Good afternoon. Welcome to the MongoDB third quarter fiscal 2020 earnings conference call. All participants will be in listen-only mode. Should you need assistance, please signal a conference specialist by pressing the Star key followed by 0. After today's presentation, there will be an opportunity to ask questions. Please note this event is being recorded. I would now like to turn the conference over to Brian Denyeau of ICR. Please go ahead.
Thank you, Andrea. Good afternoon, and thank you for joining us today to review MongoDB's third quarter fiscal 2020 financial results, which we announced in our press release issued after the close of the market today. Joining the call today are Dev Ittycheria, President and CEO of MongoDB, and Michael Gordon, MongoDB's COO and CFO. During this call, we may make statements related to our business that are forward-looking under federal securities laws.
These statements are made pursuant to the safe harbor provisions of the Private Securities Litigation Reform Act of 1995, including statements related to our financial guidance for the fourth quarter and full year fiscal 2020, our market opportunity, and our ability to capture a meaningful share of the global database software market, the opportunity created by and scalability of our go-to-market and growth strategies, the potential advantages, timing, and likelihood of success of our new products, product enhancements and planned integrations, including as a result of our recent acquisition of Realm, the anticipated impact of Atlas sales expansion on our gross margins and other financial results, and our anticipated investments in the business and their impact on future cash flows. The words anticipate, continue, estimate, expect, intend, will, and similar expressions are intended to identify forward-looking statements or similar indications of future expectations.
These statements reflect our views only as of today and should not be construed as representing our views as of any subsequent date. We do not have plans to update these statements except as required by law. These statements are subject to a variety of risks and uncertainties that could cause actual results to differ materially from expectations. For a discussion of material risks and other important factors that could affect our actual results, please refer to those in our quarterly report on Form 10-Q filed with the SEC on September fifth, 2019, and our other periodic filings with the SEC. These documents are available in the investor relations section of our website at www.mongodb.com. A replay of this call will also be available there for a limited time. Additionally, non-GAAP financial measures will be discussed on this conference call.
Please refer to the tables in our earnings release on the investor relations portion of our website for a reconciliation of these measures to the most directly comparable GAAP financial measure. With that, I'd like to turn the call over to Dev.
Thanks, Brian, and thanks to everyone for joining us today to review our third quarter results. MongoDB continued to deliver terrific financial and operational results that further strengthen our position as the most popular modern data platform in the world. Our performance is driven by the unique capabilities of our technology and our go-to-market strategy that enables customers of all sizes to consume MongoDB how they see fit. We are in the early stages of a once-in-a-generation shift in database technology, and we remain confident in our ability to capture a meaningful share of the $64 billion global database software market. Looking quickly at our third quarter financial results, we generated revenue of $109.4 million, a 52% year-over-year increase and above the high end of our guidance. We grew subscription revenue 56% year-over-year. Atlas revenue grew more than 185% year-over-year and now represents 40% of revenue.
We had another strong quarter of customer growth, ending the quarter with over 15,900 customers. This was another outstanding quarter for Atlas, which is now $175 million revenue runway business, just over three years from launch. That's only the tip of the iceberg. Atlas is driving change across our entire business as we increasingly become a cloud-first company. I would like to showcase three ways in which Atlas is making us a better company. As our business shifts towards the cloud, we are able to increase our pace of innovation with new products, features, and capabilities. Cloud services give us better insight into how customers use our products, and cloud services enable us to expand the different ways we go to market, increasing our ability to pursue a larger variety of customers around the world. All three together create a flywheel effect for our business.
First, since we deliver MongoDB as a cloud service through Atlas, we can deploy new features continuously rather than on an annual release cadence of an on-premise product. Furthermore, since new products are initially built for the cloud, our product teams can work quickly and independently to introduce new products and features, increasing the pace of innovation and providing more value to our customers more quickly. A great example of our continuous innovation is the recent introduction of autoscaling on Atlas, which we announced in September. This feature brings automated capacity management to Atlas. When enabled, Atlas will now track key resource utilization metrics in real time and autonomously adjust instance sizes up or down as needed by using predictive modeling and proven matching practices developed from matching tens of thousands of MongoDB deployments.
With this feature, Atlas now efficiently and elastically scales the customer's deployment to match their business needs without the need for manually intensive monitoring, analysis, and intervention. Early customer feedback provides great validation for this feature, and in the future, it will influence how we introduce this capability for private clouds. Second, with Atlas, we also get granular visibility to feature adoption and user behavior. This continuous and data-driven feedback loop allows us to understand what features customers find most valuable. Where in the product they may get stuck, or where they're not taking advantage of all the capabilities available. This enables us to iterate more often and more intelligently, ensuring that our product teams are focused on the most important opportunities. As a result, we can identify in real-time ways we can help customers optimize their MongoDB deployments with more timely and relevant assistance.
For example, we realized that a premier global financial institution was running an internal HR application without mission-critical scalability and security features, which negatively impacted the application performance and created significant risk. By proactively identifying these issues and sharing best practices, we developed a deep and trusted relationship with this large customer. As another example, through database usage metrics such as increased IO throughput or turning on backup, we can infer when an application's likely to go into production. These types of insights allow us to scalably engage with our customers at key moments when they're likely to increase the complexity and criticality of their MongoDB deployment.
Moreover, when we learn a great deal from how beta customers interact with our new cloud products, including Atlas Data Lake and Full Text Search, we expect the same benefits when introducing future products, such as the public beta of the revamped Realm product. Our confidence in the successful GA rollout of these new products will only increase after learnings from beta customers. We regularly incorporate these learnings into our Enterprise Advanced release, improving the value of our on-premises product as well. Consequently, we're looking to further invest in technology and data science to harness these types of signals to build even stronger relationships and find more opportunities to win more business. Ultimately, this positive feedback cycle gives us greater confidence in the returns on the R&D investments and the durability of our customer relationships. Finally, Atlas is broadening our reach into the global database market.
MongoDB is not only one of the most popular databases, but Atlas also enables customers anywhere in the world to easily use MongoDB through our free tier and self-service offerings, whether they are startups or development teams in large organizations. Since Atlas is available on the 3 largest cloud providers outside of mainland China, AWS, Azure, and GCP, it is the most widely available managed cloud database service in the world. By having relationships with all 3 cloud providers, we benefit from access to their customers and geographic reach, and we are seeing positive results with all 3. The 3rd quarter was the first full quarter of our expanded partnership with GCP, and we are encouraged by the number of deals closed and the level of pipeline activity. As the latest example of our successful partner strategy, we announced a new partnership with Alibaba Cloud.
Together, we are providing the first authorized MongoDB as a service in mainland China. Partnering with the largest tech company in China extends our reach into one of the largest markets in software and validates the value of our IP and the strength of the SSPL licensing model. Thanks to these types of partnerships, we continue to propel MongoDB as a leading data platform around the world. I'd like to spend a few minutes reviewing some customer wins and interesting use cases from the third quarter. We continue to see an increasing wide array of customers choosing MongoDB to run mission-critical applications. Axiata, one of the leading telecom groups in Asia, selected MongoDB Atlas and Enterprise Advanced to power Axiata's digital key FinTech application. MongoDB was chosen for its flexibility and how cost-effective it was compared to the company's legacy Oracle database.
MongoDB provides the high performance, availability, and scalability that the application needs to bring cashless convenience to the fingertips of both consumers and merchants throughout Malaysia. The Washington Post has chosen MongoDB to provide mission-critical support for its primary digital content management platform, Arc Publishing. MongoDB Enterprise Advanced serves as the backbone for its more than 750 million unique visitors a month and helps content teams power their businesses through innovation, better customer experiences, and storytelling on an integrated suite of tools. MongoDB gives them the flexibility to run both on premises and in the cloud. Foursquare, a location technology platform dedicated to powering business solutions and consumer products through a deep understanding of location, chose MongoDB Atlas to power all of its B2B enterprise location services.
These services, which include tools for developers, location analytics, advertising, targeting, measurement, and more, have been chosen by more than 50% of the Fortune 100. As Foursquare completes its digital transformation, MongoDB will continue to scale with them in the cloud. Keller Williams Realty, one of the world's largest real estate franchises, selected MongoDB Atlas as the core database to power its new game-changing real estate mobile app. The app gives Keller Williams real estate brokers a competitive advantage by using AI to reveal the most relevant information to potential buyers at the right time. Netskope is a next-generation cloud security platform that helps customers manage evolving security needs driven by digital transformation. They chose MongoDB to provide the security and scale they needed to expand internationally.
MongoDB gives Netskope the critical data ingestion and visualization capabilities to provide unrivaled visibility, real-time data, and threat protection when accessing cloud services, websites, and private applications. To summarize, MongoDB's third quarter performance was strong across the board. We are executing well on our strategic priorities and believe we are well-positioned to continue gaining share in the global database market. Before I transition to Michael, I want to add that we're thrilled to welcome Frank D'Souza, the co-founder and former CEO of Cognizant, to our board. Frank scaled Cognizant from inception to a $16 billion global professional services company. His experience and insight, in particular, massively scaling a technology company will be invaluable to us given our long-term aspirations. With that, let me now turn the call over to Michael.
Thanks, Dev. As mentioned, we delivered another strong performance in the third quarter, both financially and operationally. I'll begin with a detailed review of our third quarter results and then finish with our outlook for the fourth quarter and full fiscal year 2020. First, I'll start with our third quarter results. Total revenue in the quarter was $109.4 million, up 52% year-over-year. Subscription revenue was $103.8 million, up 56% year-over-year, and professional services revenue was $5.6 million, up 8% year-over-year. Our performance in the quarter benefited from broad-based strengths from both Atlas and Enterprise Advanced, as well as positive demand trends in all major regions. We continue to benefit from an attractive combination of new logo additions and continued strength in upsell activity, which we believe positions us well going forward.
The rapid adoption of Atlas continues to be the largest contributor to our growth. Atlas grew over 185% in the quarter and now represents 40% of total revenue, compared to 21% in the third quarter of fiscal 2019 and 37% last quarter. Atlas's continued strong growth at an increasing scale reflects a strong product market fit and a clear market shift amongst customers of all sizes towards a fully managed global cloud database offering. We are confident that these positive market trends, coupled with the operational initiatives Dev mentioned earlier, can drive continued strong growth in Atlas for the foreseeable future. During the third quarter, we grew our customer base by over 900 customers sequentially, bringing our total customer count to over 15,900, which is up from over 8,300 in the year ago period.
Of our total customer count, over 1,900 are direct sales customers, which compares to over 1,700 in the year ago period. The growth in our total customer count is being driven in large part by Atlas, which had over 14,200 customers at the end of the quarter, compared to over 6,200 in the year ago period. The sequential growth in total customers includes growth in our Enterprise Advanced customers, as well as new Atlas customers. It's important to keep in mind that the growth in our Atlas customer count reflects new customers to MongoDB, in addition to existing Enterprise Advanced customers adding incremental Atlas workloads. We also continue to see healthy expansion from existing customers, which is a key component of our growth strategy. Our net ARR expansion rate in the third quarter remained above 120% for the 20th consecutive quarter.
We ended the quarter with 688 customers with at least $100,000 in ARR and annualized MRR, which is up from 490 in the year ago period. Moving down the P&L, I'll be discussing our results on a non-GAAP basis unless otherwise noted. Gross profit in the third quarter was $79.3 million, representing a gross margin of 72%, compared to 72% last quarter and 77% in the year ago period. As a reminder, the year ago period benefited from the larger license mix from multi-year Enterprise Advanced transactions, and the current period margin is impacted by the mLab acquisition. Overall, we're pleased with our gross margin performance, which reflects greater efficiency and scale in our Atlas business. However, we continue to expect that we will see some modest reduction in overall gross margin as Atlas continues to become a bigger portion of our revenue.
Our operating loss was $14.3 million, or -13% operating margin for the third quarter, compared to a -11% margin in the year-ago period. Given the size and early stage of our market opportunity and our demonstrated product market fit, our primary focus will continue to be on investing in the business. Net loss in the third quarter was $14.6 million, or $0.26 per share based on 56.4 million weighted average shares outstanding. This compares to a loss of $0.13 per share on 52.7 million shares outstanding in the year-ago period. Turning to the balance sheet and cash flow, we ended the quarter with $426.4 million in cash equivalents, short-term investments, and restricted cash. Operating cash flow in the third quarter was negative $11.5 million.
After taking into consideration approximately $1.6 million in capital expenditures and principal repayments of finance lease liabilities, free cash flow was negative $13.1 million in the quarter. This compares to negative free cash flow of $9.7 million in the third quarter of fiscal 2019. As a reminder, the continuing growth of Atlas as a percentage of our overall business impacts our reported financial results in several ways. First, Atlas revenue is recorded on a consumption basis, whereas Enterprise Advanced includes a term license component that is recognized upfront. For a comparable dollar size contract, both Enterprise Advanced and Atlas will recognize the same amount over the contract term, but the Atlas contract will generate less initial revenue. As Atlas grows as a percentage of our revenue over time, this timing impact is greater.
Atlas has a lower overall gross margin than Enterprise Advanced because of its infrastructure component. That said, on an apples-to-apples functionality basis, Atlas is accretive to dollars of gross profit. Self-service Atlas customers and a growing portion of our direct sales Atlas customers pay us monthly in arrears versus annually upfront for Enterprise Advanced. We would expect the percentage of customers paying monthly in arrears to grow over time as we increase our focus on driving adoption of Atlas. We believe that facilitating the ease of consumption will ultimately maximize long-term revenues and cash flow. I'd now like to turn to our outlook for the fourth quarter of fiscal 2020. For the fourth quarter, we expect revenue to be in the range of $109 million-$111 million.
We expect non-GAAP loss from operations to be $16.5 million-$15.5 million, and non-GAAP net loss per share to be in the range of $0.29-$0.27, based on 56.8 million weighted average shares outstanding. As a reminder, we anniversaried the acquisition of mLab on November 1st. For the full fiscal year 2020, we now expect revenue to be in the range of $407 million-$409 million, compared to our previous guidance of $390 million-$395 million. For the full fiscal year 2020, we expect non-GAAP loss from operations to be $58 million-$57 million, and non-GAAP net loss per share to be in the range of $1.04-$1.02 per share, based on 55.9 million weighted average shares outstanding. To summarize, MongoDB delivered excellent third quarter results.
Our focus on executing on our product roadmap and expanding our go-to-market reach is driving high levels of growth at scale. The steps we've taken to establish MongoDB as the modern general data platform of choice positions us for continued success. With that, we'd like to open it up to questions. Operator?
We will now begin the question and answer session. To ask a question, you may press star then one on your touch tone phone. If you are using a speakerphone, please pick up your handset before pressing the keys. To withdraw your question, please press star then two. At this time, we will pause momentarily to assemble our roster. Our first question comes from Raimo Lenschow of Barclays. Please go ahead.
Hey, congratulations on another strong quarter. Dev, first question for you and then one for Mike. We obviously had the big AWS conference last week, and I'm just wondering, what was interesting there is just the lack of focus from the competing product they kind of launched. Can you just talk a little bit about how the last year has played out for you in terms of initially some noise, and now it looks like it's kind of somewhat going away. What are you seeing in the market in terms of customers understanding really the advantages of going with kind of Mongo rather than some kind of duplication of that? Then for Mike, since billing, if I listened to you correctly, then billing is going to have more and more noise in there. Have you thought about giving us a different metric?
How should we think about the future of the business if we want to kind of analyze that? Thank you.
Thanks, Raimo. With regards to DocumentDB, I think it was January of this year that Amazon officially launched DocumentDB, and I know there was some trepidation with investors about the potential impact on our business. As you've seen now for the last four quarters, our business has performed really well. I think there's been a number of reasons for that. One, this market is truly massive. It's one of the largest markets in enterprise software, I think there's room for a lot of people. Two, unlike a lot of other open source technologies, we have a strong moat around our business, especially our own IP.
It's very difficult, or actually, it's illegal for a cloud provider to take our native code and wrap it into their platform and offer it as a service. What Amazon did was basically tried to emulate MongoDB through a different architecture, which has severe performance and feature trade-offs. At last count, DocumentDB failed about 65% of the corrective compatibility tests against MongoDB. For those customers who have primitive requirements, DocumentDB could work. But for most customers who have mission-critical requirements and want the breadth of use cases that MongoDB supports, MongoDB is by far the best choice, and I think you see that reflected in our numbers. You're right, we didn't see any real news about DocumentDB. We just heard a lot of other announcements, and so we feel quite good about our position in the marketplace.
Raimo, on your kind of billings related question, yes, I think from the very beginning and outset of our life as a public company, we've highlighted how that's not a metric that we use internally or one that we focus on. Obviously, understand that it's a popular one that investors and other folks look at. We tend to focus on the revenue, specifically, in the business. I think the other piece that's particularly challenging about billings, which we've talked about, is if you're looking at sort of the revenue plus the change in deferred. Historically, the change in deferred has really been the place where people have a chance to outperform in recurring revenue subscription businesses. I think Atlas sort of changes that. Looking at deferreds in isolation in particular, we've heavily discouraged people from doing.
We also are going to be, as I mentioned in the prepared remarks, continuing to focus on driving ease of adoption and ease of use of adopting Atlas. That will continue to most likely involve a higher percentage of Atlas-related customers paying monthly for invoicing in arrears rather than signing big, large commitments up front and paying for those, which would generate incremental deferred revenue. I think while it's not a metric that we've ever focused on, I think you're correct, it's of decreasing value or insight relative to understanding the business. We have looked at other metrics to try to discern or determine if there are other metrics that are useful. Unfortunately, we haven't found anything that actually provides an incrementally helpful view of the world. We'll continue to look.
Lovely. Thank you. Congrats.
Thanks, Raimo.
Our next question comes from Sanjit Singh of Morgan Stanley. Please go ahead.
Thank you for taking the question and another great quarter from you guys. Dev, I wanted to talk a little bit about Atlas. From my perspective, this was one of the stronger Atlas quarters, and Atlas has been strong for quite some time, but it really felt like this quarter was even stronger than we expected. I was wondering if you could sort of break out the strength between the self-service part of the business on Atlas and what you're seeing on the enterprise expansion side, any details there?
Yes. Thanks, Sanjit. I would just say that I think we saw a broad performance and good performance across the board. I think we're feeling really good about sales sold Atlas. Our sales teams, as you know, have really found a lot of success with Atlas, with both mid-sized and large enterprises, and that business and that channel's doing very well with Atlas. We're also seeing increasing success in our self-serve business. We continue to increase the velocity of experimentation, to test in terms of search and paid search and organic search. We are expanding our business more and more overseas, and we're seeing a lot of similarities in terms of needs. We're seeing registration from corporate accounts increase nicely for us, so that these are sometimes development teams and large organizations who are testing Atlas, maybe as a precursor to doing a big deployment.
We're also getting better at using our data and analytics to get more insight into our business. Our self-serve business is also very iterative, but we feel really good about that channel and the sales channel. I wouldn't say there's any one channel that outperformed the other.
Understood. You mentioned in your script about the traction you're seeing across the three major public cloud providers. I was wondering if you could contextualize the opportunity with the Alibaba partnership in terms of what did sort of MongoDB traction look like historically in China. What do you see the opportunity, if you could sort of frame out the opportunity with Alibaba going forward in the Chinese market?
Yeah, we've seen, just through downloads, that China is a very, very big and popular market for MongoDB. I've been to China multiple times. We have a number of large enterprise customers there. When we run our user conferences, they tend to be packed. We know that there's a lot of interest and appetite for using MongoDB for a wide variety of applications use cases. The challenge we've had in China is that due to the regulations there, we could not offer Atlas directly in China because there's rules about owning data centers directly. Obviously, we did not, until recently, have any relationships with any local entities. Candidly, we've been in discussions with Alibaba for some time. There was clearly interest on both sides to do something.
One, Alibaba recognized the value of working with us, recognized the value of working with a team that's building the code and also getting another level of support. We wanted to frankly work with the largest tech company in China because they obviously have massive reach in that market. I would say we're very excited about the opportunity. It's still very early days. We just had some people in China two weeks ago, doing some events there, and the interest level and activity level is very, very high.
Perfect. Congrats again.
Thanks.
Our next question comes from Heather Bellini of Goldman Sachs. Please go ahead.
Hi, this is Dan Turchon for Heather Bellini. Thanks for taking the question. Earlier this year at MongoDB World, you guys announced a number of new products and features like Data Lake, Full Text Search. Can you give us an update on the traction you're seeing and what the early feedback's been? For something like Full Text Search, how are you thinking about the potential uplift in customer spend?
Yes. Well, thanks for the question. Just to remind everyone, we announced Atlas Data Lake and Full Text Search at MongoDB World this past summer. It's still in beta, we're engaging with hundreds of customers who are right now using the product. We expect it to go GA next year at our World Conference around the May timeframe. The feedback has been really, really positive. We're really excited about the feedback because it helps us sharpen the product, understand how customers are using the product, understand potentially details around user experience as well as the full feature and functionality and the use cases that people may be using, and we're very excited. Just to remind everyone, Atlas Data Lake really enables customers to use the same semantics with MongoDB to basically query offline data as they do with online data.
It's a very easy product to use, adds immediate value. It's a pay-as-you-go model, the upfront investment or there's very low entry level of cost. The ROI is very compelling, and we're really excited about that product. Full Text Search is basically a request that customers have been making of us for some time because rather than using disparate systems, they'd rather use one platform. We're excited about that opportunity as well, and the feedback there has also been incredibly positive. It's hard for me to tell you what the uplift will be, but we're feeling very bullish about both products.
Thanks. Helpful. Thank you. Last quarter, I believe you called out a tough comp coming up in 4Q just based on a handful of Atlas customers that were consuming in excess of their run rate. I was just wondering if you could put a little context or commentary as to the size and what you're seeing in terms of utilization today and then how the pipeline looks as we get into 4Q and out into fiscal 2021?
I think there are three things to keep in mind in terms of the difficult compare that we have in Q4. First is this is the first quarter where we'll have the anniversary of mLab at the base period, and to recall that the mLab cohorts are declining. That's sort of the first point to make sure that you keep in mind. Secondly, we did call out last year that we did have a handful of accounts who are over-consuming relative to their commitment levels. We don't have any reason to believe that's seasonal or anything else related to that. I think we should look at that as sort of not reproducible or replicatable. Then third, it was just a really great quarter, frankly, last year.
While we're eager to have more good quarters, we don't want to count on all of them looking like that.
Thank you.
Our next question comes from Brad Reback of Stifel. Please go ahead.
Great. Thanks very much. Some of our customer work has shown that some large customers out there are beginning to take existing Enterprise Advanced installs and move them over to Atlas. Maybe you can give us a sense of what type of opportunity that is for you from a financial standpoint. Thanks.
Yeah. Brad, thanks for the question. I would say that the scenario you outlined is still one that's, I would say, more of an exception than the rule. That tends to happen less frequently. That's really more driven by the fact that customers have decided that rather than managing their own infrastructure, they'd rather consume MongoDB as a service. We still believe in the value of offering an on-prem product because there are many customers who either have a lot of sunk costs and want to preserve those investments, or they are in certain regulated industries that prevent them from using cloud services, or frankly, are just taking a more conservative approach to the cloud and are on a slower path to moving to the cloud.
What MongoDB does is just enable them to use MongoDB wherever they are on that journey from on-prem to the cloud. While there are definitely some customers who may be migrating their Enterprise Advanced implementation on-premise to the cloud, that's driven typically by their own needs, the desire to maybe move to the cloud themselves, a desire to reduce the amount of investments in their own infrastructure, and maybe the ability to iterate and innovate more quickly. It's not driven by us.
Got it. Thanks very much.
Our next question comes from Brent Bracelin of Piper Jaffray. Please go ahead.
Good afternoon. One for Dev and a follow-up for Michael, if I could. Dev, I wanted to go back to the prepared remarks. We heard a lot of talk around your now cloud-first strategy, very similar to the commentary we heard at MongoDB World, where you're now adding features to the cloud version, the Atlas version of Mongo here first. My question is really around the customer response to that. As you look at your pipeline, are you seeing a greater mix of customers and a customer appetite to also go cloud first? Just trying to understand as you make this pivot, as you make a broader narrative around your own cloud-first strategy, is that really customer-driven? Or is there a variance in customer appetite around Enterprise Advanced and Atlas? Thanks.
Yes. Great question. I would say that we tend to be a very customer-centric or customer-obsessed organization. We tend to listen to our customers very carefully. It's no secret that many customers do desire to move to the cloud. The pay-as-you-go model, the low entry cost, the ability to elastically scale, are all compelling reasons to move to the cloud. What's also interesting is smaller customers and early-stage customers tend to prefer to go to the cloud as fast as possible because that allows them to avoid having to invest in their own infrastructure and really allows them to focus their limited resources on the work that really differentiates their business. What Atlas has done is really open the aperture of the customers we can go after and frankly, monetize.
As you may remember, with Atlas, we can monetize every interaction with every customer, depending if they're an early-stage startup in a garage or a large-scale enterprise in the Fortune 500. What Atlas does is basically enable them to start in whatever scale they want to start and then grow from there. What we find is customers may just have one app and then that app grows. More typically, we find as soon as they launch one app, they very quickly start launching other apps because they find the value of Atlas so compelling. Our growth and expansion of Atlas is very strong.
Just a quick follow-up here for Michael. As we think about Enterprise Advanced, we have that modeled about 15% growth this quarter. I know it was a tough compare. You have ASC 606. Next quarter, we have the implied growth closer to 5%. Should we think about this as being the new norm for kind of Enterprise Advanced with the customer appetite shifting to Atlas? Should we think about Enterprise Advanced as kind of a lower growth type business, just given the mix shift going on? Any color there would be helpful. Thanks.
I think the key thing to think about Enterprise Advanced is unfortunately, it relies on the 606 accounting. It's very much driven by the components of the term license revenue, which is driven off of when the subscription start date is. As we've talked about previously, can also be influenced by the mix of multi-year contracts or not, which certainly affects the Q3 base period, as we've previously indicated. I think it's much more an indicator of sort of the variability that we'll see quarter-over-quarter and year-over-year that make those kinds of comparisons, either sequentially or annually, very difficult to make just as a result of the 606 accounting.
Got it. More noise than anything else.
Yep.
Thank you.
Our next question comes from Tyler Radke of Citi. Please go ahead.
Hey, thanks, guys. Appreciate it. Dev, I was hoping you could talk about just how, looking back maybe on the last three, four years of Atlas, just how the customer use cases have evolved, maybe how the quality of customer or quality of revenue has evolved. Maybe for Michael, you talked about potentially a higher mix of customers paying in arrears. Just how are you thinking about that from a guidance perspective? Obviously if you're having less of that Atlas revenue come from deferred, I presume it makes it a little bit more difficult to guide. Thank you.
Sure. Thanks, Tyler. Just to be clear, we launched Atlas about a little over 3 years ago, the summer of 2016. The 1st year was really driven predominantly by self-service, where we had a lot of organic demand from developers around the world who really wanted to leverage MongoDB as a service. It took us about a good year to really enable our sales force to know how to sell a cloud service, because historically all they had been doing was selling an enterprise product. Then the sales force kicked in the 2nd year, and that's when we really started seeing penetration in large and small to medium accounts.
I would actually say that our traction was first in the small to medium market, where those cohort customers really saw the value of Atlas because they had very little interest in investing in their own operational infrastructure. Slowly we started to pick away use cases and workloads with the very large enterprise customers. Now we're kind of seeing a healthy mix across all three channels. That's essentially how that business has evolved. Just on the guidance, I would say the key thing is where we guide to really is on the income statement, so the revenue and the net loss position, and whether, given that Atlas is all consumption-based, as opposed to Enterprise Advanced, which has that term license component.
Given that Atlas is all consumption-based, whether it's signed forward an annual contract with an annual commitment that's paid upfront, that would generate that deferred revenue, or whether or not it's monthly invoicing and arrears. The revenue recognition is consumption-based, so it doesn't really impact or affect our guidance.
Thank you.
Our next question comes from Jack Andrews of Needham. Please go ahead.
Sure. Thanks for taking the question. Dev, I was wondering if you could give us an update on maybe the depth and breadth of strategic C-level conversations you're having these days. To the extent that you're having them, are they still more all about digital transformations, or are there some other drivers that are happening there? Just how do these conversations translate into your business? Should we be thinking about potentially larger deal sizes over time?
Yes. I've mentioned in the past that we have tended to go a lot higher in organizations than historically we've ever done as a company. We also spend a lot of time focused on the developer community, driving awareness, and making sure that they're fully updated on all our products and services. I would say over the last 18 months or so, one of the big themes with the senior-level stakeholders and key accounts is the increasing awareness of multi-cloud. Customers are obviously very sensitive about being locked into any one platform. They also value the ability to arbitrage between one cloud provider and another cloud provider. They also want the opportunity to maybe have certain access to different capabilities and strengths of each of the different cloud providers. That's been a big theme.
I would say the other big theme is obviously being able to drive developer productivity. Every organization's under tremendous pressure to innovate quickly, either disrupt or be disrupted. Using MongoDB to drive developer throughput and developer velocity has also become very important. The rise of mobile, the rise of new use cases like IoT, machine learning, deep learning, et cetera, are also top of mind of these customers. Obviously analytics. That's an area that they're always very interested in because they want to gain deeper insights in their business. Something like Atlas Data Lake becomes very interesting to them because they can leverage the existing investments they've made in their people and get the value of a deep analytics platform to basically run analytics on both your online and offline data.
I would say, obviously, every customer has own unique requirements and priorities, but those are some of the themes that we hear from senior stakeholders in these organizations.
I appreciate the color. Just as a quick follow-up, are you seeing any change in appetite for just the whole legacy application modernization opportunity? Is there any inflection in demand on those types of use cases?
That's going in full force. We see legacy modernization as one of the key drivers to our business. The move to the cloud is another that could be tied to some legacy modernization initiative, but it could be also just a desire to get more and more of the infrastructure off-premise and into the cloud. Those are huge initiatives for us. A lot of times, we're helping customers think through which applications to modernize first and where we can be most helpful. Those are some of the conversations we're having.
Thanks. Congratulations on the results.
Thank you.
Again, if you would like to ask a question, please press star, then one. Our next question comes from Rishi Jaluria of D.A. Davidson. Please go ahead.
Hey, guys. Thanks so much for taking my questions. Wanted to maybe follow up and thinking about China. Appreciate it. It's obviously a huge opportunity. Maybe can you help us understand, from the licensing perspective, what gives you the confidence in SSPL being able to protect you and not having a situation where the code is either replicated or emulated to a heavy extent? I've got a follow-up.
Well, thanks for your question. China is obviously a massive market. It's got a huge number of developers. There's a big focus in China to using software to drive a lot of innovation. We are an open source company, so our source code is available to all. As we said when we introduced our new SSPL licensing, in certain parts of the world, it's very hard to adjudicate those issues. Alibaba felt it was important for them to have a direct relationship with us, to get access to the latest features, to get access to our engineers and our support infrastructure, because they see such a big opportunity for MongoDB in that market. I want to be clear, the Alibaba relationship is not Atlas. It's a separate OEM relationship. It's based on some minimums that we expect every year.
Then we will true up on an annual basis. It will not be counted as Atlas revenue. It is a way to access the Chinese market in a way that we've never been able to do before, by having the largest tech company in China partner with you. We think that that's a very compelling reason to go into partnership with Alibaba.
All right. Thanks. I appreciate that. Just if we were to think about, you're clearly seeing strength in both Atlas as well as Enterprise Advanced, and then you've put up 20 quarters of 120% plus net expansion rates. I think just directionally, if we were to think about how would expansion rates look like with Enterprise Advanced customers versus with those using Atlas? Anything directionally I think would be helpful. Thanks.
Yeah. What we've talked about before is the number that we give is really done by channel, so that's the direct sales channel. It's not really done by product, is the way to think about it. We see strong expansion in enterprise customers, whether it's Enterprise Advanced and then staying with Enterprise Advanced, whether it's Enterprise Advanced and adding Atlas workloads, or whether the relation starts off with Atlas and expands either to more Atlas or then Enterprise Advanced for certain workloads, regardless of what the trajectory is, because there's not one common monolithic pattern or typical consistent pattern that people fall into.
What we have publicly said is if you separately look at the self-service cohorts and the self-service behavior of customers who are signing up for Atlas, the net AR expansion of those cohorts is even greater than the net AR expansion we see in the direct sales channel. I think that sort of speaks to the incredibly strong product market fit that we see in Atlas, both in the direct channel as well as self-serve. Hopefully that gives a little bit of incremental color that we've shared.
All right, great. That's helpful. Thank you.
Thank you.
Our next question comes from Pat Walravens of JMP Securities. Please go ahead.
Oh, great. Thank you, and congratulations. Dev, the venture community seems to be advising their portfolio companies more and more to use a product-led growth strategy for their go-to market as opposed to sales or marketing-led. Is that basically what you're talking about here in terms of how Atlas is changing the business? I have a follow-up.
I would say that, yeah, product-led growth is clearly a new trend in terms of how people drive top-line growth. I would say a company that has done that well was Atlassian. Clearly, that's something that we're doing here, especially on the self-serve side of the business, where we're looking for data signals in terms of how people are using the product. As I shared in my prepared remarks, there's certain signals that tell us if you're doing something really important, and we want to know that in advance or know that as soon as that happens, because we want to be there with the customer so that we can help them through those important moments of their application, whether it's going to production or they're expanding their capacity or expanding geographically, et cetera.
Those are the kind of signals that we use to help drive our business. Those signals tell us if a customer's expanding. Another example would be if for some reason we suddenly see a spike in utilization, we don't automatically assume that's a good thing. That could be that they've misconfigured their MongoDB implementation, and we could actually have a very grumpy customer. Identifying that customer quickly, making sure that they know what they're doing, and in many cases, they're very grateful that we quickly identify those issues, help them solve any misconfigurations, because that builds only a deeper and more trusted relationship with our customer. Those are examples of how we're using product signals to drive growth. As you can imagine, with a lot of data science, because we have lots of customers around the world, so you can't do this manually.
Obviously, we add a lot of human resources, especially for larger customers and even some mid-size customers.
That's super helpful. Thank you. Michael, for you, 52% growth to scale is obviously awesome. One question I get a lot is the business has decelerated a couple quarters. Pat, where does that stop and even out? I don't know if that's something that you would be willing to address or at least maybe how to think about it.
Yeah, I think when you look at the Q4 guidance, I'd say overall, we feel really good about the guidance. Clearly, when we gave our Q3 guidance and full year guidance on the September call, there was some implied or embedded Q4 guidance implied in that number. Clearly, we've raised that when you do the math. I think that shows that we're feeling quite good. I would iterate what we've talked about in terms of Q4 and just remembering the mLab impact, both in terms of it being the first time that we've hit the anniversary of that, as well as those cohorts decelerating. In general, I think we feel really good about where we are.
All right, great. Thank you both.
Thanks, Pat.
Our next question comes from Vinod Srinivasaraghavan of Oppenheimer. Please go ahead.
Hi. Thanks for taking my question and congrats on the strong quarter. I have a question on Atlas first. You had mentioned that expansion rates were strong with both self-serve and direct customers, are you seeing new self-serve and direct sales lands come on board who are spending much more now within their first year compared to when you IPO'd?
No question. As Atlas becomes more popular, as we add new features and capabilities, as we add more coverage around the world, as customers themselves get more comfortable, the word of mouth and virality of the usage, we're clearly seeing deal sizes go up in Atlas.
I think you can see that not just in the sales sold side of things, which of sort of the two channels within Atlas is actually the fastest-growing channel. Also you can see increasing spend per self-service customer, which at this point is most vast majority of which is Atlas related. We report that level of detail in the 10-Qs. You can see that there.
Okay, great. Thanks. One more. Can you just give us an update on some of the sales capacity investments you've been making? How is your hiring new sales reps progressing?
It's progressing well. This is something that we focus a lot on because a big vector for our growth is the amount of feet on the street that we have. We spend a lot of time recruiting the right people and then spending a lot of time investing in them and developing those people to ultimately get them to produce results and produce results consistently over time. That's a big part of the attention of the management's bandwidth is just about how we scale infrastructure. As you can imagine, it's not just hiring salespeople, but sales leadership, technical pre-sales to complement our salespeople in the field, partner resources in different parts of the world. That's all part and parcel of our formula in terms of how we scale the business.
Great. Thank you.
This concludes our question and answer session. I would like to turn the conference back over to Dev Ittycheria for any closing remarks.
Thank you. Thanks to everyone for joining us today. Just to reiterate, we feel really good about our business. We think we had a terrific quarter, and we look forward to speaking to some of you in the next couple of days. I know that Michael and Serge will be at a couple investor conferences. I suspect some of you will be there as well. Talk to you soon. Take care. Bye-bye.
The conference is now concluded. Thank you for attending today's presentation. You may now disconnect.