MongoDB, Inc. (MDB) Earnings Call Transcript & Summary

Good afternoon, and good morning, everybody. My name is Tyler Radke. I co-head the U.S. software sector here at Citi, and welcome to day 2 of our virtual tech conference. We are happy to have MongoDB. We have Michael Gordon, the CFO and COO; and Serge Tanjga, the VP of Finance. Guys, thanks for joining us. And I think...

Thanks for having us.

And I was going to say, I think you're only a few blocks away from us in the conference here in Midtown in a virtual conference room. So hopefully -- we're getting closer. Hopefully, next year, we'll actually be in person.

Yes. So there was that time in June when we were hoping that we would be in person even now, but I guess there's always next year.

We took a baby step and we're at least here together.

Yes.

Yes. There we go. Slowly but surely. So guys, thanks so much for joining.

And I thought it would be a very appropriate place to start just talking about the last quarter. It really seemed like the business hit an inflection, particularly in Atlas, your cloud business. We saw a big reacceleration in the growth rate in Atlas as well as on total revenue. Maybe just give us a quick recap of the quarter and what do you think is kind of driving this inflecting momentum that you're seeing in the Atlas business?

Sure. So a few things. Overall, yes, it's definitely a good quarter. We're really pleased with the results across the board. For those who don't know the details or don't have them handy or need the quick refresher, revenue was up 44% year-over-year, just under $199 million in revenue. Subscription revenue also up 44% year-over-year. Atlas, which is our Database as a Service business, as Tyler mentioned, at 83% year-over-year growth and now represents 53 -- 56%, excuse me, of total revenue. So pretty strong across the board. Atlas exceeded our expectations in the quarter, but so did Enterprise Advance, and so it was really quite strong. And I think ultimately, what we're seeing, if we just take a step back is, well, this quarter was particularly strong. And our coming up on 4 years as a public company, what we've seen is a fairly consistent theme, which is we're attacking one of the largest markets. IDC forecasts databases to be $73 billion growing -- in 2021, growing to $119 billion in 2025. So a very, very, very large market. We have exceptionally strong and demonstrated product market fit. Developers love MongoDB and building applications with MongoDB. And we've got a sort of differentiated general purpose positioning that's resonating with -- in the market with customers, with developers. And we've paired that with really solid execution in terms of actually going out in the market and sort of converting on that opportunity. And so while this quarter was particularly good, I wouldn't say there's anything that we've done differently or any one thing you'd point to as being radically different. I think it's just sort of a continuation. Some quarters -- generally, all of our quarters have been good, but some a little better here and there. But generally all consistent of us executing. And at this point, we're now about 1% market share and these very large markets are really just getting started.

Yes. Yes. And Mike, I think you've been at our conference, I think this might be the fourth year in a row. So first, thanks for joining us. But I think last year, we were talking about some of the record net adds that you saw in Atlas, which I guess, in some ways, it was kind of a preview of what was to come. I guess if you were to kind of put yourself in your shoes a year ago and kind of seeing what you saw in the net adds, like is this the type of Atlas growth that you were expecting? Kind of what has changed for the company and maybe the overall market relative to last year?

Yes. So certainly, in general, we've exceeded our expectations. And certainly, Atlas has been a large part of that and really driven that. And if I kind of take a step back and maybe think about the business or talk about the business on our channel basis, which ultimately how we run the business, we provide lots of disclosure, so people can kind of slice and dice things different ways. And certainly, the Atlas numbers are very strong or tantalizing or whatever. And so I know that's why people want to focus there. But we don't really run the business on a product basis. We run the business on a channel basis. And so if you think about the channels, the 2 main are sort of the self-serve channel and the direct channel, we can kind of take each of them in turn. I think the self-serve channel, we have been having more success. We have talked for a long time about the prevalence of MongoDB, the popularity of MongoDB, the usage of MongoDB. And now with Atlas, we have a direct way, particularly on the longer tail of usage for sort of monetizing that with Atlas. We've talked previously about this self-service motion, product-led growth, other ways of talking about it, that weren't quite in our DNA. And over the last couple of years, we've made investments in those teams. And I think you see that showing up in the customer additions, in the success that we're having. And we've been very pleased with our progress. We still think there's more to do given sort of the magnitude of the opportunity, but we're really pleased there. And then on the direct sales side, we've similarly outperformed really kind of 2 flavors of the direct sales. There's the mid-market team, which is mostly sort of -- think inside sales team. And that team at this point is almost -- the vast majority of which is Atlas oriented. Just given the nature of those customers, they tend to be very cloud forward. So the Atlas product market fit there was exceptionally strong with them. And then the enterprise channel. Within enterprise, we have increasingly seen success there. I think we talked in our June call about roughly 2/3 of the new business that we did in the quarter, even from the field team was Atlas, which I think sort of indicates the fact that large enterprises are sort of increasingly looking at and getting comfortable with Atlas for their mission-critical, quite meaty workloads. The other thing that we talked about and stat that we shared is we report out each quarter our customer count above 100,000. And we kind of shared a view that said if you only looked at people's Atlas spend, right, and you ignore what they spent on EA, about 60% would still qualify for that metric. And I think that was helpful insight or perspective from what we see, which is sort of the increasing comfort with Atlas for mission-critical enterprise workloads, but I think were sort of new and helpful for some of the folks sort of on the outside. The last comment that I'd make, just because the database market is different than a lot of markets in software, is people tend not to be early adopters. People don't tend to adopt a new release when it comes out right away. There tends to be a little bit more waiting for seasoning and maturing. And so some of the adoption curves, in general, like including overall for all of MongoDB, tend to take a little bit longer to play out, and play out mostly on an application-by-application or workload-by-workload basis. And so logically, these things take time. But I think we're reaping good returns on the investments that we've made over the last few years.

Okay. Okay. Yes. And I did want to kind of touch on those main channels that you talked about. And I think one of the things that we've seen pretty consistently is your new customer additions have remained at or greater than 2,000 a quarter in terms of net new customer additions, which has been pretty strong and I think kind of helped to continue to drive this accelerating growth in overall customer growth. Could you just remind us some of the go-to-market changes that you made that have helped to drive that acceleration in customer growth? And how should we think about the sustainability of kind of this 2,000-plus customer cadence going forward?

Yes. So let me sort of break down a few of the pieces and then happy to take any of the follow-ups. So of the 2,000, just to level set, the vast majority of that is self-serve customers, right? So that goes back to some of the things that Michael was mentioning before as sort of investment that we made, in gross marketing investment that we've made in sort of building the product-like growth DNA inside the company that sort of in earnest really began a couple of years ago. And those are the initiatives, those are the efforts that drive the majority of the customer count because -- and again, self-serve customers have to be very small when they come in and then they grow like all our customers do over time, and some portion of them sort of eventually qualify to sort of move over and get direct sales coverage, which I'll get to in a second. But as you think about sort of the -- our journey on the self-serve side, think of it as kind of crawl-walk-run, and we're making progress in that continuum. We've done a lot of progress in some of the sort of most obvious channels, sort of deepest panels for us to sort of improve both coverage and conversion, although there's still more to go. But we continue to see opportunity in terms of diversifying top of our funnel and continuing to invest and sort of improve either through sort of product experience or through data signals to improve conversion as the customer moves through the funnel. Now what that means in terms of sustainability of the 2,000, like it's hard to comment. But we definitely feel like we made significant progress, but also feel like it's still relatively early innings in terms of improving our expertise. Now when you move to the other side of the business, the direct sales business, we've seen an acceleration in customer count there as well. And that's -- that really in earnest began at the beginning of fiscal year '21. And that had to do with changes that we made first in our back-end system and then in our sales compensation to facilitate customers getting on our platform, particularly around Atlas. And so what we found is that our customers, when they get on the platform, they really like the product and they consume more and they bring more workloads over time. And so where we were slowing ourselves down was focusing too much on their first contract, the size of the first commitment, and that was basically friction that wasn't necessary or long-term additive for us. So as we thought about ways and succeeded to remove or reduce that friction, we've seen an uptick in customer count on the direct side. And so if you break down the direct additions into 3 pieces, one is sort of in our field sales team like Michael was saying, we've seen an acceleration in customer count there. In the inside sales team, we've seen an acceleration in the customer count there. And then the final piece is the migrations that I mentioned before. So customers who start off as self-serve, we get them to make a relationship with us on the direct sales side, we've also seen an acceleration there. And sort of as we look at the health of those customers in terms of newer cohorts versus the prior cohorts, we're very happy with the behavior. It's very sort of consistent with history, although it is early days. The one thing that's worth pointing out is that we've had more success in the inside sales channel and the self-serve migration, so that influences the mix a little bit. But on the apples-to-apples basis, it's been very consistent behavior across the board to history.

Right. Right. Okay. That's a good overview. So I think that -- you mentioned how some of the changes you made, I think, on the inside sales team in terms of how you are compensating reps to get folks up and running, prioritizing new logos versus large upfront commitments, that was about 1.5 years ago, right?

Correct. Yes.

And I guess were there any tweaks you made this year? What were kind of the highlights of some of the changes you made from a go-to-market either incentive or strategy perspectives?

Yes. I guess I would just say we make changes all the time. It's a constant sort of sense of iteration. And so Dave mentioned on the most recent call, some experiments that we've been running with covering high potential accounts and giving them more resources. That's something that we started. We've been working on them for kind of 2 years and we're liking what we're seeing there. So I think we're constantly working on different ways. And the macro backdrop against all of it is we have an incredibly large market, we have really strong product market fit, we have incredibly thin footprint coverage and so we're operationally constrained, right? The way we start out each of the planning processes is how many sales people think we can credibly go higher operationally, right, and because we're not opportunity constrained and so we have the luxury of being able to approach it that way. And then what we look for are ways that we can sort of incrementally drive productivity, synthetically extend the reach of reps, make sure reps are focused on the things that are unique to sales reps to do best and sort of declutter the rest of their to-do list. So we're constantly working and iterating through all those kinds of things, and I expect we will continue to do that for years to come.

Right. Okay. And maybe on that point around the experiments you're making in terms of resourcing the highest value accounts. I think another comment that -- from the last quarter call that stood out to me, which is talking about kind of adding in more high enterprise sales capacity and sophistication there. And I'm curious, like, is that being -- are you seeing kind of the decision-making in this market at the enterprise happening at a higher level? I know in the past you've talked about how often database technology can be decided on a workload-by-workload basis. But presumably, once you get to a certain size at a large enterprise and you kind of get the use cases proven out, I imagine that, that conversation gets elevated. So just talk us through what you're seeing in terms of how the decision-making has changed and if that's influencing your decision to add more kind of enterprise sales capacity.

No. So I think on the sales capacity, again, it really goes back to the sort of footprint coverage, where our headcount of sales -- quota-carrying sales rep is measured in the hundreds versus the thousands or tens of thousands for our competitors. And so that's really what drives the -- our aspiration to scale the sales team as quickly and responsibly as we can. I think in terms of the buying decisions, there are a couple of different flavors, right? There are more -- some of the workload-by-workload decisions can happen at more junior level developer, application team levels, kind of bottoms-up levels. But similarly, you can jump-start or accelerate adoption in the account if you're calling higher into the account and describing not just this feature set or how we might be better in this particular application, but more broadly what leveraging our application data platform can do for you in terms of your overall speed and pace of innovation and successive innovation, and ultimately, your competitiveness. That's a message that resonates much higher up within an organization. So the macro message is all consistent, but you're sort of going to emphasize different things depending on the audience. And I think over time, we've gotten better and invested more at the sort of calling in [ higher-tiered ] account. I think -- we still think we have work to do there. But I think they're complementary to each other.

Yes. And just to tie that into some of the commentary from the last call, the best way to have a successful strategy of calling high is if you have internal proof points that we're talking about. And this is where the incremental investment comes in, which is a more focus, more resources to sort of further develop that sort of bottoms-up adoption and then sort of up-level the conversation at the next level at a more higher-level message around innovation, and ultimately around becoming a standard that would first -- facilitates adoption.

I see. Okay. Okay. Maybe shifting gears to product. Is there a way to think about how much of your R&D today is spent on Atlas versus EA and how that mix has evolved historically?

Yes. So I don't quite think about it that way. And I think it's important to recognize that the underlying database that's packaged in Enterprise Advance is the same database that runs Atlas, right? The Atlas work tends to be the user interface on top, and then all the other products and services that we're offering, right, search or online archive or things like that. And so I don't think you can quite -- it doesn't quite parse out in the way that you might think because a lot of the core investments benefit sort of both product lines equally, which is really very consistent with our theme, which is we just want to make Mongo to be easy for customers to consume regardless of where they are in their cloud journey. Certainly, there was a period of time when we first introduced Atlas, where Atlas lagged EA. We closed that gap, brought it up to parity, and then subsequently, over the last few years, have really led with Atlas. And so if you are using Atlas, you will get more benefits of the platform and ecosystem, but it's not because you're getting a better or different underlying database.

Right. Right. Okay. And as you're thinking about building out new features, how are you kind of deciding where to prioritize future development? Are you kind of pushing most of the new features at this point onto Atlas and then ultimately releasing that back onto EA maybe a couple of quarters later? Just help us think through that philosophy on how you're trying to differentiate the Atlas product maybe to incentivize customers.

Yes. No, you're exactly right. And so kind of the brief in reminder history, like Atlas used to lag, it reached parity in 2018, but we've been leading with Atlas since then. And why is that the case? A, we can iterate it faster and ship innovations to the market faster. And then importantly, we have feedback faster, right, because we see what customers are doing. We see what sticks, what doesn't and that sort of makes for more efficient R&D process and better ROI. And then we take, to your point, sort of the learnings of what sort of sort of the rapid innovation in Atlas and sort of deploy that and have the EA customer base benefit from that as well.

Okay. Okay. Got you. Just shifting gears to the financials a little bit. I think billings has been kind of a noisier metric for you just given the growth you've seen in Atlas. What -- how would you encourage investors to just kind of assess the health and momentum of the business? Is there a KPI that you would focus people on?

Yes. And it tends to be revenue. As imperfect as it is, it's certainly the best that we think investors should focus on and one we focus the most internally. So just briefly on sort of the -- what you mentioned on just for the benefit of the entire audience. Some of the changes that we've made to facilitate adoption of Atlas also means that billings, which we never thought were a particularly good metric for us, becomes increasingly less useful over time. So because if you're not anchored on a large upfront commitment, if you're instead paying as you go or if you pay monthly, like that just means the billing -- the relationship between billings and revenue sort of [ diverts ] over time. And at the end of the day, we look at we look at revenue growth. Obviously, there's a little bit of lumpiness around EA, but, a, that's sort of declining as Atlas is growing as a percentage of the business, and both we and I believe the investor base is getting more used to sort of the EA variability caused by the ASC 606 term license piece of it. And also, whenever that's relevant, we try to call it out and give you the context and the color so that you can sort of properly calibrate to sort of guide around those challenges, but it ultimately comes down to revenue.

I think the last part is really important for everyone to understand. There is no perfect metric. We certainly understand that. We provide lots of disclosure and ways for people to sort of slice and dice the business, and obviously have conversations like this and other things. And so I just think it's important to -- when we think about it, I think from like a communications philosophy standpoint, we're focused on revenue and then we're trying to provide a lot of context and color about the contour of that. And there are always things that happen below the surface. But when there are material things, positive or negative, that's when we try and call them out just so people can understand and recognize that there's no one perfect metric.

Right. Right. Okay. So shifting to the EA side of the business, Enterprise Advanced, that's obviously been overshadowed, I think, by the success we've seen in Atlas, but still kind of growing mid-teens over the past few quarters. How do you think about the kind of annualized growth trajectory from here? Is it -- are you seeing any headwinds there just as you are trying to see more enterprise customers choose Atlas? And have you seen any migrations thus far from EA to Atlas?

Yes. So I guess the first thing I would start with, and forgive me if we're starting to sound repetitive, is that we really tend to focus and plan and run the business on a channel basis, and that comes down to like, frankly, how quickly can we scale our sales force and then how quickly can we invest in the top of funnel around self-serve. And those are really the drivers and sort of the things that we sweat internally, the product mix, particularly when you're more than 1 quarter out, is really the output of the business -- sort of how the business works as opposed to an input. That said, on EA, some incremental context that might be helpful is, as we look at our customer base, we sort of see pretty much a barbell distribution, which is a vast majority of our customers are 1 of 2 categories. One is exclusively or almost exclusively EA, and then on the other side is exclusively or almost exclusively [ app ]. Now you might ask yourself why, and I think the answer as we observe it is that, that is fundamentally a function of their IT strategy and sort of the willingness that they are -- the speed with which they're moving to the public cloud, right? And for regulatory and security and other reasons, some companies or industries, frankly, are further down the journey than others. But if you focus on the piece that is EA or that customer base, some of the major truths still hold, right? Like our penetration is low, customer satisfaction with the product is high. Back to the -- some of the investments that we made around our highest-quality, best prospect customers to accelerate growth, some of those were focused on EA customers. So Atlas is exciting. Atlas is growing faster. Michael walked through the channels. We expect that to continue in the business to continue moving in the direction of Atlas, but that's not to say that there isn't a big role for EA to play going forward. One final thing I would say is on migrations from EA to Atlas, we see very little of it. Now obviously, in the fullness of time, we expect that to change, but we haven't seen it yet. And sort of it kind of makes sense, because at the end of the day, an application running on EA works. And so IT department of any company is going to focus on either new stuff or stuff that doesn't work, not on the stuff that is working just fine. So we don't necessarily think that there's any reason for that to accelerate in the very near term here.

Okay. Makes sense. So Mongo has always been a really popular technology within the broader development -- developer community. Community downloads is something that you've talked about since the IPO. And I think last earnings call, you referenced over 100 million-plus downloads. How would you encourage investors to just -- to view the download activity? And how do you guys look at it in terms of the opportunity and potential ways to monetize it going forward?

Yes. I look at it sort of as an indicator ultimately of mindshare. I don't get overly fixated on like a specific number, the 200 million cumulative downloads that we mentioned was just sort of interesting because of attraction of round numbers as milestones and 75 million within the last 12 months. And so I think you can see sort of the increasing adoption. I don't think it's incredibly unique or insightful in the sense of I think people are coming to appreciate how popular MongoDB is with developers, how much developers like using MongoDB. And so to see the mindshare and the increased adoption shouldn't be particularly surprising, but I think it's helpful context, nevertheless. And certainly, Dave has said in multiple different settings and multiple occasions about sort of the business starts to stop with developers. And so I think it's sort of important not to lose sight of ultimately the people who are using the product hopefully getting the benefit out of our technology and out of our products, and we continue to see very strong usage engagement and adoption.

Yes. Okay. So another kind of higher-level question. I mean it feels like it was about a decade ago, maybe a little bit longer since kind of this idea of NoSQLs as a category really started to go mainstream, at least in investor conversations. And clearly, Mongo has built an amazing brand and developer following. But at the same time, the world still does seem pretty dominated by SQL-based databases. So I'm curious, where do you think we are from a new SQL maturity perspective? Do you feel like the awareness is there to kind of see a step function increase in adoption over the next 10 years compared to the last?

Yes. So I think about this in a few different ways. I mostly don't think about the term NoSQL for the first part, in part because the whole reason why it came up is because end users are frustrated with the challenges of scaling relation, right, that's what created this whole opportunity and it's this $73 billion market, where that's been dominated by one technology, but it's a technology that faces a lot of challenges when you think about the scale and the speed and pace of innovation and iteration in today's more complicated world, relation was just built 4-plus decades ago to solve a different set of problems. And so it's totally logical that in a big, large market that hasn't been disrupted in decades, that you would expect there to be new entrants. The reason why it's a little bit more than Symantec is a number of the players that were sort of trying to disrupt the market, in our opinion, kind of threw the baby out with the bathwater, hence the sort of NoSQL; as opposed to the MongoDB approach, which was to keep some of the best, most important things of relational databases, but also build in the scalability, agility, et cetera, et cetera, that MongoDB represents. Which is why, I would argue, we've done a better job than others and sort of been able to separate ourselves from this sort of crop of would be challengers 8, 10 years ago that were all sort of in the same pool and the same size of scale, because we have this technology that really took the best of both worlds and had the scalability but also the reliability and the consistency and the speed with secondary indices and things like that of relational databases. And so as a result, with that better technology married with the developer mindshare that we were just talking about, and again, combined with execution, that's sort of what's allowed for the separation. What I think is important, though, to focus on that's sort of embedded in your question is the reality that despite MongoDB's success, SQL is still the dominant language, orientation, mindset default in many, many accounts. And so the opportunity set that we have, but it's also the risk, is when there are situations where the application itself isn't struggling, right, when the application is struggling, people understand that relational -- replacing relational with relational isn't going to be the answer, right? And so I think MongoDB is a pretty widely accepted or understood opportunity to go fix that situation. Our challenge and opportunity is that, too often, it would be easy to say, okay, I need to modernize that. That means getting off of Oracle and going into the cloud, and what the customer misses, and therefore we miss, at least in this kind of application or life cycle, is the opportunity for them to get better application performance by moving from relational to MongoDB, right? And so there is still sort of bias inertia, fluency. And I think it's changing. Certainly, if you were trained in databases or came up as a developer 20 years ago, SQL is sort of the not just the default but the sort of the gold standard. And the paradigm shift that MongoDB has introduced, including things like multi-document asset support, are such like so mind-bogglingly different than what you were taught, that's a little harder to wrap your head around it. Whereas, I think, people who've been steeped in modern application development are much more familiar, fluent, comfortable and they have a certain different set of expectations that MongoDB delivers well against. And so I do think it will take time, but that's sort of some of the tension and some of what we observe out there.

Yes. And a related question. I mean do you feel like there's kind of a demographic change on -- that's helping you? I mean I think, certainly, if you walk the floors of MongoDB World versus Oracle OpenWorld, you might find that...

Yes. No, there's a generational dynamic for sure. But because it is vintage or tenure-based or whatever, it doesn't just change suddenly 1 year or the next, but like the -- it's certainly a tailwind as you think about what will the massive developers look like 10 or 20 years from now. They'll look pretty different and they'll -- I'm sure they'll -- some will still know or speak or even prefer SQL. But I think the mix will change pretty significantly over time.

The other way to maybe get at it, Tyler, is also just to think about the evolution of our product. So the original idea in the power of document model has been there for a while, but then we needed to build all this functionality around it and sort of harden the database as it is truly sort of transactionally ready and mission-critical. And that -- sort of the final major piece of that was multi-document asset which was completed in 2019. So that's like the product sort of quality. Then there's the perception of it, right, which lags the quality itself. So if you find developers who play with MongoDB 7 or 10 years ago, they will tell you many things that haven't been true about our product for years. So it's our job to sort of educate them and bring them up to speed to where they -- where we actually are. And all of that in the context of fighting 50 years of inertia and sort of the slowly changing attitude and the demographic shift that comes with regulation.

Right, right. Okay. Got you. And as we think about the hyperscale vendors, clearly, you partner very closely with them. They also have competing products. How do you see those relationships evolving over the next 5 years? Do you kind of feel like you'll be able to grow deeper with them in terms of more integrated partnerships and joint sales initiatives in the space? Or do you see kind of competition increasing from them just based on where you see them investing? Just kind of walk us through that kind of tricky coopetition dynamic?

Yes. There's definitely a competition angle. There's also a very strong partnership angle. Each of the 3 are slightly different and have different kind of underlying dynamics. In general, I would say, while there is competition for sure, I think we've done a very good job of sort of walking that balancing act. And I think that each of the 3, in different ways and in different orders of magnitude, understand the -- they all understand the popularity of MongoDB and they all understand the power of MongoDB. And 2 of them, Amazon and Microsoft, saw it so much that they created imitation competing offerings. We -- the good news is we have a very strong product. We continue to invest in that product and our competitive win rates, not just against the big hyperscale cloud players but really across the board, are exceptionally high. Those win rates are actually highest directly against imitation offerings. And so I would never suggest that they're going to stop investing and never suggest that they don't have much, much deeper pockets than we do, but we're eager to continue to look for ways to partner together as well, and we'll obviously have to see how it unfolds.

Got you. Okay. And one of the other cloud providers that you've done business with is Alibaba in China. And I think it's a different kind of revenue sharing arrangement that you have, not recognized in Atlas. Could you just help us understand the progress that you've seen in terms of that relationship? And how are you thinking about maybe approaching other international markets that may have a kind of similar type of presence?

Yes. So mostly, the China market has some differences in most markets, not all, but most markets around the globe. The 3 hyperscale players have a presence, and we're able to offer Atlas widely and globally, such that Atlas is the world's most widely available Database as a Service. We're in 80-plus regions. It's probably even higher now. But -- and so we benefited as they continue to sort of expand their own footprints, we benefit that and our customers benefit from that. China is a different market. We can't offer Atlas directly in China because the regulations there. And so the 3 players don't have direct offerings in those markets. We already had a field presence and a field team that was having success selling Enterprise Advanced into Chinese entities, including state-owned entities as well as private companies. But we wanted to be able to have a cloud offering. We couldn't offer one directly, as I mentioned. And so we had conversations and have now secured relationships with both Alibaba and Tencent to bring their -- they were offering MongoDB as a service. We've brought those offerings into compliance. We're in year 2 with Alibaba. The partnership is strong and going well and so far running ahead of the minimum commitments and the expectations that we and the Alibaba team kind of signed up for. It's too early to tell on Tencent. We have -- in smaller markets when there've been other players who either wanted to proactively offer MongoDB as a service, we partner with them. Or if there were rogue offerings that were not licensed, we brought those into compliance. And so I think it's -- we'd love to offer a direct Atlas offering in every market, but in markets where we don't have that -- and in China, in particular, I think this has been an effective route so far, and we're really excited about the partnership with both Alibaba and Tencent. We'll have to see how it goes. But again, sort of good initial results from Alibaba, and we'll also have to wait and see.

Okay. I know we only got a couple of minutes left, but maybe just for both of you, just kind of let us know where you're focused throughout the rest of the year? And if there was anything we wanted to cover that we didn't hit on.

I think we hit all the highlights. I mean I think the macro theme is pretty consistent, including from way back when in their IPO. It's a very large market. We've obviously grown the business significantly since those 4 years ago, but we've just crossed 1% market share. So we're just getting started. We continue to invest, to capitalize on that opportunity. We have liked and have seen good returns on those investments in both R&D and the sales and marketing side. We'll continue making those investments given our long-term orientation. But I think those are probably the highlights that we didn't cover.

Yes. I mean it's as simple as just continuing to execute. We talked about scaling, we talked about, particularly on the go-to-market side and sort of the key focus on view by channel and growing the sales force, as job one or has been really for a long time. And so the good news is that the market, as Michael mentioned, is large and the product market fit is excellent, so it's up to us to capture the opportunity.

Great. Well, I think we're a minute over. So thanks so much for the time. And I appreciate you for joining the conference, and for the investors who tuned in. So thanks a lot.

Thank you for having us.

Thanks for having us. Good to see you. Good luck with the rest of the conference.

Yes. Hey, thanks a lot. Talk to everyone soon.

Bye.

Bye.