MongoDB, Inc. ($MDB)

Okay. Great. Thanks, everyone, for joining us. For those of you who don't know me, I'm Rudy Kessinger. I cover security and infrastructure software at D.A. Davidson. With me today, CFO of MongoDB, Mike Berry. Mike, thank you for joining us.

Thanks for having us, Rudy.

Yes, for sure. We'll make best use of the time, jump right into it. Mike, I think the biggest debate and topic that I talk about with investors is just is MongoDB an AI winner? And if so, what does that look like? When do they start maybe to see material benefit? How are they positioned, et cetera? So let's start there. Mongo, what's it look like ahead for you guys with AI?

Yes. Thanks for the question. Look, we certainly view ourselves as an AI winner, and we've started to talk about it on the last call in terms of seeing some movement in the enterprise. So let's take a step back. Why we think we are? We think our product is perfectly suited for the AI world. Back when the founders built Mongo, good or bad, the way they built it was, we think, exactly what we need today. So you think about it, document model, built-in JSON native JSON built in. You don't have to do plug-ins or anything. Scales horizontally, so it's much more economical, flexible schema, which is really good for unstructured data. Then you add the stuff we've done in the platform around vector search, around search. So that's all in the platform. You don't have to bolt on any other products. And then I think a couple of big things that people -- we talk about it, we need to talk about it more is the multi-cloud capability, not only on-prem, but you can run Atlas not only in the individual clouds, you can run the same workload at the same time across all 3 clouds. And now with all the economics as well as the geopolitical situation, multi-cloud is becoming a big piece. And we've talked more about -- it's also a great place as they deploy agents for the longer-term memory to start storing that data. So we feel really good about the product. We've talked about, hey, not only the frontier labs, AI natives and then enterprise, we think it's only a matter of time, and we're starting to see some movement in enterprise. That's going to take a while because they have reputational risk. They have governance, they have security. But as they start to deploy that, we think that we're in a really good spot.

Yes. So you guys have talked about probably being downstream from the likes of Snowflake and others. And some investors, they hear that. They ask why they don't really maybe understand the nuances and the differences of where they sit, what they're doing, where you sit, what you're doing and how it's going to take real enterprise consumer-facing workloads. So -- but just expand on that. I guess, why are you downstream? And how long might it take until we start to see more of this enterprise AI consumer-facing workloads?

Sure. So if you look at where enterprises have deployed, everybody has deployed almost everybody internally. And it's the use cases that they're using around running their business. Certainly, you've seen everything with Claude and code, but also internally in terms of marketing, finance, other areas, and that's very well suited to where they sit in terms of using a data warehouse for that. If you have marketing people accessing data, great, finance people, you can control that. And what we've seen is a lot of deployment internally. Everyone is dealing with the cost issue, separate discussion. And so that's why we think that they've seen that bump. For us, this is really -- if you take that data warehouse, now you take where we are in terms of the database, real-time data, that's where we think they're going to deploy customer-facing enterprise workloads at scale, and that's where we fit in much better. So again, we feel very good in terms of that movement. And why do we say that? We've seen a much bigger uptick of our platform, especially vector that we know is related to AI workloads. Voyage is starting to actually have a material impact in terms of customer adds around embeddings, and we're starting to see more POCs. And again, it's tough for us to see the activity in the clusters because they may be putting agents on top of enterprise data. For us, it's just another workload, and that's what we see. So as it relates to where they are and where we sit, yes, there's some overlap, but that's why we feel good about, again, as they move for enterprise deployment that, that is where we're going to really see growth.

Yes. And CJ has talked about enterprise deployment, he thinks probably 12 to 18 months from now. What's the likelihood that could happen sooner in terms of you guys seeing more meaningful benefits? But on the flip side, what's the risk it could take even longer?

Yes. So pinning a time line is super tough. And that's why we started to talk this quarter about we're starting to see activity in the enterprises. A lot of that is, I don't want to say out of our control in terms of completely, but this is where enterprises have to get comfortable with the deployment of an agent to a customer in terms of reputational risk, governance, security. We're in there helping them do that. Is it 12 or 18? Is it 6 to 12? We'll see. We -- I think all this -- it's such a fast-moving market. And now you've seen even some people talk about, hey, maybe we need to lower the prices of tokens. That right there, I think, is -- it's just what happened in cloud. How many times did the cloud providers take their fees down. So that may move it a lot faster. But -- so it's -- we certainly hope it's sooner than that. We'd like to see some inflection by the end of the year, but it's something we'll watch.

Yes. Okay. So if you put the enterprise AI workload deployments aside, there's kind of 2 other areas. You have AI natives or digital natives and then you have the frontier labs. And I want to ask on each of those. Firstly, with the AI natives, going back to what you said earlier, you feel like multi-cloud has become even more important today. Do you feel like with these digital natives and AI natives that you are catching them and winning those workloads earlier in their life cycle than Mongo has historically won workloads in other customer segments? Because one of the things I think I hear and a lot of people here when they talk to customers is they started with an offering from the hyperscaler and then they scale and they upgrade to Mongo, if you will. But do you feel like with those AI and digital natives now because of the requirements of those workloads, you are able to catch them and penetrate them earlier?

So thank you. Great question. We want to make sure that we do both. And we spent a lot of time trying to get them before they even start their company in terms of using Mongo. This is around the reclaim the [ base ], all the stuff that we've done with our self-serve motion. This is where Voyage really helps because a lot of those folks will start with Voyage to do embeddings and they don't use Mongo and then we'll move them over. So we -- and that's really a marketing awareness. But we feel very comfortable as they start to scale and run into scalability issues, run into performance, run into cost issues that we will get them there. That's been the majority of the AI natives. And in that group, we're super excited about that. There's a couple that have gotten, I would call it, breakout size, but they're largely still pretty small. At some point, they'll just move into enterprise because they'll become real companies. And for us, those use cases are great because they are building their company and running their business on MongoDB, which is super important. And those are great use cases.

Yes. And then on the Frontier labs, I know they've asked you not to share much, so I won't push too hard on this. But what can you tell us about the frontier labs and the opportunity there with them near term?

Yes. There are certainly huge companies growing very quickly. So we feel really good about that they are using us and that their customers, they've asked us to not talk about the details. They are important use cases. They vary by the individual labs. This is something where we'll continue to work with them. Again, we're super excited about having them as customers. Hopefully, there's more growth there. And there are a bunch of different use cases for us to go after, but it's going to take time, and we'll see where that goes.

Yes. Okay. And then lastly, just on this thread, I guess, if we think about Mongo for the balance of this year, next 12 months, which of those 3, I guess, verticals do you think represents the biggest near-term opportunity? But then also long term, it sounds like it's definitely the enterprise AI workloads. But near term, frontier labs, digital natives, enterprise AI workloads, I guess, what do you expect to drive the most consumption and new business near term?

So we haven't built in much inflection in either of those 3 in the guidance that we've given. Absolutely, the enterprise is the long-term growth. I also think that as they start to deploy, hopefully, if we see better results this year, and hopefully, we do, I think it will because we'll see enterprise start to deploy more. The AI natives are great, again, but they're still relatively small. And then the frontier labs, we'll see what that looks like. I think it still all revolves around the enterprise.

Got it. I want to shift gears. One of CJ's big initiatives when he joined was to drive more of a top-down selling motion at the C-suite level and get broader buy-in for the MongoDB platform for -- as opposed to workload by workload, 100 workloads, things like that. We have seen your RPOs inflect, obviously, that tells us you're signing longer-term deals with larger commitments. But just tell me how that initiative has been going? And I guess, what inning are we in, in terms of that motion starting to complement that workload-by-workload winning?

Yes. So this -- I'm going to go back about 18 months. So CJ, I would say, has accelerated it. This was a movement that Mongo did even before that with Dave and the team in terms of moving more resources upmarket. As self-serve became a better and better customer acquisition vehicle, and the ability to move upmarket, call it, into that small to mid-market, we were able to move more resources to the enterprise. So we had already started that movement. I think we're -- and then last year, even in the second half of the year, we started to push more incentives for the sales team, hey, long-term incentives and long-term deals are good as long as it makes sense for both companies. So that's what you've seen in RPO. And then CJ has come in to help accelerate that. And then with Ryan as well, he understands the consumption business, but he also understands enterprise. So it's a process that started several years ago. Hey, the workload-by-workload and marketing awareness to your first question, is still super important. And we feel like we can do both in terms of continuing to drive net new adds, build marketing awareness, get them before they get started while we also build awareness at the enterprise level. And it's the same thing, developer awareness, either in an AI native or a large multinational bank, very much the same motion.

Yes. I just want to double-click on the RPOs because I think one thing maybe investors I've heard at least is they see RPO growth inflecting and then your Atlas growth is very strong. It's been 29% plus, I think, the last 4 or 5 quarters, but not a one-for-one kind of inflection there. And so just talk about how we should expect RPOs to look on a sequential basis from here throughout the year, but also how should they relate to your Atlas growth and your total revenue is growing going forward?

So RPO is both EA and Atlas. Atlas has become a bigger percentage of that as we went through fiscal '26. Keep in mind that, that is the -- those are the RPOs for multiyear deals, everything above 12 months. So at some point, it will find its way to revenue. It won't be a one-for-one growth, clearly. And also keep in mind that if a customer does a longer-term commitment, they still need to consume. It's not a seat-based model where you're going to recognize that revenue without the consumption. So we feel good about that. Will they move together? Sure, but not exactly.

Yes. Got it. You made a comment on this past earnings call that we should not expect significant variability to your next quarter Atlas guidance. And I know why you made that comment, but some investors have asked, was there a change in guidance philosophy? Is there more to that? What's the right level of upside to expect? Just expand on that comment. What did you mean by that? And if you look at the last couple of quarters, basically 2 to 3, maybe 3.5 points of upside in your current quarter Atlas guide. Is that the right range that we should think of if consumption is in line to a bit better going forward? Or how should we think about upside to next quarter guidance?

Yes. So thanks for the question. And we've talked about this a lot. And in retrospect, I should have just done the math on the call because I think there were some questions. Let's back up for a second. The company historically did not give specific Atlas growth guidance. And I think there's been some questions about why wouldn't it be the 500 basis points we've seen in the past? Folks, that's total company, right? And EA, historically, we've talked about this publicly, has been the biggest driver of outperformance historically. In the last 2 quarters, we've given specific Atlas growth. As you mentioned, in Q4, we came in above that by 220 basis points, and we said consumption was relatively in line. In Q1, it was 340 basis points and consumption came in a little bit better than expected. So what we're -- and keep in mind that 2 years ago, Atlas was a $1 billion business. It's now run rate over $2 billion. It is getting bigger. The ability for one cohort to move it, it's just a fact, is less so in the past. So we feel really good about Atlas, 29% growth the last 4 quarters. Going forward, that's the kind of range you should expect given the guidance and consumption will drive whether we're a little bit better or a little bit higher.

Yes. Got it. Okay. So in Q1, -- you did say that consumption was actually a bit better than expected. So first, a 2-part question there. Firstly, what drove that? I guess, if there's anything in particular you'd call out or if it was broad-based? And then second question on that is your Q2 Atlas growth guidance of 26% was still similar to your Q1 guide. And so given that the consumption improved, I know your Q2 through Q4 consumption assumptions have not changed, but why didn't we see Q2 guidance be at least maybe 1 point higher given the better ending point in Q1?

So keep in mind that the compares get tougher as we go through the year. Last year in Q2 is really when we saw the inflection in growth up to the 29% to 30%. So that's one thing. What happened in Q1 is we saw better performance in the enterprise, and we talked about this across mostly the U.S. and Europe and in the larger accounts. So because you then enter Q2 at a higher level, that's what we rolled through. We did, in the raise across the board for our modeling and everybody has their own, the very large majority of that is in Atlas for the rest of the year because you start with a higher base. To your point, we haven't changed the consumption patterns and seasonality we would expect through the year. So higher base rolls through the year, and then it is a little bit tougher compare in Q2, and that's how we guided.

Yes. And then I think as we look to the second half of the year, obviously, I think Q3, Q4, that leaves more variability for potential upside. How should we think about second half, potentially the year? And what would be the upside drivers to maybe sustain a similar Atlas growth rate in the high 20s all throughout the year?

I think the big variable there, Rudy, is just how does the economy perform. And that's the part that we don't have visibility to in terms of, hey, is inflation an issue, what happens with rates, all the geopolitical stuff, that's the part because it is a consumption business where we feel really good about driving Atlas growth. Customers are also able to spool up or down as they go through the year. So that's the part that we don't have visibility to. As we get through each quarter, obviously, we get better visibility. So that's really the big driver for us. And then if we start to see some inflection in AI, hopefully, that's a positive as well.

Got it. Okay. I want to jump to EA because EA has been, one, a big source of upside actually the last number of quarters now, but also there's a renewed focus on it, bringing the feature parity with Atlas. Why the renewed focus, right? I think you're going to have some of the vector search and search capabilities available within just a few weeks. What does that mean for EA? Do you have pent-up demand for those capabilities on-premises to build AI workloads on-premise? Just talk about that.

Yes. So this one to me is -- it makes me scratch my head a little bit. So let's back up. The EA business, I think, in the past and even at Mongo, I don't think we've done a good enough job talking about the benefits of EA and the customers that are on it. It's 20-plus percent of the business. It's some of the largest companies in the world. They signed multiyear deals to commit to MongoDB, which is great, awesome. A lot of the work around EA has been customers asking us to add that functionality. And this is a question, not a statement. Is this part of the push for more AI on-prem? And that's certainly what we're starting to see is I want to run those AI workloads in my own data centers or the colos that I have. Hey, economics matter a lot. And I think they're saying, if I have the -- why would I move the data, data has gravity, I'll run it on-prem. So I need the same functionality you have in Atlas on-prem. So great. Is there a pent-up demand? Yes, because everything we do in EA is coming from customer requests. And that's a big part. And I'll get on my soapbox a little bit. Some investors will say, all we care about is Atlas. As a CFO, I love Atlas. I also love EA. Biggest customers in the world, long-term commitments, very good profitability. And at some point, if they transition to the cloud, you want them on-prem. And multi-cloud is not only multi-cloud, it's also on-prem and in the cloud. So for us, it's a super important driver of growth. Can it be better than the mid-single digits that we guided? We certainly hope so. And again, it's huge customers making commitment and a lot of those bigger ones will run both on-prem and Atlas. So there's a big piece of that as well. So we feel really good about the EA business, and we'll continue to look at what else we add. For instance, we're going down the path to get FedRAMP High certified. That's all on Atlas. At some point, we should also bring that on-prem. So we'll look at every feature like Voyage and all that and decide do we bring it on EA or not. And a lot of that is customer requests to add functionality. And the last thing I'd say is you talk about AI winner. That is where we believe a lot of AI workloads will be, and we'll talk more about that, not only the Atlas growth, but the EA growth being driven by customers running AI workloads on-prem.

If I listen all that back, it somewhat maybe it sounds like you actually expect some of the large enterprise AI workloads to maybe take place first on EA as opposed to Atlas. Is that a, fair, maybe readback? Or how should we think about that?

So that is a fair readback, and that's what we expect. And that's what customers are telling us, which is, I'm not going to move all my data to the public cloud. I'm going to run those on-prem and maybe they train it in the cloud and then they run inference or other -- or build agents on-prem. We'll see. We want to meet the customers where they are, and I think that's a big part of the adds to EA.

And how does that change? I'm just trying to think like if we actually see the enterprise AI workloads on-premise first as opposed to the cloud, how does that change maybe the competitive environment and the competitors that are available to do that? Is that materially different than relative to the cloud? How do you think about that?

So we think -- and 2 things. One is, for us, this isn't -- it's an and, not an or. We still fully expect to grow Atlas at healthy rates and EA. So for us, it is additional TAM we can go get. And to that point, from a competitive environment, we feel really good about, again, this is where multi-cloud comes in. This is where the platform scalability across those, and we feel really good about being able to do that both on-prem and in the cloud.

Okay. And then you also said hope to maybe exceed the mid-single-digit growth there on EA this year. I guess I know ARR growth in EA was 11% in Q1. I guess if we think longer term, I'm not asking you for guidance for the next year or anything like that. But if you get this AI capabilities and feature parity on-premise, there is pent-up demand. Is it within the realm of possibility that EA could be a durable double-digit growth business over the next few years for you guys?

My answer to that is we wouldn't make the investment if we didn't think we could drive better growth.

Good answer. I like that. Okay. The messaging around margins has been super clear since you joined, and I think very much appreciated by myself and investors. You've made a very strong commitment to, I think, 100 to 200 basis points of EBIT margin improvement each and every year. Obviously, last year was a big year way ahead of that schedule. If you do see kind of an inflection in the enterprise AI workloads, whether it's 6 months from now or 12 months, 18 months, whatever, would you potentially pull back from that commitment a little if it really inflects and there's a massive opportunity to go win market share? Just how should we think about that if and when we see that AI inflection?

Yes, it's a great question. So I'll answer that 2 ways. One is we've been super clear. The driver of margin growth and expansion is going to be driven by revenue growth. When it comes through gross margin at 75%, there is a ton of investment to do. So we still believe that. We don't believe that the enterprise inflection is because we don't have enough resources. Again, if you go back to the work that they have to do to be ready, if we thought, Rudy, that we could put more dollars in go-to-market or there were product things that we could add to accelerate that growth, we do that now. That's all based in our margin goals. So we fully expect we can continue to grow revenue, margins are going to go with it. If there are some trade-offs, we'll make those. And then the third piece I'd add is keep in mind, and we've talked about this a lot, Mongo is a great company. We also have efficiencies we can drive internally. We have not pushed nearly enough offshore. We're going to do that. There's areas for us to drive more efficiency. We're going to do that. And we are a little bit late to the AI game internally. So there's a lot of areas for us to reallocate internally to go drive that growth, and we're doing that now. So we don't think that the enterprise inflection in AI is because we haven't allocated resources. If there's a clear return, we will certainly look at it.

Yes. Okay. I want to -- going back to the enterprise workloads because I agree. I hear it in our checks, these customers, they'll get a workload, they're excited about it, but there's a 0.01% risk that it does something it shouldn't and then it gets shut down by compliance, things like that. Is there anything that you feel Mongo does have in its control to maybe help avoid that or to help limit the chance that this workload screws up 0.01% of the time. Is there anything -- like, again, I know I would agree, it mostly seems out of your control, but anything that Mongo can do to help alleviate that and therefore, pull forward maybe some of these enterprise AI workloads?

I think the big area of focus for us is, number one, making sure that the product is there so they feel comfortable. The ability to run it across multi-cloud, I think, is super important for their risk. From a security perspective, we think that we bring great reliability and security. The other thing is to bring use cases that we see within our customer base to make them more comfortable. There is that part, which is, hey, what are you seeing across your customer base? What are people deploying? How are they doing it? That all helps. But at the end of the day, that company still needs to say I'm going to go.

Yes. Got it. Okay. I also want to ask, you touched on competition a little bit at you're in the multi-cloud stuff. Postgres, we still get asked about it all the time. What's the update there? How do you see them as a competitor? Is it still an overblown fear? Just what's your view?

So I don't want to say something is overblown. They're a competitor. I think that we see them mostly in the smaller companies where it's really not a decision and where they don't look at building out the infrastructure. They just want to deploy an app. But once they get to a certain scale, that's when we start to see them come to us. And I know they've added as a plug-in JSONB, and we look at that and say, interesting. But if you are an enterprise running a real enterprise workload around scalability, performance, again, multi-cloud, MongoDB is the place to be.

Got it. And then maybe one last one before we wrap up here. Another question I think I get a lot is, if you look at these customers who are running tons of internal AI agents, Copilot, Claude, et cetera, does that have the potential to meaningfully benefit Mongo? I mean what does that tech stack look like? Do those types of agents use Mongo? Do they use a different database? Do they use no database? I think some investors are just curious why doesn't -- I think they understand maybe it's going to take longer for the enterprise AI workloads. But with those agents, like why doesn't that meaningfully benefit you?

So it's a multifaceted answer. So again, it goes back to the use cases that the foundational models are using us for. And some of those will scale with their growth. Some of them will not. The other thing is when you get, there's also all the vertical applications or specific that they may use us as a database. That's where the AI natives, I think, will drive growth, which is if they're actually building their product on Mongo, then that would scale more with their business in general. So there's a lot to that question. And again, this is why we go back to Atlas is a great business. It's $2 billion of run rate. It takes a lot to move that, thankfully. And again, we feel really good about driving durable growth there. It depends on the use case.

Got it. Super helpful. That's it. I think we'll go and wrap it up there. Mike. Appreciate the time.

Thank you, Rudy. Appreciate you having us.

Yes.