MongoDB (MDB) Stock Q1 FY2021 Earnings Results Review

MongoDB (MDB) released their Q1 (April end) FY2021 earnings report on June 4. They reported a significant beat for Q1, exceeding revenue estimates by 12% and improving profitability metrics. Guidance for the remainder of the year was mixed, with revenue targets raised slightly and profitability mostly inline. Most impressive was growth in the cloud offering, Atlas, which now contributes 42% of total revenue. MongoDB also released a number of product enhancements just after the earnings report, coinciding with their annual user event. MongoDB is continuing to expand their reach beyond a core database solution into ancillary functions that enable easy application of data to common use cases, like mobile apps, analytics, data visualization and search. In this post, I review MongoDB’s earnings results and other business updates. I also dig into the product releases and competitive landscape. For a refresher on the MongoDB investment thesis, please see my full analysis published in November of last year.

Headline Financial Results (EPS is Non-GAAP)

• Q1 FY2021 Revenue of $130.3M, up 46% year/year. This compares to the consensus estimate of $119.7M, which would have represented growth of about 34%. MongoDB beat estimates by 12% of annualized growth. Q4 FY2020 revenue growth was 44%, so a slight acceleration.
• Q1 Non-GAAP net loss was $7.3M for an EPS of ($0.13) vs. ($0.25) consensus, representing a beat of $0.12. This compares to Q1 FY2020 net loss of $12.1M or ($0.22) per share. Q4 FY2020 EPS was ($0.25).
• Q1 Non-GAAP operating loss was $7.4M, for an operating margin of -5.7%. This compares to an operating loss of $12.6M in the year-ago period, representing an operating margin of -14%. Q4 operating margin was -10%.
• Q1 FCF was ($8.5M) for a FCF margin -6.5%, compared to FCF of $2.8M or FCF margin of 3.1% in the year-ago period.
• Q2 FY2021 Revenue estimate of $125-127M, representing growth of 26.7% at the midpoint. This compares to the consensus revenue estimate of $120M, or 21% growth. This represents a raise of almost 6% of annualized revenue.
• Q2 FY2021 Non-GAAP operating loss of $(24.0)M to $(22.0)M, or an operating margin of -18% at the midpoint.
• Q2 EPS of ($0.41) to ($0.38), compared to ($0.36) consensus estimate.
• FY2021 Revenue estimate of $520-530M, representing growth of 25% at the midpoint. This compares to the consensus revenue estimate of $517.6M, for growth of 23% and the company’s prior guidance of $520M at the midpoint.
• FY2021 operating loss of $(78.0)M to $(70.0)M, for an operating margin of -14% at the midpoint.
• FY2021 EPS of ($1.34) to ($1.21), compared to ($1.34) estimated.
• As of April 30, 2020, MongoDB had $977.5M in cash, cash equivalents, short-term investments and restricted cash. 

Other Performance Indicators

• Subscription revenue was $124.9M, growing 49% year-over-year. Services revenue was $5.5M, an increase of 1% year-over-year. Growth in subscription revenue is a better indicator of demand for MongoDB’s core products. In Q4, subscription revenue grew by 46%, so Q1 represented an acceleration.
• Non-GAAP gross margin was 73% in Q1 FY2021. This compares to 70% in Q1 FY2020. Gross margins will probably remain around this level. MongoDB is driving efficiencies as usage grows, but also will experience headwinds to gross margin as use of Atlas increases (due to hosting costs).
• MongoDB Atlas (cloud offering) revenue increased 75% year/year and now makes up 42% of total revenue. The growth of Atlas is impressive and represents the main driver of future expansion as more enterprises conduct cloud migrations and re-architect software offerings. Additionally, on the earnings call, the CFO mentioned “if you actually back out the contraction or the drag from mLab, organic Atlas more than doubled.” Atlas revenue was 35% of total in Q1 FY2020 and 41% in Q4.
• Net ARR expansion rate was above 120% this quarter, which it has been consistently for many quarters. 
• Breaking down Q1 Non-GAAP expenses, we see slight year/year reductions in relative percentage across all categories.
  • R&D = 26% (versus 29% in Q1 FY2020)
  • S&M = 45% (versus 46% in Q1 FY2020)
  • G&A = 11% (versus 13% in Q1 FY2020)
• Had a stronger than expected quarter in terms of new business coming from Enterprise Advanced, which is the self-hosted solution. This helped drive revenue outperformance due to the upfront licensing revenue recognition component under ASC 606.
• After the global lockdown began in mid-March, leadership observed a modest slowdown in the growth from existing Atlas customers, particularly in self-serve. This spanned many industries and wasn’t concentrated. MongoDB recognizes Atlas revenues based on consumption, so the slowdown would directly impact Q1 Atlas revenue performance (which was still high). CFO stated they haven’t seen any increase in customer churn or slowdown in new customer adds. So, under normal circumstances, Atlas revenue would have been higher in Q1.
• MongoDB experienced strong customer growth metrics for Q1.
  • 18,400 total customers, representing year/year growth of 30%
  • 2,200 direct sales customers, for year/year growth of 22%
  • 16,800 Atlas customer, with year/year growth of 37%
  • Customers with ARR over $100k grew 30%

• During Q1, leadership talked about customers in even some of the most negatively impacted industries double down on their spend as part of digital transformation efforts.
  • Mentioned that Johns Hopkins University’s COVID-19 dashboard is hosted on MongoDB Atlas.
  • Closed a seven figure deal with one of the largest global auto companies. The company is making MongoDB as a service available to internal users on their private cloud. There are currently more than 1,000 MongoDB servers in production, supporting critical use cases within the various areas of the company’s digital support infrastructure, including their connected car initiative.
  • One of the world’s most popular consumer video chat applications is built on MongoDB technology. It was able to withstand a 120x increase in concurrent users during the weekend of March 14. Leadership believes that the current social distancing environment will only increase these kinds of transformations.
  • Expanded their Atlas relationship with a leading North American airline. Their strategic focus is to accelerate their move to the cloud in order to modernize their applications and reduce their dependence on the mainframe. MongoDB is helping them build an operational data layer in the cloud.
  • Boxed, a wholesale grocery delivery service, doubled down on MongoDB Atlas and started using MongoDB Charts to help with capacity planning. Usage of their service has soared and they needed to ensure they could scale to meet the extreme increases in customer demand.
  • Sanoma Learning, the leader in interactive and personalized online learning in northern Europe, turned to MongoDB on March 13 to handle the increasing demand when the government shutdown all schools because of COVID-19. The company rapidly migrated services to MongoDB and now effortlessly handles more than 12M online exercises a day.
  • Forbes migrated its digital asset management system to use Atlas. The publisher has been working with MongoDB to transition more of its digital footprint to MongoDB. Forbes credits MongoDB for improving its ability to serve dynamic content, decreasing its total cost of ownership and making it possible to replace some of its legacy technologies.
  • Zomato, one of the largest restaurant discovery and food delivery service in the world with over 80M monthly active users in 24 countries, recently increased its commitment to MongoDB to power its logistics application. This helps Zomato ensure that are able to keep up with the increased demand, efficiently map their food delivery orders to the right delivery partner, track their journey and ensure on-time deliveries.
  • French multinational Schneider Electric, which provides energy and digital solutions for sustainability, recently expanded its commitment to MongoDB. The company chose MongoDB to help rapidly scale its new IoT enabled platform EcoStruxure. The company leverages Atlas to reduce costs and help its team manage large volumes of data more effectively.
  • Nets Group, the leading payment service provider in Scandinavia and across Europe, chose MongoDB as its anchor data platform to modernize its payment services and establish a distributed microservices architecture. With MongoDB, the Nets Group makes it even easier and more intuitive for its customers to handle digital payments and related solutions.
  • Woolworths Group, one of the largest retailers in Australia and New Zealand, decided to start offering digital receipts to it’s 11.7M reward customers in order to minimize human contact during COVID-19. The retailer used Atlas to create a new platform to ingest all of its point of sale data, with more than 350 transactions per second, and serve it out to customers in real time. MongoDB was able to help them get the system up and running in under three weeks.

• Made some adjustments to marketing approaches in response to COVID-19, particularly around the self-service channel. Invested more in online advertising to promote Atlas, as a result of lower online ad rates.
• Have over 40,000 people registered for the revamped MongoDB.Live event, which is a replacement for the in-person annual user conference. In comparison, in the prior year 2,000 people attended the in-person event.
• Transitioned their summer internship program to be a remote experience. Will have 83 interns chosen from the over 20,000 applicants. Interest in working at MongoDB is off the charts.
• On guidance, leadership offered the following commentary. In March, they stated that they expected a greater impact on new business activity in Q2 as compared to Q1, as Q1 benefited from deal pipeline already in place when the pandemic started. They still expect that to be the case and now expect the Q3 environment to be similar to that of Q2, followed by a more modest improvement in Q4.
• Regarding Atlas specifically, leadership observed weakness in existing customers expanding their spend. There hasn’t been an increase in churn or slowdown in new customers. Just that existing customer spend growth leveled in some cases, as either the customer’s business decreased, or they entered into a cost containment mode. No substantial increase in customers requesting discounts.
• Leadership also talked about continued investment going forward, rather than pulling back on spending. Due to the constraints caused by the prolonged COVID-19 disruption, MongoDB is generating savings in travel, event and facility expenses. Leadership decided to reinvest those savings in two high-return areas. First, online advertising rates have declined significantly, in some cases as much as 30%, since the pandemic started. This provides an excellent opportunity to increase investment in digital marketing, to expand upon recent strength in self-serve customer new additions. Second, in the current environment, they are finding an opportunity to incrementally add strong engineering talent that is not normally on the market and is now looking to join a well-funded company. As a result, they have decided to pull forward a portion of expected fiscal year 2022 R&D hiring to take advantage of talent availability and further accelerate their ambitious product roadmap.

Analyst Conferences

MongoDB participated in two analyst conferences following earnings results. These were the Stifel Cross Sector Insight Conference on June 8th and the William Blair Growth Stock Conference on June 10th. In both cases, the primary MongoDB representative was CFO and COO Michael Gordon.

I listened to both calls. Here are my notes, based on Michael’s comments.

• Why do customers adopt MongoDB vs. cloud provider solutions? Two main reasons. First, the MongoDB product offering is more feature complete, particularly given that competitive solutions are built in-house with limited resources or pinned to an old version of MongoDB software as a result of their licensing change. Second, customers are incredibly concerned about getting locked into one cloud provider and value the optionality to easily shift workloads without changing the underlying software application.
• Google is only cloud provider that doesn’t offer a competitive product. Google likes to lead with best of breed solutions, rather than trying to build all IT solutions themselves as cloud service add-ons. MongoDB even won a Google award – partner of the year.
• AWS and Azure can’t use the MongoDB open source project due to license restrictions. Product quality for their solutions isn’t there, fails MongoDB’s own validation tests. Have other database products, but really just a shelf space game. Want to keep customers in their ecosystem. Customers are very concerned about cloud vendor lock-in. Don’t want to have a proprietary database solution that they can only run in one cloud. Can move application executable, but don’t want two different sets of code for different database implementations.
• The Alibaba relationship is progressing. They license MDB for their users to utilize. As usage increases, MDB would generate more revenue.
• Database is the largest sector of IT spend. Projected to be $63B in 2020 and growing to $89B in 2024. Currently, dominated by legacy vendors like Oracle. As enterprises re-architect software applications, they will often revisit their database selection. This is where MongoDB can win new business.
• In addition, enterprises are increasingly competing for customers based on their technology, which relies on custom software applications. Databases are the heart of every software application. MongoDB is winning a larger share of new software application business.
• Very excited about the Realm acquisition and its ability to provide a ready-made solution for mobile. Have an aggressive product roadmap overall. Redeployed savings from travel and facilities to hiring in R&D to accelerate feature delivery.
• Multi-document ACID support, which reduces data loss, was introduced over a year ago. This was the main technical barrier to adoption of MongoDB by customers as a system of record.
• Snowflake is not a direct competitor. Feels like an adjacent product. Snowflake is for data warehouse. Don’t want to run transactional queries on the data warehouse. Requires a schema to make queries run faster for large analytics. Don’t see Snowflake competitively – comes up more on investor calls than in the marketplace.
• Channel view of Atlas. Self-serve channel is larger, but spend level per customer is lower ($6-7k each). Direct sales customers spend >$100k year. Atlas is consumption oriented for pricing. Self-serve cohort also experiences more churn – experimentation, not definite business lines sometimes. Self-serve are paying with credit card in arrears. Direct sales signs a contract and pays in advance. Self-serve is just over 20% of revenue currently.

Analyst Reactions

Following MongoDB’s Q1 earnings on June 4th, eight sell-side analysts provided updated coverage reports. They all raised their price targets, some more than doubling it, in reaction to the results. Five analysts rated the stock at a buy equivalent, while three maintained their neutral rating. The average price target for these updates is about $236, representing a 19% increase over the closing price on June 5th of $197.98.

Date	Analyst	Rating	Price Target
6/2	Needham	Buy	Raised from $170 to $253
6/2	Barclays	Overweight	Raised from $130 to $275
6/5	Oppenheimer	Buy	Raised from $140 to $240
6/5	Canaccord	Buy	Raised from $130 to $245
6/5	Piper Sandler	Overweight	Raised from $138 to $226
6/5	DA Davidson	Neutral	Raised from $105 to $215
6/5	Needham	Buy	Raised from $253 to $259
6/5	Barclays	Overweight	Raised from $275 to $280
6/11	Morgan Stanley	Equal Weight	Raised from $203 to $205
6/24	Goldman Sachs	Neutral	Raised from $169 to $215

Prior to the earnings results, Barclay’s set the highest price target, substantially raising the price target from $130 to $275. After earnings, they raised again to $280. Analyst Raimo Lenschow provided the following commentary.

Barclays analyst Raimo Lenschow raised the firm’s price target on MongoDB to $275 from $130 and keeps an Overweight rating on the shares ahead of the company’s Q1 results. MongoDB will see some growth pressure from the pandemic like most other software vendors, but Street estimates factor in an appropriate level of conservatism, which positions the company for a beat, Lenschow tells investors in a research note. The analyst expects a “beat and raise” for Q1, but believes that is already priced into the stock and thus does not see the print as a share catalyst. However, Lenschow continues to see “scarcity value” for a long term structural growth story like MongoDB and believes the current valuation level remains attractive.

TheFly.com, june 2, 2020

After earnings, Canaccord issued the second highest price target, raising from $130 to $245. Analyst Richard Davis provided the following commentary.

Canaccord analyst Richard Davis raised the firm’s price target on MongoDB to $245 from $130 and keeps a Buy rating on the shares. The analyst noted its strong Q1 results with revenue, billings, operating income and cash flow each topping estimates. He believes the company’s premium valuation is justified and is in rarefied air when it comes to TAM and long-term growth prospects.

Thefly.com, June 5, 2020

Neutral ratings were primarily concerned with valuation. Here is one example from DA Davidson.

DA Davidson analyst Rishi Jaluria raised the firm’s price target on MongoDB (MDB) to $215 from $105 and keeps a Neutral rating on the shares. The company’s Q1 results were solid but shares traded lower after-hours on deceleration seen in its Q2 guidance, the analyst tells investors in a research note. MongoDB is benefiting from accelerating digital transformation, Jaluria adds, as Enterprise Advanced outperformed in the quarter and Atlas had strong momentum. Jaluria further notes that MongoDB’s offerings continue to compete well against Amazon (AMZN) and Microsoft (MSFT).

Thefly.com, June 5, 2020

While sell-side analyst ratings can get out of synch with the market, I was surprised by the magnitude of price increases for most analysts in the case of MongoDB this quarter.

MongoDB Product Development Activity

I’ll provide some updates on the MongoDB product roadmap since the last earnings release. If you aren’t familiar with the design and use cases for their underlying database technology, please see my original analysis. MongoDB also provides an explanation on their site.

The major product releases in this cycle coincided with MongoDB’s annual user conference called MongoDB.live, held June 9 -10. If you have time, it is worth watching the Keynote, which is led by the Chief Product Officer and provides more details on all the product releases, with demos. There were a lot of features packed into this release. I will provide a summary of the changes below and try to explain how each contributes to MongoDB’s competitive position and broader market opportunity.

MongoDB Database 4.4

As part of the 4.4 release, MongoDB addressed a number of features that assist with scaling clusters, read performance and programmability.

• Refinable Shard Keys. Database sharding is a mechanism to increase the storage and processing capacity of a database cluster by splitting the dataset along a specific data field, which is referred to as the shard key. The data is then directed to different clusters, depending on the value of the shard key. An example might be the customerID, where IDs in one range are directed to one cluster and IDs in the other range go to another cluster. The challenge with this approach has been that once these ranges or shard keys are set, they are difficult to alter. Data storage and access patterns may change as an application evolves, making data utilization uneven across clusters. Rebalancing a MongoDB cluster in the past to reset a shard key was a major operation. With this new feature in MongoDB of refinable shard keys, updates to the sharding approach can be applied in real-time. If the data set grows even more, an existing shard key can be further subdivided by an additional data field. This addresses a major user complaint with MongoDB clustering in the past, that its ability to scale to large workloads was limited, due to constraints with the sharding strategy.
• Hedged Reads. When a database query comes into the MongoDB cluster, a service called the query router dispatches the request to one of the cluster’s read replicas. If that particular server is busy, the query response could be delayed, causing a spike in latency for any queries sent to that replica. Hedged Reads optimizes for the fastest response by dispatching queries to multiple read replicas and returning the results from the quickest response. This will help smooth out variances in read times, referred to as P95 and P99 latency.
• Union. A union is a common operation available in SQL, which basically combines the result sets of two different queries (with matching columns, data types, etc.). MongoDB added the union operator to its query language. This effectively performs a similar function to the SQL union operator, allowing the user to blend data from multiple collections into a single result set for analysis.
• New Drivers. A database driver provides the connection between a particular development framework and the database itself. It translates common data requests into the protocol and query language used by the database. In order to ensure broad adoption with developers, it is important to provide drivers for most languages. MongoDB already has database drivers for many development languages, like Java, Node.js, PHP, Ruby and C#. This release added two more languages popular with developers, Rust and Swift.

MongoDB Atlas

MongoDB offers a fully managed, cloud-hosted version of the database called Atlas. This is available on all major cloud providers (AWS, Azure, GCP). The big advantage of utilizing this service is the neutrality of it. Cloud vendors with their own proprietary databases (at least the flavors of NoSQL) result in lock-in. The application developer would need to customize the software to utilize the query language for that database. This makes a multi-cloud strategy more cumbersome. With a single interface to MongoDB Atlas, one set of application code can handle database interactions universally across all cloud providers. MongoDB leadership claims this represents a major advantage, which I can attest to. As a CTO/VP Eng, I wouldn’t want to have to rework an application in order to migrate it to another cloud provider. There will always be some possibility of a move – cost, data locality, business requirements. I once had to move workloads from AWS to Azure simply because some of our customers were concerned that Amazon competed with them.

As discussed in the earnings results, Atlas represents MongoDB’s fastest growing product offering with year/year growth of 75% (even higher if you back out the mLab transition). In addition to the features associated with the core database 4.4 release, MongoDB added the following items to Atlas.

• Atlas Online Archive. This provides a mechanism to archive data that is used less frequently. Archiving in this case means moving the data to less expensive storage to reduce operational costs. If the data is needed, it would be restored from the offline archive. There would be a slight delay for the first request to access the archived data and then it would serve normally. This capability provides customers with a cost-effective way to scale their data usage.
• Schema Suggestions. This feature monitors the performance of the database and generates suggested changes to the database schema to improve performance. These recommendations are made passively and are formed by monitoring database logs and examining database schema metadata. The service locates common anti-patterns and surfaces suggestions for the database administrator to address these. It works in tandem with the Performance Advisor, which monitors query logs and recommends indexes that can improve performance. Indexes can then be applied with a click. Query performance of MongoDB has been a user complaint in some circumstances. Often, the issue can be traced back to bad choices with the schema design. While the document database model allows for very flexible data storage, it does perform better if a schema is pre-defined that reflects expected data access patterns. Indexes on data fields commonly used for data filtering particularly help performance.
• AWS IAM Authentication. IAM allows system administrators to control access to resources on AWS. This MongoDB feature enables developers to leverage IAM to authenticate from applications to MongoDB Atlas clusters. It represents a deeper integration with AWS infrastructure for managing MongoDB.
• Atlas Search GA. Atlas search was announced last year as a new feature to enable full text search of data on a MongoDB cluster. In this release, MongoDB moves the feature to full GA in Atlas and adds new data types beyond text for indexing. It also supports faceted search in addition to full-text search. Search capabilities are enabled through the open source Apache Lucene library, which is the same technology utilized by Elasticsearch and Solr. MongoDB’s search implementation includes the most common search features, like autocomplete, custom scoring, fuzzy matching, stemming, etc. The addition of search functionality to the MongoDB database is an interesting move by MongoDB. Their rationale is that developers can enable common search use cases on the MongoDB cluster directly, without having to stand up a separate search service (like Elastic or Solr). I think this would be useful for basic search functionality – like text search, product filters, etc. In this way, it expands the set of use cases that MongoDB Atlas can address, which would increase usage and revenue. I don’t think it will pose a significant threat to full-featured search infrastructure, like Elastic, but does make basic use cases simple for a developer to deliver on a single stack.

Realm

I think Realm is one of the more exciting product extensions from MongoDB. Realm provides a multi-language, cross-platform SDK that dramatically simplifies data management for disconnected mobile or web application development. It is similar to the Firebase service offered by Google, which has been very popular. By wrapping this capability into MongoDB, Realm will appeal to development teams with a MongoDB back-end looking to build a new mobile app.

Managing data state efficiently for a mobile app can become very complicated, particularly when trying to provide a local cache of data for speed and to support offline functionality. Keeping data in synch between this local data store and the back-end can require a lot of overhead code. Realm makes this trivial by providing the development framework and data management infrastructure to let developers offload most of the data management tasks.

As part of the general announcement associated with Realm, MongoDB added a couple of features. They also revealed that 7-Eleven is using Realm in an app for store managers that allows them to manage inventory and accept product shipments. It has been rolled out to 8,500 stores and 20,000 devices.

• Released Realm Sync to beta. Sync provides intelligent data synchronization between user devices and the back-end data store. This works across device types (iOS, Android, web) and connectivity states (online, offline). Realm provides a compact data store for the device and seamless sync with the back-end database on MongoDB Atlas.
• Released their GraphQL service to GA. GraphQL is a popular query language for mapping data requests to API endpoints. This provides a layer of abstraction for app developers by organizing data request/response in a way that maps to their app usage, versus the sometimes disjointed structure of REST APIs. With Realm GraphQL, MongoDB is extending the same simplicity to app developers. They can generate a GraphQL schema from their MongoDB database collection through a simple UI click.

Analytics

As another extension of the core database, MongoDB is providing solutions for basic analytics. These span two use cases currently. The first is to enable data querying across a broader data set than what is stored in the MongoDB database itself. The second is the Charts functionality, which supports the ability to create data visualizations directly from data in MongoDB.

• Data Lake. The MongoDB data lake allows users to query data across MongoDB Atlas and offline file storage (typically AWS S3) using the native MongoDB Query Language. Files on S3 can be in multiple data formats, including JSON, BSON, CSV, Avro and Parquet. This capability is useful because data will often be stored outside of the operational database to reduce cost or because it has different context.
• MongoDB Charts Embedding SDK. This capability allows the developer to embed charts of MongoDB sourced data directly in web applications. It leverages a JavaScript based SDK to programmatically generate and interact with charts using the controls of the app. Creating data visualizations on web pages often required the overhead of importing a separate Javascript library for data (like D3.js), mapping database values to the library’s internal structure and configuring the controls. MongoDB’s solution with embeddable Charts simplifies this flow by providing an out of the box solution.

DevOps Tools and Integrations

In order to make MongoDB easier for DevOps personnel to manage, the team launched a few new tools.

• MongoDB Shell. While administrative functions for MongoDB are available through a web UI and APIs, most modern database solutions provide a command-line interface (CLI) for administrators. This is often referred to as a shell. The MongoDB Shell does the same. Users can connect to the database, work with data and perform configuration functions.
• MongoDB for VS Code. Visual Studio Code is a popular code editor for developers. MongoDB created an extension for the tool that allows developers to connect directly to their MongoDB instance, view the schema, test CRUD operations and access the Shell. Having database connectivity tools built into code editors is standard practice.

Additional Kubernetes Support

This release added a few additional capabilities for managing MongoDB instances using the Kubernetes container orchestration system. The MongoDB team launched a new Operator for Kubernetes that allows users to automate the configuration, deployment and management of containerized MongoDB replica sets. This has been open sourced and is freely available. They also provided a Containerized Ops Manager for enterprise deployments. This makes it easier for Ops personnel to deploy, maintain and upgrade the management platform for MongoDB.

Competitive Landscape

I could write volumes covering all existing players in the database space. I tried to provide a foundation for the competitive landscape in my original analysis on MongoDB. Readers can reference that for background. For the purposes of this quarterly update, I will try to highlight new developments or changes in the competitive landscape. Additionally, I think that MongoDB’s high-level strategy for differentiation is sound and still positions it well against offerings from the cloud providers. If anything, MongoDB’s competitive position has improved over the last year, as their focus on providing a full-featured, cloud-neutral solution resonates with customers.

The first thing to consider around competitive offerings is the breadth of the data platform that MongoDB is constructing. While the core of the MongoDB offering is a document database, MongoDB is quickly surrounding that with an ecosystem of tooling that make common data storage and access tasks simpler for developers. I think this is where some investors stumble with the MongoDB opportunity – they get bogged down in the relational (SQL) versus non-relational (NoSQL or document-oriented) debate. First, many engineering organizations will utilize both types of databases. It’s not an either/or decision. With the move away from the monolithic software architecture and proliferation of microservices, modern engineering organizations will deploy multiple flavors of data storage, selecting the best tool for each function. Perhaps Postgres or MySQL (relational) for their payments service, but MongoDB or Cassandra (non-relational) to store user profiles, shopping cart data or frequently accessed usage metadata. The key consideration is that the database market is projected to be $89B by 2024. There is room for many providers and vendors who get small slices of this market will still do very well.

Over the last few years, MongoDB has moved beyond just providing a database solution. This is where they are building competitive advantage. They are expanding into developer-friendly services that surround the database, all packaged into one platform and delivered on the cloud (through Atlas). These expanding use cases now include tightly coupled data services for mobile app development, basic search capabilities, the ability to query a data lake and charting for common data visualization tasks. These capabilities are not available out of the box with other popular database solutions, whether MySQL/Postgres or cloud vendor offerings like Amazon DocumentDB. It’s likely that MongoDB will continue pushing the envelop in multiple directions, evolving from just a core database solution to a full-featured data platform. Engineering organizations will find this additional functionality appealing, particularly where delivered as a managed, multi-cloud service. It is this product trajectory that makes MongoDB interesting to me, as the feature set will build a broader moat around their platform when compared to other stand-alone database technologies.

StackOverflow Survey

As a measure of the appeal of MongoDB to professional developers, Stack Overflow conducts an annual survey in which they ask 65,000 developers about their preferences across a number of technology types. Included is input on programming languages, frameworks, tools and platforms. MongoDB is included in the databases category. For this year’s survey, conducted in February 2020, developers rank MongoDB as the most “wanted” database for the fourth year in a row. This reflects developers who are not working with the database, but have expressed a desire to use it.

I think the take-away here is that MongoDB is viewed as a viable solution amongst professional developers for addressing database functionality. While there will be healthy debates about pros/cons of applying MongoDB to certain workloads, this general perception is important and will allow MongoDB to be included in enough database selection exercises to win a reasonable share of the large database market.

Beyond this, I haven’t observed significant changes to the landscape of independent database offerings. Popular open source alternatives in the relational space remain as MySQL, Postgres (aka PostgreSQL) and MariaDB. Non-relational options are Redis, Cassandra and Couchbase. Each of these have their own advantages/disadvantages. The key consideration is that the database market is huge and there is room for many solutions. MongoDB is still popular and is rapidly expanding its platform of capabilities as a consequence of its commercial scale. Last quarter, MongoDB spent $46M on R&D.

Cloud Vendors

The cloud vendors have made significant efforts to capture share of the database market. Like the other services they have built around core compute and storage, their intent is to capture as much spend from enterprises moving infrastructure to the cloud as possible. This results in them engaging in a “shelf space” game, where they try to have an offering in every category. The effectiveness of this seems mixed to me. In some categories, they can offer a best of breed solution. In others, they are trying to pick up some spend with a basic feature set, but lag competitive offerings. This approach seemed disruptive 2-3 years ago, as the cloud vendors launched many look-alike products. With their size and resources, the market assumed the cloud providers would roll over the independents. However, independent leaders in each category continued to iterate on their feature sets and are now leaving the cloud providers behind in many categories. Examples include services from Okta, Twilio, Datadog, Elastic, Crowdstrike, Fastly and others. We could even apply the same rationale to products outside the developer ecosystem, like video conferencing (Zoom Video). While the cloud providers arguably have extensive resources, they can’t be expected to compete effectively in every category. This is particularly true around engineering mindshare. The best engineering talent in application-oriented solutions will gravitate towards the independents. The upside is greater, as the cloud vendors offer less growth opportunity given their size.

Another controversy has been the tactic employed primarily by Amazon of taking an open source project, wrapping it as a cloud service and charging fees to host it. In these cases, the cloud provider’s developers did little to contribute to the project, but as open source, they were legally allowed to engage in this practice. I won’t debate whether this is fair use. However, most companies built around open source projects have modified their licensing arrangement to prevent the cloud providers from engaging in this behavior. Elastic and MongoDB both made these changes, forcing the cloud providers to either pin their offering to an old version or fork the project in hopes of engaging the open source community in contributions to their version. As you can imagine, this approach is becoming less effective as time passes. The breadth and quality of the solution set offered by the independents continues to create distance. The independents still employ the vast majority of the core contributors to these open source projects.

Not all cloud providers are taking this approach of launching look-alike products. Google is starting to take a noticeably different tact. They are embracing the best-of-breed independents. They formed a Partner program “to make it easier to work with us and to innovate with us”. They even host an awards program, in which they recognize the best partners by segment. For the technology segment, this past year’s winners included MongoDB, Elastic, Confluent, Box and Palo Alto Networks. In these cases, Google co-markets solutions from these best-of-breed vendors and shares in monetization. This strategy may be paying off, as at least one survey of CIOs indicated that Google’s share of IaaS business was increasing from 7.6% to 11.2% over the next 3 years, while share for AWS was decreasing.

Most important, though, as a differentiator for the independent providers like MongoDB continues to be their neutrality. As mentioned above, many CTO/CIOs worry about being able to maintain optionality and avoid lock-in to a single cloud vendor. This is critical for many reasons, but if nothing else improves their negotiating position. If the cloud vendor knows that the switching cost for porting a software infrastructure is high, they will be less accommodating at contract renewal time. Additionally, many engineering organizations like the idea of running their applications across multiple clouds, particularly where they have a global footprint. Having a single programming interface for database access simplifies the codebase.

There was some specific commentary about MongoDB’s relationships with the cloud vendors on the Q1 earnings call.

• The business partnership with GCP provides deeper product integration and unified billing for GCP customers who are also MongoDB Atlas customers.
• In 2019, MongoDB announced an expanded relationship with Microsoft. The new availability of MongoDB Atlas on the Microsoft Azure Marketplace provides unified billing for joint customers of MongoDB Atlas. Microsoft will make it easier for established Azure customers to purchase and use MongoDB Atlas. MongoDB is part of Microsoft’s strategic partner program.
• MongoDB was named the 2019 Google Cloud Technology Partner of the year for Marketplace.

(Regarding competition) Frankly we’re not seeing any real changes, obviously, I think we’ve talked about this in the past. Clearly, we partner with all the major cloud providers and Amazon and Microsoft or Azure in particular have their alternatives to MongoDB, but we’ve not seen any real change in the competitive dynamics. Frankly, when we are competing, we feel really, really good about our position. Those products are clones of MongoDB, and so when we expose the full feature set and all the capabilities available in MongoDB and through Atlas, it becomes a pretty easy decision. The things that we worry about frankly are the deals that we’re not a participant in. And so just as a virtue of their reach and their brand, they’re clearly gaining business that we just don’t always have access to.

That being said, Google, as we’ve talked about in the past, it doesn’t have a competitive product and the partnership there is very strong. They named us one of the technology partners of the year and the sales teams do a lot of joint planning, joint account planning and work in various — in Europe and in North America and other parts of the world. And so that business is growing quickly, but frankly our business to all the cloud providers are growing quickly. So we feel quite good — we feel very good about our value proposition.

MOngoDB CEO, q1 EArnings Call

On the earnings call, the CEO also mentioned a Canadian security company that recently migrated its mobile security platform from DocumentDB (AWS solution) to MongoDB Atlas. In addition to reducing cost by 60% and being able to leverage all the features of MongoDB, the key objective was to create a global multi-cloud foundation to rapidly scale IoT, AI and transactional workloads.

I won’t delve into the specific offerings of each cloud provider. Amazon advertises 15 database solutions of all types available to AWS customers. The closest competitor to MongoDB is DocumentDB, which offers compatability to MongoDB, but is pinned to the version 3.6 API, which was released in late 2017. The MongoDB leadership team asserts that DocumentDB fails most of their automated quality checks at this point. The customer set for DocumentDB is much smaller than MongoDB’s published list of customers, and this list hasn’t grown noticeably. Given that DocumentDB has been an Amazon offering for over a year, and hasn’t accelerated traction, I am not concerned about this as a viable competitive offering.

Microsoft Azure similarly offers a number of database solutions. SQL Server is their own proprietary solution for relational workloads. CosmosDB is their NoSQL database product. It advertises compatibility with both Cassandra and MongoDB. Cosmos DB was launched in May 2017. Similar to Amazon’s solution, the customer list appears limited.

Alibaba

In October 2019, MongoDB and Alibaba announced a new partnership in which Alibaba Cloud would offer customers an authorized MongoDB-as-a-service solution from Alibaba Cloud’s data centers globally. This was an interesting move, as it makes a cloud-hosted version of MongoDB available legally to Alibaba Cloud customers. Previously, Alibaba was allowing hosting of unlicensed versions of MongoDB. Customers now have access to all current and future versions of MongoDB, with the ability to escalate bug fixes and file support issues.

On the Q1 earnings call, the MongoDB leadership team provided an update on the Alibaba relationship. They stated that the relationship has been in place about 6 months and that demand is higher than they expected. MongoDB leadership isn’t factoring this into guidance, but feels good about the customer traction they are getting. As part of the arrangement, Alibaba sells MongoDB to their customers and then shares revenue on usage.

Other Independent Data Platforms

While the cloud vendors represent one dimension of competition for MongoDB, another comes from independent, commercial database providers. I will discount the offerings from legacy providers, like Oracle, Teradata, etc., as there hasn’t been meaningful change. However, a couple of newer players are emerging that should be monitored.

Snowflake

Snowflake confidentially filed for an IPO in early June. While it is not clear when this will happen, many analysts and investors are eagerly anticipating this offering. Founded in 2012 and growing rapidly, Snowflake represents the newest entrant onto the data management scene. The core product is a modern, flexible, cloud-based data warehouse, which naturally competes with Oracle and big data solutions from cloud vendors like Amazon Redshift. Snowflake describes themselves as a cloud data platform, which obviously can be very broad. In the developer section of their site, they highlight some use cases that overlap into adjacent data processing categories like observability, analytics, IoT and data science.

In Snowflake’s product videos, they represent OLTP databases (like MongoDB, MySQL, Postgres) as data sources (see diagram above on left side). OLTP stands for online transaction processing. Most software applications require a fast, developer-oriented database adjacent to the application server. This is the function that MongoDB, as well as other database solutions we have discussed, provides. While in theory Snowflake could be used as a transactional data store, I think this is unlikely in the near term. They are building out data connectors for most popular languages, but don’t have full coverage yet. Also, a lot of Snowflake’s big picture strategy is to facilitate a data marketplace for large, prepared data sets (demographics, weather, sales data, etc.). This usage would not be associated with the raw, transactional databases close to disparate customer applications.

This is further corroborated by a reference architecture diagram in the developer section of the site. We see the location of a OLTP/NoSQL database in front of Snowflake, presumably for most application workloads, with an ETL process to migrate summarized data to Snowflake. Interestingly, Snowflake includes the NoSQL label in this diagram, reflecting the fact that Snowflake’s underlying data structure is relational. This further separates it from MongoDB, which would be used for workloads that favor a non-relational, flexible database schema.

With that said, Snowflake has ambitious plans and could further expand their product reach in the future. As an aside, their product extensions could also represent encroachment into other segments of data processing, like analytics, observability and security monitoring. This will be something to monitor for other software stack players like Alteryx, Datadog, Splunk, Elastic, etc.

Datastax

Datastax provides an open-source, cloud-native NoSQL database. It is built on top of Apache Cassandra. The company created its own proprietary version of Cassandra, called DataStax Enterprise. This is cloud delivered, available on AWS, Azure and GCP. Like MongoDB, Datastax is extending into adjacent use cases like search, analytics and a graph database. They have assembled a broad set of customers. There has been speculation around an IPO since early 2019. However, earlier this year, there were reports of layoffs and a management shake-up.

Cassandra is a wide-column store database. This is a flip on the traditional relational row-based data store. In a column stores, the data is serialized into columns, rather than rows. So, rather than having a database row of {customer ID, first name, last name}, a column store organizes it along each column with the associated primary key {ID:first name}, {ID:first name}, {ID:first name}. This optimizes the data for look-up at the column level, which really speeds up retrieval for certain types of data workloads, like analytics or summary meta data for an application (get count of all customers from New York).

This represents a different workload than a document store, like MongoDB, so likely doesn’t represent a conflict. However, Datastax’s addressable solution set does have overlap with MongoDB, so we will need to monitor this player.

Partnerships – SI’s

Beyond direct sales, another contributor to growth for most software stack companies is from system integrators. The SI’s are often hired by enterprises to facilitate their digital transformation projects. For popular technologies, the SI’s will even form a practice around a technology solution, to demonstrate their expertise and concentrate their focus. When hired by enterprises, the SI’s will have influence over technology choices and are responsible for implementation. Having mindshare from SI’s for a software stack company can contribute a significant amount of new customer wins, without requiring substantial direct sales effort.

MongoDB has not traditionally seen a meaningful contribution to revenue from SI’s, but leadership feels the interest is growing. On the earnings call, the CEO mentioned that interest level is very high from SI’s. He described a call with a senior exec from one of the large SI’s in which that individual asked how their organization could increase their knowledge of MongoDB quickly, because they are experiencing high demand from their customers. This is being driven by retooling and re-architecture initiatives, probably coming out of digital transformation acceleration. The SI’s want to be able to bring modern solutions, like MongoDB, to their customers when presenting options for migrations to microservices and distributed system architecture.

MongoDB recently added Frank D’Souza, the former founder and CEO of Cognizant ($16B in revenue in 2018), to the board. He is a domain expert in SI operations and has a lot of contacts in the arena. The MongoDB CEO discussed engaging Frank to improve MongoDB’s position with both the global SI’s and even regional players. This should further contribute to MongoDB’s growth, as it introduces a new channel for meaningful customer additions beyond direct sales.

My Take-aways

• If we look at Q4 and Q1 performance, revenue growth is consistently landing in the high 40% range, with some acceleration coming into Q1. Putting aside impact of COVID-19, I think this rate is sustainable. For the last couple of years, MongoDB revenue growth has decelerated, causing many investors to worry where it will land. At this point, I think it will stabilize. The factors that would affect growth going forward (market size, competition, product expansion) are at a steady state, which means that I don’t see a reason that these high revenue growth rates (in the 40% range) are not sustainable.
• Q1 profitability improvements were significant. In Q1, Non-GAAP EPS loss reduced by 50% year/year. Operating margin improved from -14% to -6%. Again, putting aside COVID-19 impact, these trends would likely have continued. MongoDB invested significantly in building out the R&D and sales organization over the last year, but should start to realize leverage.
• The growth of Atlas is impressive. Atlas now contributes 42% of total revenue and is growing at 75% year/year. This growth rate is higher if you back out the headwinds from the mLab transition. Atlas growth rate was 80% in Q4 and leadership talked about some spend growth slowdown in Q1 by existing customers, due to COVID-19, so it’s a fair bet that under normal circumstances, they would have repeated about 80% growth. And these are no longer small numbers. As Atlas revenue becomes a larger percent of total revenue, we cloud see total revenue growth rates accelerate, or at least easily sustain at these higher levels.
• The product extensions are thoughtful. I like the strategy of MongoDB to move beyond the core database into offering a platform of data services that make application solution building easier for developers. The capabilities being built through the Realm offering are significant, as mobile app development is still a large market and a complex affair. Simplifying data management and solving the offline sync problem are very appealing. The other add-ons for search, data lake and visualization are helpful too and likely just the beginning.
• MongoDB is quickly addressing many of the technical gripes around the core database and is maturing as a versatile, reliable, highly scalable solution. ACID transaction safety was addressed over a year ago. Refinable shard keys clear a common hesitation around applying MongoDB to large data sets due to the risks around sharding. As each of these limitations is ticked off, MongoDB becomes more acceptable as an enterprise solution for mainstream workloads. Customer wins with high scale, technology first companies, have further solidified this perception.

Risks and Items to Watch

• Q2 and full year revenue projections returned to deceleration in year/year growth rates. As stated above, I am encouraged by sustained revenue growth in the 40% range for two quarters in a row. However, Q2 guidance is for growth rates in the 20% range. While I expect MongoDB to exceed this, even a beat of the magnitude of Q1 would put Q2 revenue growth around 38% year/year. Further, the improvements in profitability from Q1 regress through the remainder of the year. This is likely attributable to the revenue slowdown with the same fixed costs (mostly salaries). With revenue beats, this will improve, but likely not to the same levels as Q1, at least for the remainder of this year. While I think MongoDB can return to high growth levels next calendar year, this represents a risk to the valuation in the near term.
• While the competitive dynamic appears the same as it has been for the past 12 months, this isn’t guaranteed to continue into the future. While I don’t think a cloud provider will put forth a meaningful new competitive offering, another independent could emerge. I will monitor progress from Snowflake, Datastax and even Confluent for developments in their offerings.

Investment Plan

MongoDB provides a popular data management platform that is experiencing rapid adoption. The addressable market for their solutions is huge and primarily occupied by legacy providers. As monolithic software applications are updated and re-architected, engineering organizations have the opportunity to introduce new databases into the mix. In parallel, digital transformation initiatives often involve building new, mobile-first applications that require a flexible, developer-friendly data store. These tailwinds should drive sustained growth for MongoDB for years to come.

Q1 results were strong, with sequential revenue growth acceleration from Q4, and marked improvement in profitability. Particularly encouraging is the growth of MongoDB Atlas, the cloud-based solution, which now makes up 42% of revenue and is growing at over 75%. These trends bode well for sustained revenue growth in the 40% range. If it weren’t for the COVID-19 situation, MongoDB would be performing well on track. However, Q2 and full year guidance include an expected slowdown in customer spend, as a result of the macro environment. These will impact growth rates and reverse the profitability measure improvements.

As we look towards next calendar year, assuming the macro environment improves, MongoDB will be well positioned. The pull forward of digital transformation initiatives and general migration of business to online experiences will benefit MongoDB. While they won’t win every deal for a database solution, they will be involved in enough workloads to drive meaningful growth. Additionally, MongoDB’s product roadmap continues to expand beyond the core database, into a suite of data platform services that address ancillary functions and further solidify the case for customer adoption.

I initiated coverage of MDB in November 2019 with a five year price target of $490. At the time, the stock was trading around $140 and is currently at $224. I think my price target is easily achievable, likely before 2024. For now, I will keep this price target in place and plan to revisit it at the end of 2020. This will allow visibility into full year FY2021 performance and plans for next calendar year. In the meantime, I think MDB is valued fairly relative to the opportunity. Investors with a long term perspective would benefit from ownership of MongoDB. Personally, I have 6% of my portfolio allocated to MDB and will likely increase that as I see further evidence of the story materializing.