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

Good afternoon, everyone. I'm Jason Ader, and welcome to our virtual fireside chat with MongoDB CFO, Michael Gordon; and MongoDB VP, Finance and Business Operations, Serge Tanjga. How do I say that, Serge?

Tanjga.

Tanjga. Okay. So first of all, thanks, Michael and Serge for being with us.

Thanks.

Thanks for having us.

Before we begin, I'm required to inform you that a complete list of research disclosures or potential conflicts of interest is detailed on our website at williamblair.com. [Operator Instructions] And with that out of the way, let's dive into questions.

So first, at a high level, Michael, what is the secret to MongoDB's success over the past decade? And how has your value proposition evolved over time?

Sure. Yes. So the MongoDB story for those who don't know, we make database software and have the leading modern general purpose database. And we've been continuing adding on additional products, which we sure will talk about, to build out through data platform. The database market is one of the largest in all of software, roughly $62 billion per IDC's estimates in 2020, growing to $89 billion in 2024. So an incredibly large market, mostly dominated by traditional legacy vendors, and that's part of what creates our opportunity. So the combination for us that's worked well has been outstanding new technology matched with very large market, demonstrating a really, really strong product market fit. One of the drivers for that has been how easy and intuitive MongoDB is to work with for developers as they build software applications. And that's led to MongoDB being the most preferred database by developers. So Stack Overflow runs a very large survey every year. In this past year, again, for the fourth year in a row, we were awarded the database developers most want to work with, right? And that's because of the ease of use, the scalability, the flexibility. And when you think about data today and applications today, that's fundamentally at the heart of what all businesses are competing on, right? You hear these phrases like software is eating the world or every company becoming a technology company. And really that's shorthand for the software being the defining piece of competitive edge, whether it be external customer-facing or internal. And at the heart of every one of those software applications is the database. And so I think it's a great market, tremendous -- large market, tremendous product, leading to great product market fit, huge developer following and affinity and understanding matched with really solid execution has put us in the position we're in.

Great. And then I know historically you guys have focused on the operational database market. First, maybe help people understand the difference from operational database to other types of databases. And then secondly, what are the areas that you're broadening into from the operational database?

Yes. So there are lots of different ways to sort of cleave the market and to think about it, sort of divvy it up. People often will talk about systems of record, systems of engagement, systems of intelligence and systems of insight, but there are plenty of other different ways to go in it. I think the core, from a history standpoint, has been really to, as you mentioned, the operational side, that tends to be either revenue-facing applications, customer-facing applications or critical internal applications. The first use cases of MongoDB were generally not as systems of record. Although, over time, we've continued to evolve into those. I think one of the most important things to think about and to understand about the MongoDB story is the general purpose nature of what we do, right? We are not the only company that saw this big $60-plus billion market that had been relatively unchanged over the last 4 decades, and with your good economics hat on, you should expect large markets that are ripe for competition should attract capital, right, and should attract competition. I think one of the things that's been different about our approach is we set out to build a general purpose database. So not a database that's just particularly good in only one use case, but instead has a broad range of applications. And so whether it be systems of, like I said, engagement or systems of record or systems of intelligence, which is sort of some of the newer stuff that's coming up, or even insight, which is as we move into the analytical side of things, MongoDB can be quite relevant and helpful for folks across that. So really, the whole market over the long term is what's available to us.

Got you. And what are some of the new products that you've introduced? I know you had MongoDB World this week. What are some of the, let's call it, add-on capabilities or augmenting capabilities that you've added to the core database?

Yes. So the core database was the initial product that we started at MongoDB. We sell it both in a subscription license and/or growing license for those customers who are -- want to manage their own environment, whether that'd be in their own data centers or otherwise. And then we also have MongoDB Atlas, which is our database's service product, which we launched about 3 years ago and is now 42% of revenue and growing quite quickly. We had also introduced Atlas Search and Data Lake, which are additional sort of adjacent use cases, all under the umbrella of making a developer's life easy. So last year, we introduced both of those products into beta. They're now in GA, and they're really just extending the value proposition for helping developers and people who are working with data making data stunningly easy to work with. Data is usually the problem, right? It should be the opportunity because of the power and the insight that it has. But if you're a software engineer and you're building applications, data really complicates things. And so MongoDB DB helps us streamline that and make that easier. We also announced some product introductions and GA and beta related to mobile. So mobile is increasingly an incredibly popular area for developers, and applications need to have a mobile orientation. So a little over -- I guess a little over a year ago now, we bought Realm, which is a leading independent mobile database. And we've done work integrating that into the rest of MongoDB. We also introduced, I think, a beta synchronization product and some other things because one of the biggest challenges with mobile applications is when you're off-line, you don't have connectivity, at some point, presumably, we'll start taking [ airplane trips ] again and things like that. What do you do from a syncing, right -- the sync component of a mobile application is incredibly important. And so we introduced a product in the beta there. So a wide range of continuing to push out the relevance, not just beyond the core -- not just the core database, which is obviously important. We also introduced some new functionality into -- around the broader ecosystem data platform to help developers more easily and cover more ground. So they can spend more of their time on building out the application functionality, the user experience, things like that, rather than having to worry about some of the back-end stuff.

Good. And you mentioned about systems of record earlier. Can you talk about how you guys have been able, over time, to address more of those types of use cases and maybe what were some of the gating factors historically for you to really be a major player there?

Yes. So without getting like overly technical, I would say one of the biggest -- any system of record is obviously incredibly important to an organization or any institution. And what we've seen over time is database is a complicated product. We've now spent 10, coming up on, I guess, 13 years, 10-plus years, building out the database. The last key thing that sort of folks were looking for is called multidocument asset support. And so happy to get into that if it's relevant or interesting to folks. But we introduced that probably a little over a year ago now. And so if you think about the technical reasons, right, other than inertia or other things like that, but from a technical standpoint, we've now sort of removed the critical -- I think, critical potential blockers or objections that someone might have had to running regardless of whatever, no matter how mission-critical the use case, their data on MongoDB because ultimately, what people care about specifically when you're thinking about a system of record is what are the levels of data guarantees that you have, right? And that sort of multidocument asset transactions is sort of the gold standard. And we announced, like I said, a year ago that, that was -- that we had that across the offering.

So you're seeing good adoption -- or has adoption changed since you introduced that for those types of use cases?

I think it's probably 2 years ago, from when we announced it, and I think the announcement itself was, in some ways, more important than the release, although obviously, following through on your announcements is important as we've done. But I think it's -- for most applications, developers are not using the most current version of the database, right? They tend to lag a little bit in terms of upgrades and deployments just because databases are so central -- are such a central component of the application. But I do think that when we announced multidocument asset transaction support coming, I think that really was an eye opener for a number of people, particularly for people who had maybe known or experimented with MongoDB years ago but had not kept current on the pace of innovation and the evolution of the product over time. And so I think that catalyzed people to sort of revisit and reevaluate that because typically for the sort of newer breed of databases, which we're the clear leader, historically, hasn't -- hadn't and hasn't offered that. And so that was a real change of perspective or change of framework and really underscored this sort of general position -- general purpose positioning that we have and the fact that we can sort of support mission-critical workloads regardless of what the use case was.

Good. And how much of your business today comes from new apps versus legacy migrations? And how has that been trending?

Sure. Do you want to comment on that Serge or you want me to?

Yes, why don't I go? So what we said at the time of the IPO was that about 30% of our new business comes from replatforming of existing applications. And that business has continued growing quite nicely since the IPO. But as a percentage of our total business, it's actually declined. That's a function of the fact that Atlas, our Database as a Service offering, tends to more naturally attract new workloads. So it's more a function of a mix shift versus sort of the continued robust growth in the replatforming business.

Understood. Thinking about competition here. Actually, I have a question from the audience, which I want to get to as part of this discussion. But who are your main competitors today? And really, how do you differentiate your offering?

Yes. So I think there's been really 3 main buckets. You got the legacy incumbent players, best personified by Oracle, where the vast majority of the market share and dollars are today. I think the key differentiator there is better, more modern technology that's more nimble and agile, right? Oracle and the rest of the relation universe has these rigid set of tables of rows and columns that are challenging to scale and difficult to iterate on and, as I mentioned earlier, create a fairly heavy cognitive load for a developer, figure out how do I go from my object-oriented programming language to map it back to this structure of tables of rows and columns, whereas MongoDB has a document model, and that document model is much more analogous or similar, consistent with object-oriented programming languages. So those are the big incumbents. That's where the biggest market share is. And so it's really a question of mind share and the passage of time frankly as applications sunset. One way to think about the universe, as I mentioned, is sort of the overall market sizing, right? So $62 billion growing to $89 billion over 4 years, we've got effectively $27 billion of new business being created over that 4-year period. So just under $7 billion a year in new business. And then as you look at the other $62 billion, I'm sure we could have a healthy debate around application life cycles. But just for simple math, I'm just going to say average application life cycle is 10 years, which means that there's $6 billion that's kind of coming up for grabs every year, right, as applications reach their kind of end of life. And so we've got about $13 billion every year that's sort of up for grabs in this very large market. And so I think over time, the incumbents who have the dominant share of that $62 billion have some vulnerability. And we've had success, as Serge was just saying, at sort of chipping away at that. There, as I mentioned, we were not the only one who recognize the opportunity. Although I think when you look at the rest of the crowd, I think we pretty meaningfully distanced ourselves in size and growth relative to everyone else. So we don't run into kind of the other players who were sort of founded 10-plus years ago. So we don't really run into them or see them, although it's a large market. And then the third big bucket is the cloud players. And so Amazon, Google, Microsoft, et cetera, all have database offerings, all would love to sell all possible services to all potential customers. And that's really been an example of both hard competition but also some cooperation. We have joint sales agreements with all 3 of them. We've had joint successes with all 3 of them. And so that's really the kind of high-level summary of the competitive universe.

You guys just had like a -- won an award at Google, as you saw, right?

Yes, I think we were their marketplace partner of the year or something like that. Yes.

Okay. Does that mean you were the #1 product sold through the marketplace? What does that mean?

I don't actually know if it's an entirely quantitative criteria, although it wouldn't surprise me. I certainly wouldn't want to speak for them in terms of their numbers. But yes, I think we've been able to have a big impact on their platform. I think the teams work quite collaboratively together. I think Google has taken a slightly different strategy than Amazon and Azure, particularly as it relates to the database market, both -- given the popularity of MongoDB, both Amazon and Microsoft have attempted to create a kind of imitation or competitive offerings that don't use our intellectual property because of our intellectual property rights, but they've seen the popularity. So they've tried to mimic or imitate the look and feel with what we hear from customers as an inferior product. Google has said their strategy is more about taking best of breed, and they look at MongoDB as a best-of-breed modern database technology. And so rather than trying to copy or imitate it, they've decided the best thing for their customers is just to offer them the real thing. And so we've been jointly working with them just like we've done with the other 2 but have had a lot of success with them. And I think some of that culminated to them giving us this most recent award.

And how do you differentiate versus the cloud guys?

Yes. So there are a couple of factors. I think if you're thinking about their direct competitive offerings as I mentioned, they're imitation offerings. So they can't be based on our own intellectual property. And so they're trying to imitate or mimic what working with MongoDB looks like. And as a result unfortunately for customers, there's a risk of confusion but also of poor performance. So it will fail a fairly high percentage of our correctness framework. And so ultimately, it's really a quality issue. The product quality just isn't there. Some of that's definitional in the sense that they're trying to copy our product. And so it sort of has the lag and can't be pushing a little forward, but also it is a little clunkier because they can't use our code. I think away from their MongoDB directly related products, they have other databases. I don't know how many databases Amazon is up to, but something like 15 or so in the offering. I think from their perspective, it's really more of a shelf space game, right? They just want to make sure to keep you in their ecosystem, right? They want the storage and compute and all the other value-added services or ancillary services that they can sell you. And so they want to make sure that they have one of every flavor of database or anything else. I think the reality is they understand, and certainly I've heard this consistently and loudly from customers, the customers are very concerned about vendor lock-in, in general, with the public cloud players but particularly at the database layer. And so the concept of picking a proprietary database that I can only run in Amazon or I can only run in Azure is quite daunting in terms of the magnitude of flexibility that they're surrendering. And so separate from all the product advantages that we have, I think there's also just some marketplace structure advantages that we have that inhibit people from being kind of -- I can be all in on a cloud vendor, but that doesn't mean I'm going to be all in on their database, right? I might like the rest of their stack, but ultimately, I'm probably going to have a primary, secondary strategy where I may run 80% on AWS and 20% on Azure or GCP or whatever the structure is, but that's a very different -- that's perfectly fine as it relates to storage and compute, right? I can kind of move the application executable relatively easy, but to actually have the application itself built on a proprietary database is pretty sticky and pretty cumbersome. And I haven't seen and I haven't heard customers who are eager to do that.

And then a question from the audience here is on Snowflake. Is that someone that you view as a competitor today?

Yes. So not a competitor, sort of in an adjacent space. I think, my take, and I certainly wouldn't want to speculate on the questioner what they're asking or where they're coming from, but my guess is, as we've announced products like Data Lake, and Snowflake also, I'm sure, has an ambitious road map and everything else, it's relatively easy to confuse the 2 directions. I think the key thing is Snowflake is playing in sort of the modern data warehouse space, where you have a known set of requirements that modern data warehouse or data warehouse in general got at the start because people couldn't run applications or couldn't run queries off of the operational database because it was slowing the operational database down, so it needed to offload things to a data warehouse. And then I would batch out the data to that data warehouse, so I didn't tax or impact the load or the performance of kind of my regular application, regular database. And so data warehouses were used for fairly known consistent reporting. I had to have a schema where I set it up because I knew I want to run these reports and everything else, whereas data lakes have cropped up for a little bit more exploratory, less known stuff, right? And so what our Data Lake product does is that rather than having to spend a lot of time and money and unique expertise on deploying a technology like Hadoop or some of these other things, I can very easily, with my MongoDB existing skills, have that developer run analytics not just on the operational database, meaning MongoDB because we have a thing called -- we have a capability called workload isolation, which will allow you to run those queries on the operational database without actually impacting database performance. And so you can not only do that, but then you can also read data off of some sort of common online store or online archive like an S3 bucket or something like that. And so I can get the value and insight much faster. I'd say the other big difference is we continue to be focused on the developer, whereas companies like Snowflake are much more focused on an analyst or a data analyst who's using other tools, Tableau or things like that, who's used to work with data warehouses and maybe writing a lot of SQL or things like that. There are certainly small areas of overlap. So we have a Charts product that's around business insights and visualization and things like that. And so yes, theoretically, there's a little bit of overlap. But in general, they're pretty different.

So you don't see them...

And we don't see them -- I don't see them competitively. I can't think of a single competitive situation that I've heard of that we've seen them on. They come up a lot in investor calls because, obviously, even though they're a private company, they have some level of awareness, and markets are hard to -- harder to parse out when you get into that granularity. So I totally understand [ and buttress some ] further question. But from a marketplace standpoint, we really don't see them.

Okay. Let's switch gears to the quarter you just reported last week. What are the big takeaways, I guess, from Q1, especially with respect to the impact on the business from COVID?

Yes. Serge, do you want to run through that?

Yes, why don't I start? So first of all, we were very pleased with the quarter, strong results and sort of well ahead of our guidance. Subscription revenue growth of 49%, very happy with that outcome. We -- if you sort of -- if you go back a quarter and talk about March, when we first provided the guidance for Q1 and guidance for the rest of fiscal year '21, what we said was we expect to see a business impact from COVID. We expect that impact to be material in Q1 in terms of the bookings, followed by a more material impact in Q2 and then a normalization in the back half of the year. What ended up playing out was 2 things. Number one, our ability to attract new business exceeded our expectations when it comes to Q1. And that's true of both our direct sales business and our self-serve business. And I think that just frankly speaks to 2 things: number one, the popularity of the technology; and number two, our ability and the ability of our teams to sort of very quickly pivot and execute well in a challenging environment. So that was the major positive takeaway from Q1 in terms of ability to attract new business and new customers across the company. Where we did see an impact from COVID was on the Atlas business, when it comes particularly to our existing customers. When the global lockdown began, we saw the growth from those customers slow down a little bit. So the growth continued, but it was at a slower pace than historical. And when we dug into it, we realized that it was a modest slowdown across a broad base of our customers, which when we sort of further diagnosed it, we very strongly believe has to do with the economic consequences of the slowdown and sort of the second-order -- we are experiencing second-order effects as customers' businesses slow down themselves. So that was a slowdown that -- and it's a reflection of a pretty diversified customer base that we have in Atlas. And so as we look forward for the rest of the year, we do expect to see a bigger impact to new business for the rest of the year as this crisis continues, and that's simply a reflection of the fact that we benefited from our ability to close a strong existing pipeline in Q1. And obviously, that becomes more difficult as the year goes on, and the macro continues to be challenging, if not more so. And then we do expect to continue to see this lower-than-historical growth in our customer base, and that's sort of how we triangulate to our view for the rest of the year.

Okay. A question from the audience here. It's about Atlas, and it's about Atlas self-service versus Atlas direct. Can you talk about the differences between those 2 segments and the types of customers and market sizes for those 2 areas?

Yes. So maybe at a high level, just take a step back and very much appreciate the question because one of the things it does is it underscores that we've provided a lot of disclosure, and we've tried to educate folks about the business. But one of the things that's important to understand is we look at the business on a channel basis, so direct sales versus self-serve. We have a lot of disclosure on a product level basis just given how quickly Atlas is growing and particularly at the time of the IPO, just the meaningfully divergent gross margin profile that Atlas had. We thought it was important to provide visibility and understanding to that. But I always worry that there's a risk that it sort of over-rotated or has an unintentional consequence for people disproportionately focused on a product view of the world rather than a channel view. So I think the question is great because it helps remind people that the channel view of the world is really the right way to think about it. Within the channel, I think from a volume perspective, clearly the self-serve channel is much larger, but the spend levels per customer are much smaller. So you can see in our publicly reported disclosures that the average direct sales customer spends north of $100,000 with us, on average, 6-figures-plus a year. Whereas the average self-serve customer spends between $6,000 and $7,000 a year, right? So meaningfully different spend levels. It doesn't mean the application is any less important to them, but they're just smaller businesses and smaller operations. From a revenue perspective, all of Atlas regardless of channel, Atlas is really consumption-oriented. And so if you think about a new workload coming on, whether it's a new workload that I know is going to be successful, whether it's a migration, and it usually would start at smaller spending levels and build over time. And so we tend to see that very strong expansion that Serge was talking about. And in Q1, we also saw healthy expansion, just not as significant, as healthy as we normally see. In terms of the -- I'm trying to give other sort of interesting economic or financial attributes. I guess the other thing I'd say about self-serve customers is the self-serve cohort, because it's an experimentational component, right, this is someone who's signing up with a credit card and everything else is coming and they're joining the platform, their behavior as a cohort is quite strong. The expansion of those cohorts is really attractive. But you will see some churn, right? Some of those people are just experimenting. They don't know if the application is going to be wheeled, maybe they're testing it out. And so there's a little bit of underlying different behavior. I guess the last thing economically that's probably worth calling out and making clear to people is there are different financial or payment relationships. So self-serve customers are swiping on credit card and paying every month's invoicing in arrears. Our direct sales customers have historically paid annually in advance. We've been increasingly trying to find ways to reduce the level of friction, given that once someone starts on Atlas, the expansion is quite healthy. And so we've seen, especially in the mid-market, some set of customers moving to monthly invoicing. And so we said last year that about 2/3 of Atlas did run through deferred revenue. We don't guide to billings, we don't focus on billings, but know a number of you do, and wanted to call that out because you'll see an increase in divergence. We don't think it's relevant to start, but it will become increasingly less relevant as we continue to evolve the business.

And self-service today is about 20% of your total revenue?

Yes, just over 20%.

Just over 20%. Okay. Good. All right. Well, we have about a minute left. I wanted to just, Michael, ask you if you have any final thoughts that you wanted to leave the audience with.

No. I think maybe just a quick kind of additional follow-on to what Serge said. We felt really good about the quarter. One thing we've heard from the conversations today and then the follow-up conversations we have with investors post earnings that's maybe helpful to some is obviously we don't comment on [ year-all ] expectations and everything else. We felt -- we really felt really good about the quarter. I think one thing that was interesting that's probably worth like addressing and clarifying is I do think some people seem to think that because Atlas is cloud-related that it would categorically would have this reflective increase in consumption. And I think that's true for a lot of software applications. But I think you need to think of Atlas as a little bit different, in that it's not quite as quick twitch, right? It's not as simple as just sort of saying, "Oh, suddenly we all work from home, so we need video conferencing software that I can deploy enterprise-wide or we need collaborative software that we can deploy enterprise-wide." All of the trends that we were already benefiting from that have been highlighted by coronavirus and the shutdown accrue to our benefit in the long run. It's just a longer gestational period within, okay, now I want to go build an app and launch the app and start consuming the app and everything else. And so I think it's helpful and sort of incrementally additive to our tailwinds, but there's not quite as rapid a response. We don't quite benefit from that student body left shift dynamic that some others do. And I think there are some people who just didn't fully appreciate that about how the database market is different, which I don't blame folks because it's not like there are 20 different publicly traded database companies that you can go like spend time and educate yourself around this that's worth it. So that's probably the last thing I would just add to help people's understanding.

Good. All right. Well, Michael and Serge, thanks very much for joining us today. I know you had a long day. I appreciate the time and stay healthy.

Thank you. Likewise.

Happy to. Thanks.

Thanks, everyone else, for joining us, who's on the line.

Yes. Thank you for the time.