think through the dollars of investment that they think they need and then how they're going to allocate that to the various channels. So even with the credits model, I think in the first instance, they're going to think through, "Okay, well, overall, I have a wallet size of X, and so I'm going to allocate it to kind of different areas of the contract." So nothing material to raise. And of course, the revenue recognition associated with the credits model is identical to the old world.
So I wouldn't look at the credits model as sort of an automatic headwind or tailwind across any major dimension for the company in the near term. What it does do, though, is it can accelerate a sales cycle because now you have a customer that no longer has to sit through and think through the individual levels of purchase that they want to make by country. And instead, they can just think more holistically about the investment that they want to make. And then there's a back-ended exchange rate table that will sort of work down the credits based on their usage. Right now, we only have two channels that cut across the credits model.
So it's a little bit early days to sort of say, "Hey, here's a customer who was really only going to use SMS, but look, because of this, they have now started to use WhatsApp." I think we will see more of that as it expands to a broader set of channels. And certainly, we're getting positive feedback from customers who are excited that they're going to have the fungibility of the resource because there's nothing worse than a customer who has purchased a certain level of entitlements and then gets to the end of their contract term and realizes that not only have they not used it all, but it doesn't roll over and they can't repurpose it for something else.
That is going to eliminate that frictional element, which I think customers are excited about, salespeople are excited about, and just we generally at the organization are happy about.
Good. Let's talk about the new Braze Data Platform. If I go back to a dinner I think Chris and I had maybe about two years ago, three years ago, we were talking about the CDP space. And one of the answers are, "We're not a CDP," especially at that time. "We're not a CDP. People are fusing a little bit of the functionality with the CDP." But if we fast forward a couple of years, you all now have a Braze Data Platform. I guess, how does it differ from a traditional CDP that's out there? And are you seeing customers maybe use the Braze Data Platform instead of a CDP for similar functionality?
Yeah. So I think the advantage of the Braze Data Platform is it's all kind of integrated into the kind of vertically integrated tech stack, and therefore, our concern with the traditional kind of CDP world where data might flow into a CDP first and then get organized and then flow to Braze after is we can't really live in that world. For Braze to have the highest ROI, we need to be living in the stream and living in that sort of real-time data platform.
And so having the Braze Data Platform is almost a little bit more of kind of a repackaging of features and functionality that we have had over time and enabling you to take data from various data sources, Cloud Data Ingestion and other places, and kind of bring it all together and make sure that the optimization and the decision-making process that Braze is doing and the message triggering is as efficient, effective, and as high quality as possible. So it's not that we're trying to be a CDP as such, but I think it's helping to dispel some of the kind of confusion in the industry among customers who are like, "Wait, do I need to buy a CDP in order to use you guys?" The reality is the features and functionality that Braze has can help solve a lot of those issues.
It'll eliminate some of the potential lagginess and enable all of the data to continue to live in the stream by avoiding data having to flow into a separate CDP first and just sort of fully living within Braze.
Okay. I know back on the channels kind of side of the questions, you mentioned a few minutes ago you're kind of monetizing from a channel perspective, SMS and Braze with credits.
SMS and WhatsApp.
SMS and WhatsApp, sorry.
Yeah.
With the credits and whatnot, but WhatsApp is still relatively new to the company, at least compared to some of the other channels there. What are you seeing with adoption trends there? My guess is it's probably more popular in Asia-Pac, but outside of some geographic components. Anything else interesting to note there?
So it's more popular, I would say, outside the U.S. than in the U.S. Actually, Latin America, it's pretty popular in Latin America as well. And I think the adoption trends, it was certainly popular quickly. And so there was a lot of sort of excitement around it. So we've seen healthy growth in that channel generally, which has been great and helpful. And then we're excited about just continuing to grow that channel and to grow it globally and just have it part of the credits model. And it will remain as part of the credits model, and the credits model is just going to start to incorporate more and more channels. And so right now, really just covering that SMS and WhatsApp, but growing that to more and more over time. But no, we've been very happy with the uptake in WhatsApp.
And again, more popular, I think, outside the U.S. than within.
So coming back to the AI theme, I said there'd be at least a couple of questions here. You all have certainly pressed on this AI accelerator the last couple of years. But in terms of new feature functionality to meet this increased customer demand about just trying to use this technology at all, is there a couple of features or modules or something in particular on the functionality that customers are kind of gravitating to the most over the last year or two that's kind of new that would be interesting to call out?
Yeah. I mean, things like in conjunction with our catalog feature, the monetization of the item recommendation engine is something that we're excited about and customers are excited about. And so that's leveraging transformer technology and so the T in GPT, that transformer technology to essentially do the highest quality item recommendation to individual end users based on, again, where they are with their journey with the brand. So it sounds like a very retail-focused item, but it really isn't. It's anywhere where there's kind of a list of SKUs of products and services that a customer might be able to take advantage of and then a list of buyers or end users who might be ripe for those purchases or those interactions or those events.
So you could see it with any organization that has the concept of that cataloged list of SKUs and then wanting to kind of get the right recommendation out to their list of buyers and their list of end users. So that's kind of one of the newer AI feature functionality that's really leveraging some more sophisticated transformer technology.
Okay. Last thing I have is on product. I know you all recently announced a new product called Catalyst.
Catalyst.
You know, why don't you just give the audience a little bit of an overview of, A, what it does and maybe what initial customer reception has been?
Yeah. So Catalyst is not yet even in beta. So we're right at the moment, we're collecting customers and prospects that are interested in piloting the beta. But Catalyst is essentially a sophistication extension, and it's the natural evolution of a lot of the AI that we've been building into the tool over the last many years, but really bringing it together in a much more holistic and comprehensive way. And so you can imagine a world where before you would say, "Hey, I want to optimize for people looking at my messages.
So let's make sure as I generate this campaign, the AI is selecting the right channel to send to people and the AI is sending at the right time." And so those are optimizations that you could tell the system to do, and the system would kind of run some A/B tests and could kind of run that over time. But that's fairly narrow. And so now with the sort of levels of sophistication and increase in the AI technology, now you're at a point where through Catalyst, and so we're excited to start piloting this with customers, you can actually give in natural language an overarching goal. So I want to optimize for moving people down the funnel or conversions or upsells or limiting churn. And so you can give it a more overarching economic goal or activity goal and then set parameters.
There are discount codes that you can use, but you can't use more than X in a day, and they can't be bigger than Y, and they can't be concentrated in this area, and here are the number of channels that you can use, but you can't overload on this channel or you can't send more than this volume over here, so there's a way to kind of parameterize how the system is allowed to behave, but then once that is given to the system, now imagine the computer technology running multivariate tests simultaneously and continuously optimizing across a vast number of these different dimensions that it's been given access to using that real-time first-party and zero-party data.
That is really Catalyst is kind of bringing together a lot of the feature functionality and AI technology that we've been building over time, but creating a much more comprehensive capability where you can actually give the system a goal, and then it will self-optimize in a way that I think a human just doesn't have the capacity and the speed to operate at.
Can't wait till you get a beta customer or two.
Yeah.
So we can maybe find them and chat with them.
Yeah.
That's awesome. Moving to go-to-markets, at least moving off product and going to go-to-market. You all have talked about how marketing budgets the last couple of years have been tight. We see that across a lot of our coverage in the marketing software space just kind of generically. So I don't think you're not unusual in that. But are you seeing any green shoots or opportunities as we start to move into calendar 2025? Because software recessions, at least product recessions historically, they tend to only last two to two and a half years. So we're kind of at that point already.
Yeah. So I think we're excited by kind of the direction of travel of the pipeline. So the pipeline is looking healthy. And also, it used to be a couple of quarters and a year or two ago, the pipeline was behaving in ways that was very erratic. That has ebbed as well. So you're right, marketing budgets are a little bit smaller, but they're more known. And so there's a little bit less kind of disruption that we're seeing out there. So I think we're excited about the continued globalization of our overall footprint. We're putting a few boots on the ground and a few new locations that we've talked about. And we have some new folks that we've put in São Paulo and in other parts of the world where we're seeing some good traction.
So I think that's exciting to see kind of the ongoing evolution of the globalization. I think we are excited. There's been a decent amount of pain at the enterprise level back when the market was really falling apart and seeing some of that start to recover and excited for kind of more of that, excited for our sales team to kind of improve and increase its productivity. I think we still have some capacity absorption there for next year. Our fiscal year is about to end here on January 31st. So I think there's between the free trial motion that we have launched in the recent period, the credits model that will help to kind of accelerate some of the sales cycles, some of the new boots on the ground, and some new parts of the world.
I think the combination of those things and then continued strength in certain pockets, particularly in Europe, we're excited about that for next year.
I got like four minutes and 33 seconds left, and I have a lot of questions. So let's get to that free trial question. It's kind of a newer motion for you that you recently rolled out.
Another question I get there often is, "Hey, doesn't that help the company focus more down market on smaller customers, which is not necessarily a segment you're moving away from, but you've de-emphasized that maybe the last couple of years?" How do we think about what that opportunity really drives for?
Yeah. So a couple of things. So the free trial is not necessarily something brand new that we just sort of came up one day and said, "Hey, it's time to do a free trial." It's really an extension of something that we've always done and that was really more relevant for the enterprise. And so we've always done or we've had for a long time the concept of a sandbox and really the ability for some of these more sophisticated and tech-forward thinking enterprises to really get their hands on the product. And so here, the free trial motion is also available really for the enterprise. And if it's helpful down at the SMB level for a lower touch sales cycle, that's also where we're happy for that as well. But it's not necessarily kind of the reason that it was done.
I think it's going to be very helpful at the enterprise level to just get their hands on the product and really negate some of the need for a much higher-touch proof of concept and heavyweight demos and the like throughout the sales process.
Last question in go-to-market is at your Analyst Day, you also talked a lot about engaging and leveraging kind of the GSIs and that partner environment more. Before I ask the question on that, I think what's been interesting in my coverage of you all and the other experiences in the space is what your technology partners like to say about you. You have some fantastic technology partners. Actually, one of my best checks in the space has been going to the Movable Ink conference every year. You guys are paired pretty tightly with them, I know. But the types of customers there and the management's vision about where the space is going, they all believe the exact same thing that you do. So this is where I think the vision of the space is going next couple, three, four, five years, which obviously you all believe.
But how do you get the GSIs to piggyback that same vision? Because historically, they like to sell the ExactTarget, the Adobe, the big platforms of the world. And at some point, they change their mind because they feel that's not the best technology.
Yeah.
But kind of, where are you with that to capture a pretty impactful system?
Yeah. Yeah. So I think everything you've said is sort of broadly directionally correct, and I think with the GSIs, I think it's taken us a while to sort of get enough attention from them. So, in the last, call it 24 months, probably achieved a level of size and scale and scope and relevance that we were finally able to kind of get their attention. There's also been a number of kind of a small number, but some sort of smaller organizations that have popped up that are out there actually trying to compete with the GSI. I mean, they're much, much smaller, but they're building businesses entirely on the back of Braze services, and so I think the GSIs are continuing to invest in building out their Braze practices. We're seeing a continued increase in the number of Braze certifications at these organizations.
We are continuing to focus on figuring out different services, so we are not destined to be a services organization, and actually, if you look at our financials, we have low single digits, 4% or so of our business, of our revenue is services, but we are continuously looking for ways to kind of incubate new services that we can then sort of say to the GSIs here, like, "This is the type of stuff that you should be doing on the back of Braze," because you're right, they've built these huge practices over time to service kind of the legacy clouds, and the reality is the legacy clouds need a lot of humans to kind of keep that going.
And so our more modern and sophisticated technology stack maybe doesn't require as many humans, but actually opens up the door to much more sophisticated data transformation and business transformation that actually you can't do if you're stuck in the old legacy world of the legacy marketing clouds. And so we're continuing to work with the GSIs to sort of see that vision and starting to see the traction where they really do believe that Braze is the next kind of generation technology, and they want to be seen as thought leaders. And so it is incumbent on them to be recommending Braze over some of the legacy technology because that's really the right place to be from an advisory perspective.
Two last quick ones for me. We have the CFO here. We have to at least ask a financial question or two, right? I know the question you probably get asked more often is around net revenue retention. You gave a significant amount of disclosure at the Analyst Day. Your non-COVID cohorts renewing at a rate of, I believe it's nine points higher than that COVID cohort.
That was the Q2 number, yeah.
That was the Q2 number.
Yep, that's right.
I think my watch says 2025, which means... is my mic dying? Oh, sorry. We're in 2025, which means we're about probably three years towards the end of that cohort.
Yeah.
Or at least close. Your average duration is about two years in most of your contracts. Any updated thoughts on when we kind of get through what that churn looks like?
Yeah. So look, we gave a number for the end of the year for Q4, trailing 12-month statistic of 110. And we talked about that at Analyst Day. And I have a high degree of confidence on that number, so no issues there. I think what I think people should know and understand, and we've started to kind of be transparent on this, is the revenue inflection of kind of revenue growth is likely to actually precede the inflection of dollar-based net retention given the 12-month trailing statistic on dollar-based net retention. So we're at the point now where we will likely, we're counting the distance to an inflection on dollar-based net retention. We're counting it in quarters and not years. So I think we're starting to kind of see that the path to that number starting to settle and then inflect.
But I would look for revenue growth to inflect prior to hitting the bottom on dollar-based net retention.
Okay. The last one for me is really more of a gross margin question at the end of the day. We know gross margins on your SMS business are lower because of the carrier fees generally. But as I think about, I'm an Android user. As I think about my kids that are Apple users and my friends, is almost all of their text messages have converted from your legacy SMS to RCS going forward. The charges with that are significantly lower because it's more of a data infrastructure than a cellular charge. How do you think about your customers maybe changing that message type and sending more RCS messages versus SMS? Could that actually be a tailwind for gross margins, do you think, over a period of time?
Yeah. I mean, I think there's going to be puts and takes across to sort of isolate it to kind of like one message channel and to talk about that. And the direction for gross margins, I think, is missing the vast array of puts and takes on gross margins generally. So when we think about gross margins, it's broadly split about 30% humans and 70% technology. So the human component, we've been actually moving people or not moving people, but hiring people in differentiated cost structures around the world. So Bucharest and São Paulo are kind of the two big new ones where there's obviously a bigger savings. On the technology side, I think part of what you're talking about in terms of that SMS piece narrowly could be a tailwind.
But we are also starting to sell channels like LINE in a fashion that actually takes us out of the middleman. So LINE is that Japanese channel where we are not acting as like a white label reseller. The customer actually goes to LINE directly, buys the volumes, and then we sell a relatively high margin connector that enables the triggering of those messages. And that is a motion that we may start to see more of across some net new channels and potentially some SMS, some global SMS regions as well. And that is going to mix into the overall margin profile. We are currently operating at the bottom end of our long-term range, and we're not concerned about the direction of travel within that. So I would look for us to kind of act within that long-term range until we update the range.
Fair enough. Well, with that, happy to take any questions from the audience before I lose my voice. Yes.
At the Analyst Day, you talked a little bit about operating leverage. There's some room there. You talked about one or two things besides the one you just spoke about of gross margin for an operating line that can drive that leverage. And then just kind of anything about growth versus profitability.
Yeah. Yeah. So I don't know if we have to repeat the question.
Yeah. The question's about operating leverage. No, it's fine. You don't have a mic, unfortunately.
Okay.
I'm not going to throw it. It's not a good day to throw a mic. It's just kind of where that's coming from generically.
Yeah. Yeah. So at Analyst day, we did actually set forth a framework of different revenue growth scenarios and then the associated expansion that we would expect to deliver on an operating income, non-GAAP operating income margin perspective. So when we think about, so we talked about gross margin, the other three categories, so I'll just kind of go through them. So R&D as a percent of revenue, we're kind of operating at very close to our long-term range. So I would expect that from an investment perspective, think of that line item as tracking most closely with revenue growth. On the G&A side, we're operating about 14-ish% as a percent of revenue. Ultimate goal is closer to nine. So there's kind of five-ish percentage points there. It's not a huge amount of dollars because it's just a smaller percentage of revenue.
That's as we continue to scale and invest in these cost-optimized various locations, strategic locations around the world, that we're going to see that scale continue to benefit us and that cost structure continue to benefit us. The place where we really need to get the highest level of leverage improvement is in sales and marketing. So there is an element of benefit where we are going to see enhancements from these strategic locations. We are certainly putting some marketing folks in that in some part of the world. There are obviously sales folks that are in São Paulo and in other regions. So there's definitely a globalization factor that should be helpful from a cost structure perspective.
And then we are still at a point where there's a little bit more productivity improvements that we should be getting out of our sales population, and that should help to drive that down. The other pieces are from a revenue benefit perspective. We are still experiencing levels of churn that are more elevated than where we think they should be on a sustained basis. And as that continues to improve, again, a more efficient motion from a sales perspective and a more efficient dollars of revenue per resource. If you plug up the leaky bucket, you're in a much better position. So I think those factors all kind of working together will drive the evolution in our non-GAAP operating income that we've committed to across that framework of different revenue growth scenarios.
We probably have time for one last quick question if there is one.