Couchbase, Inc. (BASE) Earnings Call Transcript & Summary

All right, everyone. We can go ahead and get started. I'm Austin Dietz, one of the software analysts here at UBS. We're thrilled to have Greg Henry, Couchbase's CFO, on stage with us. Greg, thanks for being here.

Thanks for having me. Appreciate it. First time. I love it so far.

Fantastic. And as a reminder, you can send in your questions via the QR codes on the table, and we'll get to them at the end. But with that said, let's go ahead and get started. Greg, the database market's one of the larger more strategic software markets out there. As a result, that's invited a number of entrants into it. So how should we think about Couchbase's opportunity given the number of database providers out there?

Yes. So you're right, Austin. It's a massive market. It's $140 billion, $150 billion. It's a market that we don't have a route to really concern ourselves about because it's so big and it's basically split between half of it's on the analytical side and half is on the operational side where Couchbase plays. And look, the reality is there's been a very nice shift in the market from relational to nonrelational over the last decade or so. But still the majority of the market sits on the relational side. And so there's amazing tens of billions of opportunity on the relational side that we can go after as well as the majority of the new workloads that are being built are going to a nonrelational technology. So where we sit at the center of this is we've been built and architected to be the most scalable performance database with caching and cloud edge capabilities. And so that really opens up a very nice swath of the market that we can go after. We always say we're not really market-constrained. Demand is good. This is about executing on our strategy and delivering on that.

Yes, fantastic. Could you maybe touch on how you're similar to but more importantly different than MongoDB? It's a name that some are familiar with in the room. That would be great.

Yes. We obviously compete against Mongo. They've done a great job. We give them all the credit in the world for doing that. We are both JSON document store databases. I think where we differentiate is the architectural -- underlying architectural differences. So we are a shared nothing scale-out, memory first architecture. So scale-out is similar between both us and Mongo. The memory first is the caching capability. So we are caching technology as well as a JSON document store database. They are JSON only. And then the shared nothing is the architecture of how each node within our underlying database architecture is sort of equally intelligent. And so if one node is going to fail or run out of room or whatever, they equally know how to go across each other, which drives low latency. Mongo is more of a, what we'll call a leader-follower. So there's one leader node that's instructing all the followers so that it generates some latency there. And then the other difference, I would say, is Mongo recently announced they were shutting down their Realm, their mobile business. We were the first to market with mobile technology in 2014. It's built natively on the platform and all the capabilities that are there with our on-prem or cloud capabilities also go to the edge now.

That's great. And what about the cloud-native databases from hyperscalers? Could you maybe talk about the relative positioning of your own cloud-based, database Capella, relative to some of offerings there? And what types of applications are you best suited to underpin?

Yes. So the hyperscalers have very nice sized database businesses, and they have a variety of databases, which just shows that the need -- no one can sort of solve everything with a singular database. Our moat or our defensibility against the cloud vendors obviously is there is no vendor lock-in. When you use Couchbase, you can use it across any cloud you want. In fact, when you buy Capella, you buy a Capella credit, it doesn't say AWS or Azure or Google. You can use it wherever you want. So that and then the cloud vendors don't have any mobile technology as well. So edge. Where we compete against them is they are what we would consider a good technology. But for the mission-critical applications that we serve in the large enterprises, good enough isn't good enough. They need to have performance and scale that allows -- because a lot of these applications that they are building are mission-critical, they're running their business, they're running revenue and they're outward customer-facing and they just cannot afford any latency, downtime, et cetera.

As long as I've got you, over the last 6 to 9 months, have you noticed any change in the competitive intensity of that database market with the hyperscalers? Anything you'd flag or call out?

No, nothing in particular. I think it's still sort of the same competitive environment. I think for us, every quarter we go, we get further into Capella, which is really being helpful for us in terms of it's a way to have a buy-from, self-serve motion from go-to-market. So it's helping on our efficiency. And then also Capella, now that we have a perpetual-free tier, allows people ease of use and need of access to Capella and to Couchbase that they could never have had before. Well, it couldn't have -- it was much more difficult before. So making it a lot more accessible for developers that maybe hadn't sort of tried and played with Couchbase.

Yes. Yes. So while we're on Capella, it now represents 15% of your ARR mix today, up nicely last night. And you've talked about Capella reaching 30% to 50% of that total ARR mix over the coming years. So what initiatives are you pursuing to drive more Capella adoption?

Yes. So obviously, we've highly incentivized our field teams to go after Capella, particularly with new customers, but also with existing customers. And that's where we're going to get the -- so 15% ARR, 33% of the customers. Obviously, we're landing small with new customers, but the big dollars are coming from the migrations of our existing customers. And so we still have 85% of our business sitting on the enterprise model. And so we're spending a lot of time internally with sales. We have Scott Anderson, who leads sort of growth in operations focused on migrations and working with some of our larger customers on building plants and migration plans. So that's one of the biggest drivers that's going to not only help the Capella penetration but also help growth.

And Greg, what would need to happen for Capella to hit that 50% of ARR mark sooner than anticipated? So maybe more towards the 4-year end of that time frame relative to the 6-year end?

Yes. It's all going to be driven by migrations. We talked about this past quarter yesterday, last night, we talked about we had our largest single migration in company history, a multimillion-dollar account that was on enterprise is now migrated. That customer has effectively doubled and they're going to migrate over a couple of more years here. And by the time they're done, there'll probably be triple or more of what they were on the enterprise. So you can just see how in a very short period of time, you can take a really nice account and make it a top 25 account within our installed base.

Yes. Maybe just to tease out that migration you just brought up. Was there a specific trigger? Or what was the catalyst that drove such a large migration to Capella?

Yes, a lot of it was -- look, a lot of our customers will come to us and tell us like we are not in the business of running and managing databases. So the fact that you all can do it. And when you do the TCO analysis and you say, well, here's how much you're paying for Couchbase today, then layer on the infrastructure, then layer all the resourcing and consulting and everything, the TCO becomes pretty compelling. And the fact that then they don't have to manage all these people and consultants and buying infrastructure, it just makes it easier. It's the same reason why all of us have gone to the cloud from being -- having our own data centers, they'll all take care of it. We don't have to worry about it. It's the same concept. So it was TCO, but it was also like, "Hey, I don't have to now worry about managing the database. Couchbase will handle that. I can focus on the core business that I'm trying to run here."

Yes, that makes sense. And it ties into your medium-term target. You talked about a 20% plus ARR and revenue growth medium-term target. So could you talk about the construction of that 20%-plus growth outlook? And how much of that depends on Capella's success?

Yes. A lot of it depends on Capella's success because the way we've articulated is if you have $1 of enterprise today in Couchbase and you move to Capella, it becomes $1.50 to $2. I told you that example I just gave you where they basically doubled, so that's what's going to be the driver of that. So that's where a lot of the growth is. So the further we get into this penetration of migrating customers to Capella, that's what's going to start showing some of that accelerated growth. And you will see that in ARR before you see it in revenue because in this case, as we migrated this customer over, we picked up all that nice revenue, but -- or the ARR, but the revenue is on a consumption basis. So that will come in over time. So the revenue will trail. As I've told many investors and anyone I can talk to, my goal and our goal is to keep the growth rate of ARR ahead of revenue. That will all come behind it.

Yes. Well, maybe we stick on -- stick to ARR here. And as a follow-up to that, you obviously reported 3Q last night. You reported 16% ARR growth in constant currency. You're guiding to 17% ARR growth in 4Q. So could you talk about what you would need to see for ARR growth to return to that 20% plus level relative to the 16% to 17% range you're in right now?

Yes. This year, for Couchbase, our fiscal '25 was sort of a different year in the sense that on the enterprise side of the business, we tend to expand at the point of renewal. Our renewal pool, just based on how the timing of contracts works, was down versus last year. I articulated last night that next year it goes back up. And this year, it was way more heavily weighted to the second half and particularly Q4. Next year, it's much more balanced. So as I look at Q4 and particularly the first half of next year, the opportunity we have to grow the existing installed base that's in enterprise is greater and the opportunity, every quarter that goes by, there's more opportunities to continue to migrate customers. And so that's what gives us the ability and confidence that we can return to being a 20-plus percent grower.

Yes, fantastic.

And I can just -- one other thing, look, we are not running this company to be a high-teens grower. We are 100% focused on driving growth and getting back to being a 20-plus percent grower.

Yes. That's great. And yes, maybe just to your point on the renewal pool, perhaps you could elaborate a little bit more. So next year, I think you talked about it being larger. Is there any way to think about sort of how much larger and then just evenly distributed? So how are you guys going about that opportunity? Maybe you could speak to your focus there.

Yes. It all depends on the timing of the contracts, right, of when they're next up for renewals. So we typically do 1- and 3-year contracts, so our duration is kind of typically around 18 months. So it just so happens that we have these years where we have peak years and then we have trough years. Again, next year, it's a bigger renewal pool. It's modestly bigger, so I don't want to give the impression that it's going to be double, but it's certainly nicely going to be up next year. And again, that's all based on where we are today. I have to ultimately see in Q4 what happens with particularly as the sales force looks at what's in Q1 and whether they want to pull things forward because they're in accelerators and things like that. So that gives us that opportunity. And again, we're just much more progressed with a larger pool of big customers on either renewing and upselling or migrating to Capella. So it just -- there is, I'd say, better momentum in things we see at the company that I think a lot of folks aren't necessarily seeing through the financial results of the company to date.

Yes. Yes. So I'd love to ask you a little bit more specifically about the 4Q ARR guide before we get to next year, there's obviously 4Q because it implies a healthy step-up in the net new ARR dollars you'll add in 4Q. And you talked about having some contribution from a precontracted amount of ARR that will step up. So that's great. But then you've also talked about these larger strategic deals that you have in the pipeline. So first, could you provide more color on these large strategic deals slated for Q4? How do they compare to your average deal size? How many are we talking about? More color on those deals I think we'd all love to hear about.

Yes, we started the second half, and we had a pool of, call it, 20 large strategic deals. Obviously, we closed several of those in Q3 and then the larger amount sits in Q4. Now some of them have renewal points, right, so there's compelling events. Other ones are customers that have actually come to us and said, "Hey, we need to like to do something." They might be a Q1 customer like to do something early. We're working with a large financial institution that's come to us and said, "Hey, we'd like to" -- they're already a nice 7-figure customer, but, "we like the technology, and we like to go a little bit bigger." They don't have what I would call a compelling event like a renewal point, but they have come to us. So it just is a sizable pool of deals that allow us the sort of comfort and confidence that we'll get enough of that done to deliver. And we understand that we've delivered about $16 million of ARR this year so far through 3 quarters and the midpoint of the guide for Q4 is $18 million. So we understand why people are saying, "Well, you're going to do more than half of the year in Q4." And we understand why there's questions around it, but we obviously get the benefit of seeing what's supposed to happen, the precontracted ARR component. And so what we actually have to build in quarter is well within the range of what we've historically done.

Yes. I mean it's a fair point. And then just also given the size of these deals, oftentimes, there can be more hands in the pot, more sign-off required. So I guess just given the nature of these more strategic deals, can you talk about your confidence level in pipeline conversion in this all-important fourth quarter here in closing the number of deals you need to hit to hit the 4Q ARR guide?

Yes. So we obviously have a larger pool of deals that are in progress and in flight and dollars that are associated with them that would get us well beyond where we've guided. So we obviously have derisked that we're not going to get all of them because that doesn't really -- or rarely happens. And so we feel like we've derisked that enough. We have enough that have compelling events. And again, the other thing that happens is in Q4, because we're at January 31, a lot of our customers are 12/31. And so budgets reopen on January 1 and they have more certainty and clarity about what budget they're going to have and it tends to lead to getting some deals done in January that where they have budget and our year is still open.

And was that a -- because you made this comment last night on the call that you might see some of those deals pull forward from Q1 into Q4. Is that the primary driver? Or...

No, We don't plan on that. Again, we have a modest amount of assumed sort of early renewal activity coming in. But we do -- we have that every year. And so all I say is that there's always a greater opportunity because of the budget comment and the fact that our sales team, a number of them will be in accelerators and them doing the deal in January, we'll pay them out exponentially more than if they do the deal in February. So they are incentivized to then do it. It's a matter of getting the customer to want to do it as well. But for a lot of our customers because they're on calendar, during January and February, it's all Q1 for them. So this isn't a year/quarter dynamic like it is for us.

Yes. That makes sense. And then last night, you had talked about -- you had obviously brought up earlier, the largest Capella migrations to date. And I think there was reference to having line of sight to several more of these larger Capella migrations. So how does -- from a timing perspective, is that relevant for this Q4 ARR discussion? Or is that comment was more referenced over the coming 12 to 24 months, you could see more of those migrations play out?

Yes. I think it's more over the long term because, again, it's hard to predict the timing of some of those. But look, we're having conversations. We talked about on our earnings call for Q4 last year, we signed our largest deal in company history. And at the time, we were having discussions that, hey, what about some Capella, put some Capella. They're like, no, no, no, don't put any in there. We don't want to talk about it. And here we are 9 months later, and we're having meaningful conversations about Capella. Like we want pricing. We want to go there. I'm not sure how big we want to do it, but we want to start. And so these things can change very quickly, where there was nothing, and now there's something, and this is all incremental on top of the running deal they have. So we are in dialogue with our largest customers on Capella. This year, in particular on the go-to-market side, we carved out a strategic team. So we, for the first time, have really hunter-farmer model sort of evolving and we looked at our -- that team looks after our 25 to 30 largest customers. So they have no new logo target. They're only renewing, upselling and migrating. And we've seen the real benefit of that. And I think we'll probably look to move next year to have even more than the top 25 or 30 and expand that because that's where we get our biggest sort of return, if you will, from an ARR perspective.

And when did that hunter-farmer model go into effect?

Just this fiscal year.

Just this fiscal year.

Yes. I mean, you can imagine that all of our 25 or 30 top customers, not all of them are even up for renewal this year. So it's just building, but that focus has really helped us and we've seen the benefits of it. And that's why when we were planning for fiscal '26 with our CRO, Huw, we're having conversations about do we expand that team and focus more on, say, the top 75 or 100 because we're seeing that, and that makes up such a big part of our ARR base and when we're talking about these large strategic deals, is that a customer pool we're dealing with.

Yes. Yes, that makes sense. And what about the new logo piece, Greg? I think you had added around 34 new logos, last night, healthy, but not blowout. So I guess, what are you doing to drive greater new logos coming on the Couchbase's platform? And where are we in just that evolution of that new logo machine or [ coming to ] fruition?

Well, we're actually pleased with the new logo performance this year, particularly in comparison to last year. So we're starting to see where we're getting 30-plus new logos a quarter or more like we were teens or low 20s, and we got a quarter when we did over 60 this year. So we're seeing -- sort of start seeing that benefit of Capella. But again, it goes back to that go-to-market because these -- this group of sellers is now focused on strategic with no new logos. When you're dealing with these large enterprise customers, it takes a lot of time. And that's where a lot of the ARR dollars can be derived from. So a salesperson's mentality is that I can get most of what I need from this large renewal and upsell versus going after new logos. Now that we've taken those off, they're dealing with more modest customer sizes and have more capacity to do new logos. So we think it's benefiting for both getting more dollars out of those top strategic customers and getting a better new logo performance.

Yes. And how should we think about there's now a free tier out there as well? That was, I think, a fairly recent introduction. So maybe you could elaborate a little bit more on that and its potential, that free tier, to drive new logos with time? How do you see that playing out? Are there specific triggers where a customer then tips over but -- yes, to a paid seat? Or if you could talk about that opportunity, that would be great.

Yes. So when we launched Capella, we had 30-day free trial, so you had free, but it was only for 30 days. So it was probably interesting for developers knowing that they had a 1-month time frame. And now they have, I can go put it out there. If I don't get back to it for a month, it's still going to be out there. And so I think that just opened up the aperture for more people who want to jump in and try it because they don't have a time associated with it. It obviously gives us more time to look at who's using it, what ways they're using it. So between the free trials and what we see in the marketplaces from an on-demand perspective, it just creates a whole new pool of pipeline opportunities for us to go after. And the fact that it's not just sourced pipeline, but we can actually see what they're doing, and we can have interactions with them and it just helps us, again, sort of broaden the pipeline but have it sort of more mature than just sort of the first touch point.

And I think going -- you had referenced the strong new logo number in 2Q. I think there might have been some ISV or partner involvement in that. So how built out is that partner ecosystem for you guys? And could that be an additional driver of new logos over time?

Yes. We've been working on this partner motion probably for 8 years now. When Matt, our CEO, joined, he'd come from Cisco and he saw the power of the partner program there, and he immediately brought somebody in to lead it. So we feel like it's well built out, and it's contributing nicely. I mean we pushed this hard to the sales org about partner involvement. So we have a pretty high percentage of our deals that are partner sourced or influenced. That said, I think we can get better and continue to expand it. We've definitely seen in our sort of non-U.S. regions where they've really embraced that and adopted that. And that's where we're seeing some of that really nice activity, that ISV was outside the U.S. So that's where we have really nice opportunities.

Okay. That's great. Yes. I mean maybe just to step back, a broader question for you, Greg. I think investors had all hoped that AI would lead to a pull-through of the data stack layer as enterprises got their data states in order. And a couple of years into this, we haven't really seen any signs of a lift in the results. So why is that? And will AI lead to a pull-through of the data stack at some point?

Yes. I think the AI applications are still being developed. I always sort of try to sense out besides things like chat and LLMs, what are the AI applications that any of us are experiencing today, and there's not that many out there, quite frankly. And so I think they're still building and that's all to come, but it will happen. And I do think that AI will play a huge role. We've already seen it with our Copilot, with Columnar analytics, the AI services we just announced a couple of days ago, we think that's going to be really important. And the more we can get people on to Capella, that's just another opportunity for them to use, engage and ultimately consume on the Couchbase platform.

Yes, that's great. Yes. And maybe just on AI. To your point, there's only so many applications out there, but you guys are certainly a part of that modern NoSQL database layer and modern AI stack, where, yes, we just love to hear some about the early AI use cases that Couchbase -- that you're seeing on the platform, that you're being pulled along for. Yes, so may we start there. And then timing-wise, I'd be curious to hear when enterprises could put more of these AI applications in the production.

Yes. I mean the 2 things that we really see is through Vector Search and the Couchbase Columnar, which is a real-time analytics. So we're starting to see customers really use this. Again, I think the AI services is going to really help this along because with that, you can sort of within your own data set, create your own model, have sort of auto vectoring. So it's going to make it sort of easier and simpler to use, so you're not vectoring across your whole data set, you can create a subset because it is it is sort of heavy to do it across the whole dataset versus a micro set. So we think that's going to really help accelerate the use of some of that. And we're absolutely starting to see that both on the -- like I said, the analytics side and the -- and vectoring. But again, I think the AI services for us is going to enable that as it gets more cost efficient and simpler to use.

Yes. I mean it's a good segue to like has there been any early read on how the consumption profile of these AI applications, how they might compare to traditional applications that are in Couchbase's platform? Or is it too early to say?

No, I think it's going to be much -- again, every time we add a capability or a tool set to platform within Capella, it's just -- we don't sell it separately, right, because we still sell credits, but it just gives customers more opportunities to use the Couchbase platform and ultimately consume. So that's how we're going to monetize that versus actually selling it as a service. It's just going to go through consumption. But the more we put out there and the more we give the customers the opportunity to take advantage of the platform in new and different ways, that's going to drive the consumption, which is why we want to continue to push customers over to Capella because that's where we're going to get not only the initial uplift as you move into Capella but the opportunity to use much more than you probably would have before.

Yes. That makes sense. And then you guys just announced Capella AI Services to streamline the development of a GenAI application. So can you talk a little bit more about this latest offering and what you're seeing?

Yes, it's just going to be -- there's 5 new capabilities that we added into this, but it's going to allow you to create sub-models from your data set. It's going to sort of allow you then to auto-vectorize that subset of data. It's going to have agentic capabilities to do things that the human would have normally done. And so it's just going to help with development, help with the capabilities of what's going to come out of the database to enhance the application. So we are extremely excited. And what's amazing is this AI services concept was just born a year ago, and here we are putting it into private preview a year later. So Matt talked about on the call yesterday how our engineering team is rolling and the pace of innovation is faster than ever at Couchbase. This is a perfect example of this, where it took -- it went from, "Hey, we got to do something more on AI" a year ago to here we are with launching AI services.

Yes. Does it feel like the potential uptake of these agentic AI application, could it be faster over the next 12 to 24 months than the pace of adoption of what we've seen over the last 12 months of more Vector Search, RAG-related use cases? How do you think about this sort of next chapter?

Yes, absolutely. I mean, one of the feedback we're hearing, particularly on Vector, and it wasn't just Couchbase, it was others around, it was costly to do the vectorization. And so that's why we want to do with these AI services is so that you can, again, create a sort of a sub-model or a smaller data set to vectorize off of, which makes it much more efficient from a cost perspective. So I do think that's going to aid in the ability for things to move much further and faster. And the reality is we're just another year into it. I would imagine if we're back here in a year that we will have actually seen a multiple of the momentum of what we've seen in the last year.

Yes. And then, Greg, this new offering -- one of your new offerings, columnar service, looks interesting. We're seeing this more broadly in the industry where it's the convergence of both operational data with analytical data. We're seeing it from Microsoft, we're seeing it from Snowflake, yourself. So I guess what's going on in the industry? What's happening that we're seeing a greater conversions between these 2? And maybe you could talk about the monetization of that service as well. Would Columnar add net new applications to Couchbase's platform? Is this about driving greater usage and expansion of existing applications? How do you think about that piece?

Yes. So to answer your second question, yes, it's all about usage and consumption and giving customers the ability to have tools and capabilities that will drive them to consume more. When I started out the conversation, we talked about the TAM, there was the operational and the transactional side -- sorry, the analytical and transactional side of things. There is a crossover right in the middle, which we call translytical. And there is the -- again, we don't see an operational database entering the full analytical market and vice versa, but there is this overlap where there's a need for analytics and there isn't the capability or the time to move it offline into a batch process. It has to be real-time run-time analytics. And that's where Columnar comes in and allows you to do that, where you don't have the time to go offline. You want to do that when you have the time and ability to sort of set that data set aside and run the analytics. But many customers, particularly if you think about retail, if you think about travel and hospitality, they are trying to do analytics real-time to get closer to their customer, be more responsive to their needs and that's what Columnar is going to help do.

Fantastic. Great. In the last 30 seconds, one from the audience. One of the factors weighing on the 4Q revs guide is the consumption rev rec at Capella. As these Capella migrations ramp next year, could we begin to see a greater divergence between revenue and ARR growth?

Yes, you could, yes. Because again, like, let's take that customer that just migrated this past quarter. Had that been an enterprise deal, we would have picked up $0.5 million of revenue, and we picked up next to nothing because they are only starting their consumption journey. So you get all the lift of the ARR and effectively no revenue. So we actually went backwards on revenue, but materially forwards on ARR. It will all catch up eventually. So that's why we don't get concerned about that. We are more focused on ARR. And if we keep that ARR growth at the right level, all the revenue will come behind it.

That makes sense. We're out of time. Greg, thanks for being here, and thanks, everyone, today.

Thank you, Austin. Appreciate it.