MongoDB (MDB) Q3 FY2023 Earnings Review

In Q3, MongoDB reversed a number of the trends that hampered their Q2 earnings report. They delivered a strong beat on revenue growth, a return to positive operating margin and a nice bump in customer activity. The market reacted as positively to the Q3 report as it did negatively for Q2, pushing the stock up 23% the following day. Since then, MDB stock has appreciated further and now hovers around $200 a share. This is still below the $241 close the day after the Q2 report, but is well above the 52 week low of $135 touched before Q3 earnings.

While the trends turned back in a positive direction in Q3, the broader macro environment may still pressure growth for the next couple of quarters. Atlas sequential and annual revenue growth were the lowest in the last two years. Much of the revenue outperformance for Q3 was attributed to strong growth in EA license revenue. Nonetheless, these results were much better than expected coming out of Q2. Adding to the upside momentum, MongoDB resumed its path to profitability, with positive operating income and improved gross margin.

For the long term, I still think MongoDB is well-positioned to participate meaningfully in the secular trends of data growth and platform consolidation. Going into the second half of 2023, year over year comparisons become easier and the current investment in S&M will drive further growth. The inflection to sustained positive operating margin should provide support for the valuation multiple. MongoDB’s competitive position is defensible, as industry analysts rank it above comparable alternate solutions in features and capabilities. While the hyperscalers offer competitive products, the depth of MongoDB’s collaboration and partnership agreements with each is actually increasing, providing an additional tailwind to growth.

Audio Version
View all Podcast Episodes and Subscribe

Let’s take a look at the Q3 results in detail and then loop back on the broader product strategy and competitive position. For background research on MongoDB’s performance, interested readers can see my full coverage and revisit prior posts from our partners at Cestrian Capital Research.

Growth Metrics

Q3 revenue was $333.6M, up 47.0% annually and 9.8% sequentially. This soundly beat the analyst projection for $304.7M, which would have represented just 34.3% annual growth and no sequential improvement from Q2. The analyst projection was even slightly higher than the company’s own estimate for $300M – $303M in revenue coming out of Q2. The sequential revenue growth rate in Q3 ticked up nicely from Q2’s 6.4%.

Atlas grew 61% y/y in the quarter and now contributes 63% of total revenue. The growth rate declined from Q2’s 73% annual and 13.5% sequential growth. Because of outperformance on the EA side, the share of total revenue contributed by Atlas shrunk by 100 bps from 64% in Q2. A year ago in Q3 FY2022, Atlas made up 58% of revenue and was growing at an annual rate of 84%.

For Q4, leadership projected a revenue range of $334M – $337M, representing 25.9% annual growth at the midpoint and 0.6% sequential growth. If MongoDB were to beat their estimate by the same amount as the Q3 beat (about 13% of annualized growth), then Q4’s actual revenue growth would be closer to 40%. Analysts had modeled for just $318.4M in revenue for Q4, which would have represented about 18% annual growth. The significant beat in Q3 helped propel the Q4 estimate above the analyst projection , even with a negligible sequential raise.

For the full year, MongoDB is now targeting a range of $1.257B to $1.260B in revenue, for annual growth of 44.0%. This represented a substantial raise of $54M – $61M from their prior range issued in Q2 of  $1.196B to $1.206B, which was 37.4% annual growth and a raise of just $19M. The large raise in Q3’s report exceeded the roughly $50M of reported beat and raise amounts, implying further outperformance for the Q4 report.

In reaction to their weak performance in Q2, MongoDB leadership provided some commentary on the demand environment, specifically calling out those customer segments and geographies that showed weakness. Further, coming into Q2, management had already factored in $30M – $35M of annualized revenue loss to account for the macro impact. This turned out to be a low estimate when the Q2 results were reported, implying the impact to the full year would be worse than expected (versus conservative, which was the initial assumption).

During the Q2 earnings call, management discussed the different factors at play, broken out by customer types:

Self-serve. These are generally small businesses and individuals who sign-up and consume MongoDB resources directly online and don’t interact with a salesperson. Consumption softness first observed in Q1 continued through Q2, but performed better than expected.

Mid-market. This category includes a lot of fast-growth, digital native companies that selected MongoDB as their primary database from the beginning. The slowdown in this category observed in Europe in Q1 extended globally in Q2. Additionally, it was more pronounced than expected.

Enterprise. Enterprise customers represent MongoDB’s largest source of revenue. In Q1, MongoDB didn’t experience an impact on consumption, but they modeled for some in Q2. As it turns out, consumption growth in North America for enterprise customers was above management’s expectations. But, Europe experienced macroeconomic headwinds greater than anticipated. This resulted in less Atlas consumption. Customers in this category would be Boots, Conrad Electronic and Otto.

Further, in Q2, MongoDB leadership had projected weakness in EA product performance for the remainder of the year. Patterns in EA purchases can have an outsized effect on revenue recognition. When a customer purchases an EA license, a greater share of the revenue is recognized upfront. This is different from Atlas where revenue generation for a new customer starts at $0.

We expect a sequential decline in Enterprise Advance in Q3 as the renewal base is sequentially lower. Looking into Q4, we expect a seasonal uptick in revenue from EA. But recall, we faced a very difficult year-over-year comparison given strong EA new business activity we saw last year.

MONGODB Q2 EARNINGS CALL

As it turns out, Q3’s actual performance was better than expected. Management pointed out improvement in the trends from Q2 that contributed to the outperformance in Q3. Preliminary explanations were shared during the earnings call. The Finance team was able to further expound on these in subsequent analyst events, particularly the Barclays Technology Conference just a couple days after earnings.

• EA revenue actually grew sequentially. This was unexpected, with the Q2 narrative indicating linear EA growth at best. Leadership highlighted the fact that some enterprise customers increased their EA license commitments as a precursor to their cloud migration. Additionally, some customers signed multi-year EA deals, which would normally be for a single year. Due to accounting rules, a portion of that additional deal value is realized as revenue upfront. These larger commitments by customers are counterintuitive in this macro environment unless they consider MongoDB mission critical and want to lock in volume pricing over a longer term.
• Atlas consumption trends improved in Q3, following a drop in Q2. They are not back to historical levels, however. Management attributed some of the growth to “seasonality”, implying that usage of applications was higher in the September – October period than over the summer. Additionally, Q2’s sequential Atlas growth was driven by three more days in the Q2 calendar quarter than Q1. This inflates the Q2 sequential growth and tempered it for Q3.
• The mid-market bounced back in Q3, versus the dip that occurred in Q2. This applied globally. Some of the digital native customers (like Coinbase and food delivery) were put into this category in Q2. This segment slowed down usage more than other categories in Q2, but then bounced back more than others in Q3. The segment only makes up a “mid-teens” percentage of revenue, but the digital natives are a large portion of that.
• Enterprise customers in Europe resumed their spending trends. After a dip in Q2, they increased usage again in Q3.

Looking forward to next quarter, management expects the seasonal benefit for Atlas to decrease. While November usage looks good, they expect some drop-off in December and January, due to the Holidays. However, the incremental Atlas ARR booked in Q3 will carry over to Q4, which drove much of the raise to Q4 revenue. EA revenue, on the other hand, drove outperformance in Q3. For Q4, it is expected to be sequentially flat, due to the large outperformance in Q3.

But also, just going back to your question in the guide, most of our raise for Q4 is actually Atlas driven. Because that better than expected starting ARR point actually helps our Q4 numbers. And that’s part of the reason — although the beat was mostly EA, the raise is mostly Atlas.

Barclays Technology Conference, December 2022

Finally, management pointed out that they did not experience “meaningful delays or deal slippage” in Q3. This contrasts with comments from other software infrastructure peers during their most recent earnings reports. Management highlighted the large net increase in new customers in the Direct Sales channel (more below), as the strongest indicator of the new business trends for the medium and long term.

---

Cestrian Capital Research provides extensive investor education content, including a free stocks board focused on helping people become better investors, webinars covering market direction and deep dives on individual stocks in order to teach financial and technical analysis.

The Cestrian Tech Select newsletter delivers professional investment research on the technology sector, presented in an easy-to-use, down-to-earth style. Sign-up for the basic newsletter is free, with an option to subscribe for deeper coverage.

Software Stack Investing members can subscribe to the premium version of the newsletter with a 33% discount.

Cestrian Capital Research’s services are a great complement to Software Stack Investing, as they offer investor education and financial analysis that go beyond the scope of this blog. The Tech Select newsletter covers a broad range of technology companies with a deep focus on financial and chart analysis.

---

Profitability Measures

A big part of the market’s negative reaction to MongoDB’s Q2 results was their reversal to negative operating margin and reduction of the full year estimate for the remainder of the year. This provided a double whammy for the stock, as a marked slowdown in revenue growth projections was magnified by loss of operating leverage. It was easy for analysts to conclude that MongoDB’s growth was slowing and they needed to spend significantly more to offset that.

In the Q3 results, profitability measures improved substantially, putting MongoDB back on the expected path towards positive operating margins and free cash flow. In the current investing environment, this factor is being weighed more heavily in valuation multiples. Companies with significant negative margins are being penalized more than than a year ago, in spite of high revenue growth.

Non-GAAP gross margin continued its upward movement, reaching 74.3% in Q3. This is up 110 bps from 73.2% a year ago. In Q2, gross margin was 73.5%. Gross profit grew nicely as well year/year, up 49.2%. This was slightly higher than the 47.0% annual growth in revenue. Management once again attributed the gross margin improvement to increased efficiencies in the Atlas cloud offering. Greater scale provides opportunities for volume discounts and optimizations when hosting solutions on infrastructure provided by the hyperscalers.

The outperformance in revenue and gross margin drove strong operating income. Non-GAAP income from operations was $19.8M (5.9% op margin) in Q3, up from $6.3M a year ago and a $12.4M loss in Q2. The Q3 result even beat the $17.5M in operating income delivered in Q1. The Q3 actual operating income significantly surpassed the expectation they had set in the Q2 report for $(10M) – $(8M). The outperformance was attributed to the revenue beat and cost management due to the challenging macro environment.

This strength in operating income drove a nice beat on EPS. Analysts had modeled for Non-GAAP EPS of $(0.17), which was within the company’s range of $(0.19) to $(0.16), set with the Q2 results. MongoDB actually delivered positive Non-GAAP EPS of $0.23. This result combined with the revenue beat to compound the market’s reaction to the Q3 earnings report.

Free cash flow showed some improvement, but not as marked as operating income. For Q3, MongoDB used $5.7M in cash. They spent an additional $2.7M in CapEx and finance lease liabilities, yielding total free cash flow of -$8.4M. This represents a FCF margin of -2.5%. A year ago, FCF was -$9.2M for a slightly worse FCF margin of -4.1%. In Q2, FCF took a large step down at $(48.6M) for a FCF margin of -16.0% MongoDB ended the quarter with a strong cash position of $1.8B.

Looking forward to Q4, leadership set a Non-GAAP operating income target of $6M – $8M. At the midpoint of Q4 revenue guidance, this would represent 2.1% operating margin. Given the beat from Q3 of almost $30M, actual Non-GAAP operating margin could reach 10%. For the full year, they increased the Non-GAAP operating income target substantially to a range of $30.8M – $32.8M. This is an increase of $42M from the range issued in Q2 of $(13.0M) – $(8.0M) and implies a Q4 beat of at least $12M.

Looking at spending allocations on a Non-GAAP basis by department between this Q3 and a year ago provides some color:

• S&M: Increased from $90.4M to $138.4M, up 53.1% y/y. S&M now makes up 41.4% of revenue, versus 39.8% last year. The Q3 allocation to S&M is down from 47.3% in Q2, which surged due to the MongoDB World event.
• R&D: Increased from $47.5M to $62.6M, up 31.8% y/y. R&D now makes up 18.8% of revenue, versus 20.9% a year ago. This is also down from 21.3% last quarter.
• G&A: Increased from $21.9M to $27.0M, up 23.2% y/y. G&A now makes up 8.1% of revenue, versus 9.7% a year ago. This is down from 9.0% last quarter.

The decrease of all three of these on a percentage of revenue basis relative to Q2 drove the outperformance on operating income, in addition to the revenue outperformance. Compared to last year, the allocation to S&M was the only department that increased. This reflects the greater investment in the go-to-market function to drive more Direct Sales relationships with customers. If the result is durable revenue growth through large customer usage expansion, I am okay with the increased spend.

The 10-Q provided some additional insight into hiring trends. I created the following chart of total employee counts and included those attributed to Sales and Marketing from the quarterly data. 

Looking at Sales and Marketing headcount in Q2, we see a substantial increase, up 56% year/year and 16% sequentially. MongoDB doubled the number of new hires in Sales and Marketing in Q2, as compared to the average of the prior four quarters. Besides MongoDB World, this explained the large increase in operating expenses in Q2, which is also the source of the worsening operating margins through the rest of the year.

In Q3, we saw hiring moderate back to historical rates. The increase in Sales and Marketing also slowed. If we average the Q2 and Q3 additions, then the hiring rate would be inline with prior quarters. This decrease likely helped improve operating income, in addition to the revenue outperformance in Q3. I still like the increased investment in S&M, as I think that will drive more expansion within existing customers.

Customer Activity

Customer activity provided another highlight for Q3. Total customer growth ticked back up sequentially over Q2, adding 2,100 customers to reach 39,100. This represented a marked improvement over Q2’s dip in additions to 1,800. MongoDB had been consistently increasing total customers by 2,000 to 2,200 over the prior 4 quarters. In this macro environment, it is a positive sign for MongoDB to return to the customer addition range it achieved during 2021.

As MongoDB is approaching 40,000 paying customers, the rate of customer additions is still relatively strong at 26% annually and about 6% quarterly. Combined with their net expansion rate, this should support healthy revenue growth going forward. Additionally, with less than 4% of customers spending more than $100k a year, the bigger opportunity for MongoDB is in spend expansion for existing customers.

The vast majority of new customers came on through the Atlas tier. These represent both brand new customers and EA license customers transitioning to cloud. MongoDB doesn’t break out new EA customers specifically, but we can assume this is a small number, with an outsized contribution to revenue.

As the opportunity for revenue growth going forward can be driven more by spend expansion of existing customers, MongoDB provides another designation for certain customers called Direct Sales. These represent customers that the MongoDB sales team considers high potential for future spending growth. As such, they are assigned a salesperson. While we lack an objective measure for the impact of these customers on usage growth, MongoDB reports on the count of this category and considers it an important financial signal.

In Q3, MongoDB added 500 Direct Sales customers, bringing the total to 5,900. This category of customers is growing faster than total customers, clocking in at 51% annually and 9% sequentially. MongoDB has been adding about 300 – 600 of these customers a quarter for the last year, with the rate picking up over the last two quarters. As MongoDB increases their investment in sales and marketing, this is one of the primary outcomes. The reasoning is that these direct sales relationships will drive more spend over time.

Direct Sales customers generated 87% of total subscription revenue in Q3. This is up from 86% in Q2 and 84% to 87% of revenue for the four quarters before that. A year ago this value was 85% and two years ago it was 82%, demonstrating that the contribution of Direct Sales customers is expanding.

This expansion of customer spend should drive an increase in customers spending over $100k annually. MongoDB measures $100k customers by annualizing the prior quarter’s spend. In Q3, MongoDB added 83 of this size customer, which was the same as Q2 and at the higher end of the historical range on an absolute basis. MongoDB leadership also provides a range for the net ARR expansion rate each quarter. They simply confirm that the value was over 120%, which has been the case for many quarters. This implies that customers over a year old increased their spend by 20% or more when comparing the prior period with the same a year ago.

Q4 of last year represented a spike in $100k customers, crossing 100 additions for the first time. With MongoDB’s heavy investment in sales and marketing and focus on driving Direct Sales customers, I would expect a similar increase going into Q4 this year. Given the macro headwind versus tailwind a year ago, this Q4’s result might not be as significant, but growth in $100k customers is the logical outcome for all the investment in fostering Direct Sales relationships. This will be an important signal for next quarter’s report.

In the prepared remarks, MongoDB’s CEO highlighted a handful of customer wins. This quarter, he grouped several together and provided details on a couple.

• Toyota Financial Services, Ulta Beauty, Mediastream and Vodafone were singled out as running mission-critical applications on MongoDB Atlas. Their use cases spanned search, analytics and mobile data services, in addition to the core function as a database of record. This provides evidence of MongoDB’s product strategy of expanding into adjacent workloads.
• For Vodafone specifically, the CEO described how they have made MongoDB Atlas a core data platform for hundreds of new cloud-native applications, serving their 625M customers across 65 countries. Additionally, Atlas is is the data source for their IoT ecosystem with 140M devices.
• He then called out several more customers that are using MongoDB Atlas for their migration from on-premise to cloud-based delivery. These include American Tire Distributors, Schwarz IT and Volvo Group.
• Finally, he named a few customers that are expanding their use of MongoDB Atlas across their tech stack and applying it to back new applications across different parts of their business. These include Hugging Face, Okta, Washington Post, Cisco and L&T SuFin.

In several of these examples, the customer is using MongoDB as a generalized data platform upon which any development team in the organization can build an application. This “platform” configuration provides a powerful expansion motion for usage, as each new application will have a separate MongoDB cluster provisioned.

Interfacing with these larger customers to help evangelize the suitability of MongoDB for additional workloads is the function of the Direct Sales team. With the assistance of sales engineers, they can engage with the customer’s engineering leadership to point out additional MongoDB use cases, like search, times series (for IoT), real-time analytics and mobile. They can also work with the team to migrate existing relational schemas to a document-oriented equivalent.

Another motion being captured by MongoDB is the consolidation of database vendors. A number of customer examples over the past couple of quarters have highlighted cost savings realized by moving to MongoDB Atlas, either by reducing license fees from multiple alternate solutions or decreasing staffing support required to maintain open source projects. This represents a new marketing message, reflecting the benefits of consolidating multiple application workloads onto the MongoDB platform. This aligns with messaging from other software infrastructure providers in recent quarters and plays well with overall enterprise sentiment given the macro environment.

Product Strategy

While we are experiencing volatility in the IT spend environment and macro is pressuring usage of customer applications, MongoDB’s product strategy remains largely the same. MongoDB started by building a reliable, scalable and easy to use platform to serve as the transactional data store for modern applications. This hinged on the flexible document-oriented data model, which aligned with how developers work with data within applications. This represents an evolution of the traditional relational data model. While compact, a relational model rarely maps directly to how data is stored in an object-oriented application, relying on an ORM to serve as a translation layer between the two. The document model eliminates this friction by mirroring the structure of data within applications, making it easier for developers to use in most cases.

From this foundation, MongoDB’s strategy has been to expand into adjacent data storage use cases. This expansion has taken a couple of forms. First, they built APIs to facilitate the translation of different data models into the core document store. These data types can be easily modeled in a document form and then queried through the API using the same access patterns typical for that data type. These extensions address most “document adjacent” data types like key-value, time series, graph, geospatial, search and even basic relational.

The second form of expansion applies these additional data types to application specific uses. The first was to provide a mobile app data sync, which enables mobile devices to maintain a lightweight copy of data for a user in close proximity to the device. This replaces other cloud-based mobile app data stores like Google’s Firebase.

The next use case expansion was search, which is functionally based on an index of documents, providing a natural extension to MongoDB’s core data engine as a document store. The MongoDB team leveraged the same search libraries based on Apache Lucene that are used by other dedicated search solutions. This allowed MongoDB to be used for text search and more recently faceted searches, providing a replacement for Elasticsearch and SOLR.

The benefit to using MongoDB for search is the proximity of the transactional data to the search index. A normal configuration with Elasticsearch or SOLR requires a data sync job to transmit updates from the transactional database to the search index. This process can be brittle, introduces a delay and encumbers DevOps with another system to manage. MongoDB’s search architecture eliminates these disadvantages.

More recently, the MongoDB team has been continuing this trajectory into new use cases by adding support for analytical workloads through an interface to their time series data type. This allows applications that collect, process and distribute large amounts of time series data (primarily IoT) to leverage MongoDB as their back end.

Through these additional data types and workloads, the sales team can make the argument to potential customers that MongoDB can be leveraged for many types of applications. This allows the customer to eliminate data storage point solutions that address just one of these workloads. The benefit to the customer is fewer vendor relationships to manage, a simpler data interface for developers and less cost through volume discounts.

MongoDB’s CEO summed up the approach by providing an anecdote from a customer meeting at AWS re:Invent.

“So what we’ve done is a first-class transactional platform, and now we’re expanding the platform to do things like search and analytics,” he noted. “I was just meeting with a customer who was thinking about Mongo for their transactional platform, elastic for the search platform, and a graph database for a special use case. And, and we said, ‘You can do all that on MongoDB.’”

MONGODB CEO, Interview with thecube, December 2022

Beyond this product strategy, MongoDB has been pursuing newer go-to-market approaches. These involve two pursuits – stronger relationships and collaboration with the hyperscalers and expanded partnerships with System Integrators. For the hyperscalers, MongoDB has been rolling out deeper integrations and co-selling relationships. They have been very busy. Over the last year, they have announced strategic partnerships with GCP, AWS and Microsoft Azure. These relationships have strengthened in spite of each hyperscaler offering competing products, like DocumentDB from AWS and Cosmos DB from Microsoft.

In March 2022, MongoDB announced an expanded collaboration with AWS. The agreement with AWS built on the existing multi-year relationship between the two companies, aimed at driving customer adoption of MongoDB Atlas on AWS. In an effort to further improve the customer experience, both companies agreed to collaborate across sales, customer support, solution architecture, marketing and other areas to make MongoDB Atlas a better experience for developers on AWS globally. This includes increased workload migration incentives and enhanced tools to help customers move from legacy technologies in on-premise data centers to MongoDB Atlas on AWS. Finally, the partnership will support MongoDB’s expansion into more AWS Regions across the globe and the US Public Sector with FedRAMP authorization.

Not to be left out, Google Cloud Platform (GCP) struck a similar agreement with MongoDB in April. In this case, it is a pay-as-you-go offering available directly in the GCP console. Customers will just be billed for MongoDB Atlas based on their consumption, with no up-front commitments. This makes provisioning seamless, as customers can initiate the relationship through GCP and consolidate costs onto their existing GCP bill. Atlas is deeply integrated with a number of other GCP services including BigQuery, Tensorflow, Cloud Run, App Engine, EventArc, Cloud Functions, Google Kubernetes Engine (GKE) and Dataflow.

While Microsoft Azure has the most directly competitive offering in Cosmos DB, they too are starting to collaborate more. MongoDB’s CEO highlighted the improving relationship with the Microsoft Azure sales team on the Q2 earnings call.

At the end of the day, the hyperscalers still make money from customers that utilize MongoDB Atlas on their cloud. The primary source is the additional compute and storage that MongoDB is using to process and store their customers’ data. In addition, MongoDB can bring their on-premise customers with EA licenses to the hyperscaler partner when those customers are ready for a cloud migration. The hyperscalers are realizing that pushing their internal solution on a customer who has already decided to leverage MongoDB could result in them losing the deal to another hyperscaler.

The second new go-to-market motion for MongoDB has been to build out their relationships with system integrators and ISVs. These companies generate revenue by helping enterprises plan and execute their digital transformation projects. By partnering with these companies, MongoDB is more likely to be the recommended solution for the data storage portion of a digital transformation project.

A number of large systems integrators are in the process of setting up business units focused on MongoDB given the size of the growing MongoDB practice. A growing number of ISVs continue to build their products in MongoDB. We currently have close to 200 ISVs co-selling relationships, which is up more than 2x compared to two years ago. Our growing popularity has tangible benefits for our business, especially in periods of economic uncertainty.

MongoDb Q3 FY2023 Earnings Call

The reason that the hyperscalers and ISVs are forming partnerships with MongoDB (and not smaller database competitors) is because of the growing popularity of MongoDB with developers. On the Q3 earnings call, MongoDB management shared some stats around usage of the open-source product. They claim that the community edition has been downloaded more than 150M times from the MongoDB site over the last 12 months. That is more than in the period from the company’s founding through the beginning of 2020. Granted, downloads can be triggered by a lot of activities, including rebuilds within an existing customer’s installation, but the relative velocity increase is impressive.

As another data point, in Q3, the MongoDB Atlas free tier had over 300k sign-ups. Developers use this for small projects or training. The free tier has limits on storage (512MB) and shares RAM and CPU with other users. The main benefit to MongoDB is that these developers become familiar with the database and then bring it to their full-time job. Use of the free tier has increased 15x over the last 5 years.

Competitive Landscape

I have provided in-depth analysis of the competitive landscape for transactional databases and MongoDB’s position in prior blog posts. MongoDB is the most popular document-oriented database solution on the market. Additionally, with their expansion to other document-adjacent data types (time series, search, key-value), MongoDB intends to deliver a broader multi-model data platform. Plus, the solution provides support for unique workloads like mobile, in-app analytics, charting and search.

As the data platform’s applicability expands, it becomes suitable as the backing data store for many applications within the same enterprise. Most modern enterprises are allowing developers the freedom to choose their technology stack from a preset selection of options. MongoDB’s go-to-market with large enterprises involves getting anointed as one of these sanctioned solutions. Then, the broad popularity, ease of use and familiarity of MongoDB with developers drives their selection for new applications and upgrades.

As developers were empowered by enterprise engineering teams to make tool selection, the result has been a sprawl of point solutions to address multiple data types and workloads. The MongoDB value proposition is to provide a single platform to replace most database types. A reasonable argument can be made that MongoDB is a suitable drop-in for key-value, search, time series, graph, wide column and reverse index data types, as well as a denormalized relational schema.

The benefits of a single data platform stem from cost savings, fewer vendor agreements, higher developer productivity and reduced DevOps overhead. Additionally, MongoDB clusters can be located on and share data across all three hyperscalers (GCP, AWS and Azure).

MongoDB’s strategy with enterprises is to land with a single application and then expand into many. This has been successful with a number of large enterprises, with examples like Vodafone highlighted on earnings calls. And while we would assume that the majority of their customers would be the digital natives, the opposite is true. MongoDB’s largest customers are mainstream enterprises in finance, manufacturing, healthcare, retail and technology. They even have a number of government agencies as customers.

To assess MongoDB’s position in the market, we can refer to third party industry analysts. A reasonably objective indicator of data storage engine usage across all categories is provided by DB-Engines. They maintain a ranking of popularity of solutions on their web site, with an overall score and an indication of change in magnitude compared to the prior month and the prior year. This is constructed from a combination of inputs pulled from various public forums, discussion boards, web sites and job postings, which are all heavily developer influenced.

If we look at rankings for document databases and their multi-model adjacents, MongoDB is well ahead of any competitive offerings. The second place position goes to DynamoDB and that has a score less than 1/5 of MongoDB’s ranking. Further, no other offering is making significant progress up the rankings, with all solutions maintaining about the same relative level of popularity over time.

A graph of rankings also shows the large gap that MongoDB is maintaining relative to other offerings. The only contender that has been perceptively improving their position over the last couple of years is DynamoDB. The DocumentDB product from AWS is down at position 25 with a score of 2. This product clearly didn’t gain popularity against MongoDB, even though AWS had originally unveiled it several years ago as a direct competitor.

Gartner Magic Quadrant

Recently, Gartner published their latest Magic Quadrant for Cloud Database Management Systems. They define the category very broadly – this is probably the one of the more expansive technology categories I have seen Gartner create. It spans many types of data processing workloads, including OLTP transactions, stream event processing, traditional data warehouse, logical data warehouse, data lake and stream analytics.

Gartner defines the Cloud DBMS market as follows: Core capabilities are that vendors fully supply provider-managed public or private cloud software systems that manage data on cloud storage. Data is stored in a cloud storage tier. Optionally, they may cater to multiple data models and data types — relational, non-relational (document, key value, wide column, graph), geospatial, time series and others.

Gartner Magic Quadrant for Cloud DBMS, December 2022

Given all the types of data workloads covered, it’s not surprising that 20 vendors were evaluated. Of these, 12 were placed in the Leaders Quadrant. Both of those numbers are higher than on most other Magic Quadrant reports. Vendors are normally spread across the four quadrants more evenly.

After not even being included in the 2021 report, MongoDB landed in the Leaders quadrant in 2022. Gartner further clarified that “MongoDB did not respond to requests for supplemental information for this document. Gartner’s analysis is therefore based on other credible sources.” Effectively, this means that Gartner went out of their way to include MongoDB in the 2022 report, even without their input. In fact, MongoDB didn’t even reference the Gartner report on their web site, while they did include a link to the Forrester Wave report (below).

This inclusion by Gartner is unusual, and reflects favorably on MongoDB’s position in the market. Normally, companies are jockeying with Gartner to be included in these reports, as placement in a particular quadrant can be an important input for enterprise buying decisions. Additionally, newcomers are usually relegated to a lower quadrant on their first inclusion. For MongoDB to land in the Leaders quadrant on their debut, and without the benefit of their own input, is very impressive.

Second, MongoDB occupies the highest position among those providers who are purely a transactional database for application workloads. The hyperscalers would be expected to rank above MongoDB, as they offer multiple database solutions across both transactional and analytical workloads. Similarly, other providers ahead of MongoDB like Snowflake, Databricks, Cloudera and Teradata center their platforms around analytics and ML/AI use cases.

Looking at the quadrants below the Leaders, we see the more recognized competitors to MongoDB, including Redis, Couchbase, Neo4j and Cockroach Labs. These solutions are appropriate for transactional workloads backing applications. Gartner has place these in the Challenger or Niche quadrant. They also included Aerospike, DataStax (Cassandra) and InfluxData (InfluxDB) as Honorable Mentions.

What investors need to keep in mind when looking at Gartner MQ’s is not the absolute position of a provider on the quadrant, but their movement from one year to the next. By this measure, MongoDB performed the best, going from 0 to the Leaders quadrant in one year. Graph database Neo4j made their debut as a Niche Player. MariaDB and SingleStore were dropped. The only other provider to be promoted to the Leaders quadrant was Cloudera, and they are primarily used for analytics and machine learning.

Specific to MongoDB, Gartner highlighted several strengths and cautions in their standard evaluation for each vendor. As strengths, they called out MongoDB’s strong market presence, which they described as “one of the most successful entrants into the DBMS market in the past decade” and cited their effective move to the cloud.

MongoDB received positive feedback from Gartner clients for customer satisfaction. For Gartner Peer Insights, 96% of respondents said they would recommend MongoDB — an extremely strong result that suggests that when used for use cases that are appropriate, MongoDB is an excellent choice.

Gartner Magic Quadrant for DBMS, December 2022

Gartner also highlighted very positive customer satisfaction scores among their clients (which tend to be large, traditional enterprises), with 96% of respondents saying that would recommend MongoDB. Finally, Gartner called out MongoDB’s expanding product vision, offering analytics support, SQL capabilities and new multi-model capabilities like time series and search.

There were a few cautions, as with all provider reviews. Gartner pointed out that the document model implicitly discourages heavy use of JOINs, which is at the foundation of a normalized relational model. The reliance on a JSON-like data structure may be unfamiliar to developers accustomed to working with relational databases.

Gartner also mentions MongoDB’s limited support for data science. The platform does not offer built-in capabilities to support models, algorithm libraries or feature stores. This limits MongoDB’s use for augmented transaction use cases. This will likely be a future development vector for the MongoDB product team, as they view MongoDB as appropriate for in-app analytics workloads.

Forrester Wave

As another data point from industry analysts, in November, Forrester released their Wave report on Translytical Data Platforms. In their report, they compared translytical data platform providers across 26 criteria. To do that, they identified the 15 most significant providers, which included Aerospike, Cockroach Labs, Couchbase, DataStax, GigaSpaces, GridGain, IBM, InterSystems, Microsoft, Oracle, PingCAP, Redis, SAP, SingleStore and of course MongoDB.

Translytical data platform is a relatively new term that has been embraced by Forrester. Comparing this to Gartner’s blanket DBMS category, I like Forrester’s partitioning better because it focuses in on the workloads that are relevant for modern, data-driven application development. They rightfully exclude the data platform providers that are dedicated to stand-alone analytics and machine learning processing, like the modern data warehouse / lakehouse architectures.

Translytical platforms are next-generation data platforms that are built on a single database engine to support multiple data types and data models. They are designed to support transactional, operational, and analytical workloads without sacrificing data integrity, performance, and analytics scale. Adoption of these platforms continues to grow strongly to support new and emerging business cases, including real-time integrated insights, scalable microservices, machine learning (ML), streaming analytics, and extreme transaction processing. Translytical data platforms are highly optimized for both reads and writes, leveraging distributed in-memory, multimodel, advanced workload management, AI/ML, and cloud architectures to support modern workloads.

Forrester Wave, Translytical Data Platforms, Q4 2022

Translytical data platforms are optimized for transactional workloads like a traditional OLTP database, but add support for real-time analytics, which is increasingly becoming a requirement for data-rich applications. To accomplish this, translytical data platforms operate on a consolidated database engine and support multiple data types. They can serve as the data source for real-time transactional workloads spanning microservices, streaming analytics, machine learning inference and high volume user applications.

Against this backdrop, MongoDB was evaluated within a pool of 15 providers. Forrester placed them in the Leaders circle, with only four other vendors. They ranked highest among providers that exclusively focus on database solutions with a single product. MongoDB was ranked highest on 11 out of 26 of the scored criteria.

Forrester’s review of MongoDB’s solution included a number of positives. They call out the pace of market share growth, driven by aggressive addition of new capabilities to support various translytical workloads. They cited a number of use cases by reference customers including real-time analytics, systems of insight, customer 360, IoT and mobile apps.

MongoDB continues to grow its translytical market share with new capabilities to support workload isolation, intelligent memory tiering, multimodel, streaming, and distributed transactions to support various translytical workloads. Organizations use MongoDB to support real-time analytics, systems of insight, customer 360, internet of things (IoT), and mobile applications. MongoDB’s roadmap includes improved performance and workload isolation, additional connectors, expanded self-service, and greater extensibility. MongoDB has strengths in multimodel, data modeling, streaming, fault tolerance, dev tools and APIs, and a broad set of use cases. One reference customer said, “MongoDB has added new functionality almost every year that has added to our capability to utilize the data we already have in our systems.”

Forrester wave, Translytical data platforms, Q4 2022

On the downside, Forrester commented that MongoDB’s extensibility is lagging. Some customers expressed concerns around “handling a high volume of data” and “ultra-low latency.” The MongoDB team has been addressing these limitations over the past couple of years, making notable progress in workload isolation, partitioning, distributed transactions and performance. There is more work planned in the product roadmap.

Forrester summarized their review with praise for MongoDB’s strengths in developer adoption. I think Forrester’s review aligns well with MongoDB’s product positioning and demonstrates their traction in addressing the needs of the modern enterprise for a broad-based data platform that enables high productivity for developers.

Investor Take-aways

After a disappointing break in momentum associated with the Q2 earnings report, MongoDB got back on track in Q3. While not a perfect quarter, it was much improved, particularly considering the macro backdrop. While other software infrastructure peers were posting smaller beats on the most recent quarter with a minimum or no raise over estimates for the next, MongoDB outperformed expectations by a healthy margin.

Additionally, Q2 suffered from a reversal in improving operating leverage. The Q3 report brought this back into alignment as well, with strong upward movement on gross margin and operating margin. This drove a huge beat on EPS for Q3 and a substantial increase for the full year target on Non-GAAP operating income.

Customer activity reinforced the strong demand for MongoDB’s platform, demonstrated both in an improvement in Atlas growth and a surprise surge in EA licensing. Total customer additions, Direct Sales and $100k customers all returned to the higher end of their historical ranges. The incremental investment in Sales and Marketing to form deeper relationships with larger customers through Direct Sales appears to be bearing fruit.

Third-party analyst firms issued positive reviews of the MongoDB platform as well, placing their solution at the high-end of rankings relative to competitive offerings. MongoDB made its debut onto the Gartner Magic Quadrant for DBMS in the Leaders Quadrant, which is a rare achievement for a newcomer. Similarly, Forrester ranked MongoDB as a Leader among 15 competitive providers in the modern Translytical Data Platform category. They gave MongoDB the highest score in 11 out of 26 criteria.

The positive momentum was reflected in MDB’s stock price, surging over 20% the day after Q3 earnings and continuing up to hover around $200 currently. This is still below the initial close around $240 following the Q2 report, and well above the lows before Q3 earnings.

Looking at valuation, MDB has a P/S ratio of 11.3 currently for 47% annual growth and positive Non-GAAP operating income. This is down from the 15-20x range prior to the Q2 report in 2022. Pre-Covid, this ratio was also above 15x, with higher revenue growth but worse operating margin.

Analysts currently estimate that MongoDB will finish this year (FY2023) with $1.260B in revenue, which is the high end of MongoDB’s range issued with the Q3 report. This would represent annual growth of 44.2% over the prior year. For next year (FY2024), they have modeled just $1.579B for 25.3% growth. Looking to FY2025, the projected growth rate ticks up a little to 28.3%. Analysts are modeling a substantial slowdown for next year and linear growth the following.

While the macro environment introduces a wildcard, I think that MongoDB can beat these estimates. I would point to renewed momentum in the Q3 report, the large expansion opportunity with existing customers and healthy growth in total customer additions. With a targeted TAM of $109B by end of 2024, MongoDB’s penetration would be about 2% of the market at that point.

I currently have a 6% allocation to MDB in my portfolio. I will likely add to that on dips, but wait until the next quarterly report before making it a much larger position. Given that MDB is still priced below the post Q2 earnings price drop, it may provide a reasonable entry point for new investors. The macro environment and associated pressure on IT budgets over the next year will impact software valuations across the board. Relative to other software infrastructure stocks, though, I think MongoDB is well positioned as enterprises appear willing to invest in the platform through multi-year commitments.

NOTE: This article does not represent investment advice and is solely the author’s opinion for managing his own investment portfolio. Readers are expected to perform their own due diligence before making investment decisions. Please see the Disclaimer for more detail.