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

Well, good morning, everyone. Thanks for joining us. I'm Jack Andrews. I cover the data analytics and infrastructure software space at Needham. We're very pleased to have Michael Gordon, the CFO of MongoDB with us today.

So Michael, just to kind of level set everyone. Could we just get an update on what's happening in the database market these days? I mean on the one hand, this market has been around for many years, and it's been very stable. The top 3 market share shift hasn't really been moving much over -- when you look over a longer period of time, yet there's a tremendous amount of innovation, like from the likes of MongoDB and others in this space. Just what's happening at a high level? Why is this happening? And particularly, why now at this moment in time?

Yes, sure. Thanks for having us here. It's great to see you and a bunch of familiar faces. So just to give people a sense for the database market. The database market is one of the largest in all of software. So $64 billion spent in 2019, estimated to grow per IDC to $97 billion in 2023. So just about $33 billion in growth over the next 4 years. So like I said, one of the largest markets. And the reason why the market is so large and why it's drawing so much attention and why there's so much sort of innovation and focus and everything on the market, is -- you hear these phrases like software is eating the world or every company trying to become a software company. Some of these large banks having more developers and engineers and companies like Facebook and things like that. And the reality is the reason that's driving that is that companies are increasingly trying to drive competitive advantage from software, right? They're trying, to whether it's competing for customers, improving the user experience, things like that. And at the heart of all that is a software application. And at the center of the software application is the database, right? And the database's scalability, agility, flexibility, all that kind of stuff, determines ultimately how competitive you can be in driving those improved user experiences. So if you're like me and you woke up this morning, and you got your phone, you had all these different updates for all the different applications. That's hinting or an indication of the pace of innovation that's taking place, and developers need to be innovating on all these applications. And again, if I can have a modern database as opposed to a legacy traditional database, that will put me in a materially meaningful advantage. And MongoDB is the leading modern general-purpose database. And so that's what positions us well for the opportunity.

Right. And so then fast forward a number of years, do you think that this database market will sort of reconvene over a couple of large market leaders? Is this analogous to the ERP world? Or do you think there's going to be much more of a dispersed market that -- based on use cases and things like that?

Yes. So the market in general is growing quite significantly. And one of the things that's sort of interesting, as I mentioned, the stats about the market size, if you run the calculations, you get about roughly 10% growth year-over-year. And typically, I think of more mature markets has tended to be more growing in GDP, right? And so again, I think that speaks to sort of the centrality and the strategic nature of what we're doing. I think over time, this will play out over several years, but there'll continue to be a handful of large players. We're obviously growing and focused on trying to capitalize on the opportunity for us. But I think that the market will take time to evolve. I don't think that there will be one dominant player in the way that historically it has been. I think we've seen sort of a number of changes there. But I think there's a lot of running room, continuing to go certainly for us because we're very early on in our penetration.

Sure. So when we think about your business, I mean is there a way to segment in along 2 lines. I mean there's kind of -- your leverage to certainly new application development, but there's also a lot of -- there's a migration movement happening of legacy applications. Is there a way to think about your business along those 2 dimensions? And do you think [ they're more ] important?

Sure. There are a couple of [ ways ] to slice and dice it. Maybe if we go back to the market sizing that might sort of help thinking in general. So I mentioned the markets, so $64 billion growing to $97 billion, so roughly $33 billion of growth over 4 years. If we just say, it's roughly $8 billion a year, that gives you sort of the new spend that's happening on a rough basis. IDC, I don't think, breaks it out this way. But fundamentally, my assumptions, most of that is coming from new applications, right? New workloads that are being created. And so we should -- it's the most modern -- it's the most popular modern database suite, we are well positioned for those. And then if you take the $64 billion of existing spend. That's really driven by the application life cycle, right? That $64 billion is not doing an RFP every year for every dollar of that, right? If I have an application that's up and running, like I am perfectly happy with it. And so we can have a debate or a discussion about how long is an application life cycle, is an application, 8 years, 10 years, 12 years, 15 years, 6 years, whatever it is, it's probably on balance getting shorter just given innovation cycles. For easy math, I'm going to say it's 10 years, just because it will make me be able to do the math quickly. And so that would suggest that there's roughly $6.4 billion every year that's sort of replatforming or you've got sort of an opportunity. And so it's probably more dominated by the new applications [ that deal with the ] sort of the $8 billion-ish and change a year relative to $6 billion and change a year. Certainly, we're well positioned for both, but we're particularly well positioned for the new workloads. We continue to win an increasing dollar amount of relational migrations each year. But that's maybe one way to kind of think about it. From our perspective, we're indifferent. Our goal is to make MongoDB easy for customers to consume. We have incredible developer mind share. So there's a cycle stack overflow, very popular with developers. They've run a survey, the last couple of years. And each time MongoDB has been the most popular database. The database developers most want to work with. And I think that sort of speaks to the ease-of-use, how intuitive it is, et cetera. And that's driven a lot of the popularity. And so that sort of makes sense when it comes time to tackle the new applications when developers are building a new application. And if we think about the legacy relational workloads, we announced about 1.5 years ago, the last really sort of feature gaps that we have relative to relational, which is sort of multidocument asset transactions, which is quite technical. And depending on where we want to go, we can get into it. But I think the key piece there is that was sort of the last thing that we needed to give people the peace of mind that they can move even the most demanding, even the most rigorous relational workloads. So we've had Wall Street firms move trading applications, big, large, Fortune 100 companies moving their order capture applications, office legacy systems to MongoDB because of its robustness.

Great. So when we think about MongoDB from -- you touched on the importance of developers. And I think MongoDB has been downloaded a few multiples in excess of the total number of software developers on a plane [ on this basis ].

Correct.

So the popularity is certainly very strong. But how do we think -- how do developers factor into the expansions? And how do you think about expanding on your -- into large customers. I forget the exact stats, but you have a presence in -- some sort of presence in most of the Fortune 100, for example. Is it developers that you're relying on to help you capture more workloads over time? Or how do we think about that?

Yes. So it certainly starts with developers because they're the one building applications. And to your point, MongoDB is now, from our website alone, has been downloaded more than 80 million times. And to Jack's stats, it's somewhere depending on who you talk to 20 million, 25 million developers in the world, something like that. So most developers will be familiar with MongoDB. One of our challenges is trying to make sure people are aware of the current version because we've done -- we've invested so much over the last several years and continuing to improve the product. You could have someone use the product 5 years ago and is stuck in some sort of antiquated 5-year dated view of the universe and given the rate and pace at which we're innovating, one of our challenges and one of our focal points is making sure that they're aware of the new features and focused on that. Within large customers, it tends to be a multipercent of sale, right? You're sort of getting to the developers and making sure that the developers are aware early on because they're sort of picking the database early in the life cycle, but ultimately, especially if it's a mission-critical application, certainly, if it's a relational migration, you tend to also need to be able to surround IT leadership and sort of other people with -- why is MongoDB a safe choice and helping them understand that. You mentioned Fortune 100 customers. I think we said at the time of the IPO that we had just over half of the Fortune 100. And so it's kind of an interesting way also to think about the market sizing and the opportunity, which is think theoretically, I mean, our market is obviously much bigger than 100 customers. But if you just take those 100 customers for a second, it's sort of like a little microcosm. The best we could do is not quite double our customer count, right? So when you think about sort of new customers versus expanding within existing customers is a couple of [ decent ] levers. That said, our -- then the volume that we had in terms of wallet share spend was less than 1% of the spend on databases, right? So the best you could do was more than increase your revenue from those customers by 100x. And so we're still continuing to focus on capturing new customers. But even in our largest customers, there's still a lot of opportunity with the running room that we have, just given how big the market is and [ how fractional ] that we're penetrated?

Sure. But I obviously wanted to follow-up on that, because maybe talk about perhaps the pace of expansion, because you alluded to more mission-critical apps, I think Dave has mentioned on some earnings calls about he's having more strategic conversations with folks, Chief Digital Officers, et cetera. Should we think about -- is there -- are you starting to see maybe more of a top-down mandate in terms of, hey, we need to maybe standardize our MongoDB? Or are we not there yet?

Yes. No, there continues to be an evolution. And it goes a little bit back to one of the comments that you were making earlier about the overall market dynamic. It's probably worth spending a little bit of time on this just to help people understand it because database market is a little different than some of the other markets in the software. And so what I often try and describe is that the market is not monolithic, meaning one company -- I don't know how many applications you have in Needham, but you're not running one database for every single application, right? You know, pick AT&T or Target or any company, they probably have tens of thousands of applications, and they're not running one database whereas they're running one ERP, one HR system, they have one CRM system, and those tend to be sort of monolithic applications within a firm. The database market is different in that it's built at the application level, right? And so I think that's sort of an important understanding and nuance is when you think about how do you penetrate an account, it's not that we suddenly need to go convince firm XYZ to move off of Peoplesoft and go to Workday, lock stock and barrel, but it's sort of -- there's a new application. It needs the characteristics of a modern application. We're particularly well suited. That might be the first application on our first entree into accounts that has success. You sort of build over time. It would be -- it would be atypical that a relational migration would be the first application to win, right? You sort of like build the trust and momentum, et cetera, et cetera. And then as you build critical mass and people have success, that's when customers can say, okay, we're going to declare you the modern standard or we're going to declare you or nonrelational standard or whatever it might be, but that's sort of an evolution.

Right. Okay. Well, just sort of speaking of trust and momentum, one of the other things we've noticed is that more global systems integrators are building out -- are expanding their own -- developing, expanding MongoDB practices. So could you sort of shed some light on how partners are helping you and maybe who's moving the needle for you?

Yes. So it's a really important part of the story because, from an overall market perspective, as I mentioned, we've got this massive market opportunity, you'll see -- in our reported disclosures, you'll see kind of several hundred people listed from a head count from a sales and marketing perspective, but that includes presales and marketing folks as we've been trying to build out our digital and growth marketing capabilities. So we have only a few hundred -- we've got low hundreds of quota-carrying sales reps, right? When you think about that relative to the tens of thousands of reps, a company like Oracle has. We're very, very thinly penetrated relative to the opportunity, right? One of our biggest risks is that there are opportunities or discussions that are happening that we're not a part of just simply because our footprint coverage is so thin. We don't have coverage in all the NFL cities here in North America, not to mention in geographies outside of North America and the rest of the world. So partners could be a really important sort of force multiplier for us to be able to extend our reach, become aware of opportunities, relational migrations, apps that are experiencing lots of pain that maybe we won't run in or other just initiatives at customers. And so we've seen strength really across the board, whether it's Infosys or Accenture or IBM or TCS or Capgemini in Europe or everyone else. And then we've also partnered very effectively with the different cloud players. So each of Google, Microsoft and Amazon. While, of course, there's a competitive dynamic with each of them, they've also been partners in terms of going out and selling MongoDB outlets, which is our database's service offering. They're paying their sales reps on the consumption on Atlas. We've got deals where it exists in the marketplace. And so it can burn down credits for each of those folks. And so that's another way to sort of extend our reach, given how thinly covered we are.

Right. Okay. So I wanted to shift the discussion a little bit and just talk about your product differentiation and its impact on the business. I mean, one of the differentiators of MongoDB is that it works across clouds. You've got multicloud capability versus some of the other competing products in the market. But what does that translate to in practice? I mean, do you see, for example, larger ACV deals from customers who are using it for multicloud capabilities? Does that -- does it matter at all? Is that the right way to think about the business?

Yes, I think it matters a lot. I think, first of all, at conferences like this or meetings, we talk a lot about the cloud and everything else. But then you go out and you talk to customers and customers are -- in large enterprises are still pretty early on in their cloud adoption. And I think that, that's easier for people to forget, but it's still pretty early on. That said, as they're starting to increasingly move to the cloud. Large organizations are exceptionally afraid of vendor lock-in. And what they're really worried about is not being able to move from one cloud to the other. So that's why multicloud is so important. And the database layer is particularly important in that. And customers are exceptionally worried about having a proprietary database that is unique to that cloud vendor, right? If I'm running on a proprietary Amazon or Microsoft Database that's unique to their cloud, I can't bring it back on premises, I can't run it in a hybrid environment, I can't port it to another cloud player. It's relatively easy to move the application executable code, but if I actually want to rebuild the application from a Azure proprietary database to a AWS one, like that's expensive and complicated. And so we've seen people be very, very, very hesitant about that, which is part of what MongoDB brings to the table in that we provide the flexibility where they can run it across all 3 clouds. They can run it in a hybrid environment, they will bring it back on premises, and that's sort of future-proofing and avoiding a vendor lock-in is really quite critical as people are increasingly looking towards running more applications in the cloud.

Sure. Well, just -- and when we think about the growth of Atlas has been, frankly, extraordinary. Does it matter where people are running Atlas? I mean, is there a way to think about market share of we're -- by public cloud vendor? Is that important to you at all?

It's not important to us. Typically, people have a multivendor cloud strategy. So a large organization will have sort of a primary cloud vendor, and a secondary cloud vendor and they'll sort of have worked through that. And so we're indifferent. We launched first on AWS and then rolled it out to Azure and Google. And so Amazon is probably a little bit ahead in terms of the market share within MongoDB. But from our perspective, it's not relevant. And then when -- and we partner with all 3 well. And then when you look at more broadly in terms of Atlas, we have the self-service side of the business, which is people signing up for credit cards, which at this point is all principally outlets, which is fairly straightforward and we have all that in our disclosure. I think in terms of when you think about how much of the business becomes Atlas over time, I think it really depends ultimately on your views on how quickly do enterprises adopt public cloud, right? Because while we talk about and provided a whole bunch of disclosure around Atlas on a product level, we really think about the business on a channel basis, right? We think about the direct sales channel, right, so when we're going out and interacting with customers directly, signing contracts, et cetera, et cetera. And then -- or the self-service side of the business, which is someone signing up for the business with a credit card. And so the mix -- and the self-service side is, like I said, almost entirely Atlas and then the mix of -- in the enterprise side, is really all about how cloud forward they are and what does that cloud adoption cycle look like. Because our goal is just to make MongoDB to be easy for the customer to use. They usually have an IT strategy. They have a point of view that says, we want to run this application in the cloud or we want to run this application on premises. And so we're not trying to drive our salespeople to skew the conversation to one way or the other. We just want to make sure the Atlas is easy for the customer to consume and about 1.5 years ago, I think it is, we introduced the enterprise features into Atlas, which makes from a feature parity standpoint, they're equal. And so it's really about just whatever their IT strategy is?

Right. Well, with this all that being said, you did announce a new relationship recently with Alibaba. And so how should we think about -- what that means for MongoDB?

Sure. Yes. So MongoDB Atlas is already the most widely available cloud service -- database and service, and we leverage each of the 3 major cloud providers, Amazon, Google and Microsoft. None of those folks have direct offerings in China, given the restrictions that existed in China and some of the Chinese players like Alibaba and others have been very popular cloud services there. We entered into a transaction, a relationship in October, whereby Alibaba is now licensing MongoDB to cover their MongoDB service. Historically, they had been providing a service, but it was unlicensed, unauthorized. And so after some discussions, we helped them understand that our licensing model is a little bit different. And while they may be familiar with open source software and they're used to not having to pay for open source software, our licensing model was different, which they came to agree. And they saw an opportunity to differentiate themselves in the market. And I think it's unusual that a company would have an existing business and choose to add cost to it, which effectively is what they've done. And I think it really speaks to the popularity of MongoDB. China has been one of the largest markets for us for downloads. There are an incredible number of developers there. We've had a lot of success with our field team there in selling different customers, but haven't been able to have a cloud offering in market and so this really changes that equation, which is great. So the awareness to developer mind share and then the strength of the IP licensing model are what sort of created this sort of unique opportunity for us to partner with them. It's a multiyear deal with annual commitments. It -- are -- the license fee they paid us increases as volume increases so we share proportionately. And both sides are optimistic and hopeful that we'll be exceeding the minimums. We've been super clear just that it had no impact in Q3, and there'll be limited impact in Q4 and fiscal '21 as the relationship ramps over time. But from a long-term standpoint, I think it's pretty differentiated. It's a pretty unique situation. I think, reflects the strength of our developer mind share, the business model, the IP and overall the product. So we're looking forward to it.

Right. Maybe if you could just touch on briefly the importance of your licensing model because it's different than the way most people think about open source, and that certainly plays into the discussion we just, I guess, had with Alibaba.

Yes. Sure. So most -- just a quick kind of recap of like what a typical open-source model would look like. A company -- there would be some interesting technology built inside of Google or Facebook or Yahoo! or someone like that. And they would take the technology and decide it's not core to what they do, and they would spin it off and they would say, "Hey, community go make this great," right? So illustratively, Yahoo! creates Odoo, decides it's noncore and just spins it off and say's, "This is really interesting, like, let's go make this technology better." And so what winds up happening is it gets open to the community, no one really owns the intellectual property. No one really controls the product road map. No one really controls what goes in the free version, what goes in the paid version. And because we've designed it to crowdsource, the development of the software, it has to be available under a highly permissive license, right? So for everyone in this room to be able to edit and freely modify and incorporate the software, we need property -- intellectual property that doesn't constrain any of that and doesn't prevent us from building great software. And that's pretty typical. But the consequence of that is that the big cloud players, because the software -- the IP rights are quite permissive, can take the code, package it up, run it as a service and monetize it quite effectively to the detriment of the community or anyone who's tried to sort of make real investments in the product. So MongoDB is different. We built MongoDB ourselves. There was no prior art that existed that we're leveraging. We built it all ourselves. We understood that developers were increasingly looking to open source technology and wanted to sort of test and experiment with software. And so we created an open-source version of the product. That's really much more analogous to a freemium offering to drive adoption, usage, mind share, et cetera, et cetera. People have done that. But we were able to do that because we weren't looking to crowdsource the development, we were able to do that under a much more restrictive open source license. And therefore, make sure that none of the cloud players package it up as a service, which is why you don't see Amazon or Google or Microsoft, running MongoDB as a service and why you see now Alibaba not running MongoDB as a service. And so the result is that we own our own intellectual property. We control the product road map, we decide what goes in the free and paid version and we have very strong and protective intellectual property rights.

Great. Thanks for that overview. So I just -- I want to touch on trends in your Enterprise Advanced product for a moment. Now that the Atlas and Enterprise Advanced have reached feature parity. You touched how Atlas is really sort of transformed your overall organization. How do we think about that business moving forward?

Enterprise Advanced?

Yes.

So a couple of things. So one, I think it really is a core building block and growing quite nicely, as you can see in the financial results that we publish. As I mentioned earlier, I think, ultimately, it really depends on how quick they do enterprise as a top public cloud, what the mix between Atlas and Enterprise Advanced is. And I'll say one thing that's maybe helpful. We have a fair amount of disclosure. And starting with the IPO, we provided a lot of insight into Atlas to help people understand in part because it was new, in part because it was different, but also in part because it was materially lower gross margin. So Atlas' gross margin profile includes the underlying infrastructure spend and as a result, this lower gross margin percentage. We've now introduced Enterprise Atlas and on apples-to-apples kind of future equivalent basis, Atlas represents revenue uplift, gross profit uplift, but it's gross margin percentage dilutive, relative to Enterprise Advanced. But I think fundamentally, I think there's a risk that we've sort of unintentionally steered investors to overly focus on the product view versus the channel view. And we think about the business from a channel perspective as I said. And I think the product distinction, while it's interesting and Atlas grabs a lot of headlines and is growing incredibly well as is Enterprise Advanced, but Atlas' growth is sort of particularly stellar. The rationale for sort of calling out the difference is a little less important because we made so much progress on the Atlas gross margin. And so I think, historically, at the time of the IPO, if you'd told me, we were going to report a quarter, this third quarter that we just reported, where Atlas will be 42% -- 40% of revenue and we'd have 72% non-GAAP gross margin, I would be like, we wouldn't be able to pull that off, right? And so I think we've done a really great job on the gross margin story. And the result of that is while people tend to focus on Atlas because it's sort of the easy thing and we make it easy for you and the releases and everything else, what mix Atlas is or isn't to the business, matters a lot less than it used to, given how much progress we've made on the gross margin side. So I think, ultimately, the gross margin -- I mean, ultimately, the mix shift will depend on enterprises and how quickly [ they got ] public cloud.

Sure. Well, I guess since you mentioned gross margin, the other thing that's impacting you, I guess, is the acquisition of mLab, a little over a year ago. So I mean just from a longer term, is there -- do you have updated thoughts in terms of how we should think about the trajectory of gross margins going forward now that you've sort of updated us on the intricacies of what's happening today?

Yes. So fundamentally, we've seen some slight degradation. We are not at the bottom and have said that we're not prepared to call the bottom. I think we'll sort of continue to have this sort of shallow view that we've been going through. Specifically, on mLab, that's worth calling out. So we acquired a company called mLab on November 1, 2018. So this fourth quarter that ends in a few weeks when we report, that will be the first quarter where mLab [ is in the base ]. We had talked and made it clear at the time that when we made the acquisition, when we announced it, that we were expecting the mLab revenue base to shrink over time, primarily for 2 reasons. One, any time you're doing platform consolidation and everything else, there's a risk of just people [ trading ], who have relatively small workloads or have forgotten that, that's hitting their monthly credit card. When they're asked to sort of [ re up ], they say, "Wait a second, this is invaluable." And we baked all that into our diligence in our guidance and the price that we paid. And then the other thing is, we intend to give customers sort of the better pricing. So if you -- if mLab is cheaper pricing than Atlas, you get to keep it, so that would theoretically have no revenue impact because you're paying what you're paying. But if Atlas happens to be cheaper, which it is, in some cases, we'll actually give you the benefit of the lower Atlas pricing, which obviously would result in some price deterioration because you'll be paying less than what you used to be paying. So for all those reasons, we kind of communicated that out. And like I said, that was factored in. And that's been performing well relative to our expectations.

Okay. Great. So you recently raised a significant amount of capital, $1 billion convertible notes offering. Can you just talk about the rationale for that? And how do we think about the potential use of proceeds there?

Yes. So this is the second convert issue that we did. We did this -- we announced this last week, very opportunistic capital raise, taking advantage of incredibly low cost financing. It also gave us an opportunity to repurchase about 70% of the prior convertible, which had a -- when you take the cap call into effect had roughly $106 conversion price, and so we're able to move that point of equity dilution, closer to $300 a share. So we feel good about that. And then it added some incremental opportunistic capital to the balance sheet. We -- as a public company, M&A is not our core driver to the strategy. We've pursued it somewhat opportunistically. We've made 2 small acquisitions, each under $100 million, but that did deplete a little bit of the cash, and we wanted to make sure that we had sort of cash on hand to be opportunistic, [ while ] there are no specific M&A plans.

Right. Okay. And then, as Atlas continues to become a greater portion of the overall mix. Can you talk about what the impact that has on your business front. Atlas is a usage-based model, built in arrears. How does that impact your ability to forecast the business? And do you think billings, for example, is becoming a less relevant metric for how to think about MongoDB?

Yes. So a few things. So first, on Atlas, to Jack's point, Atlas is consumption-oriented. And so as customers consume and use Atlas, we recognize the revenue. On the self-service side of the business, that tends to be fairly straightforward in the sense that you've got like a large number of relatively small spending customers. And so you've got some portfolio theories working for you in terms of forecasting. On the enterprise side of things, consumption tends to be fairly straightforward. And when it hasn't been, we've called it out. So a year ago, in Q4 of last year, we had some customers consume meaningfully in excess of their contractual amounts that we have -- would not recur in this coming fiscal year. And so there is that level of visibility. Nothing is as straightforward and clean as software revenue recognition, 605, good old ratable revenue recognition. Unfortunately, we don't have that anymore and then there's 606 for the Enterprise Advanced portion of the business, there is some lumpiness. So the term license component we're required to recognize at the start of the subscription. And that's approximately 1/4 of the revenue stream. And so that introduces some lumpiness, some increased variability and also reduced comparability when you think about both on a sequential basis as well as on a year-over-year basis. And so that's probably one of the factors that we tried to articulate even more, just so people have that understanding of that visibility. So we called out last year in the second half of the year, we had a lot of multiyear contracts that sort of have that kind of impact. And then lastly, on your billings question, we've thought for a long time, billings is not a great way to look at our business. In particular, we said in Q1 that about 2/3 of Atlas doesn't flow through deferred revenue. So that makes the billings calculation, particularly challenged. And I think that will be sort of increasingly so, that it's less useful. The last thing I'd say is, on the chance that you're going to look at it anyway, at a minimum, I would tell you to not look at deferreds in isolation because that will be particularly confusing. So I'll probably leave it there.

Okay. We are out of time, but thank you very much, Mike, for joining us today.

Thanks for having me.

Thank you for your time.