Provided here is our redacted investment memorandum, detailing our rationale for investing in Cube’s Seed in 2020. We reinvested in their Series B in 2024.
COMPANY OVERVIEW
Cube Dev offers an end to end platform for analytical applications. Their flagship product, Cube.js, is an open source project that helps developers quickly create analytics experiences within their applications; this could be internal dashboards, as well as customer facing ones. On top of that, Cube.js also has a “Playground” mode which allows non-technical users to build their own reporting.
The real value that Cube offers comes from their middleware that sits between data sources like Snowflake, PostgreSQL, Redshift, and the frontend visualizations for them, usually implemented in React, VueJS, etc. This allows developers to build fast and high performing data-driven applications without needing to worry about all the data piping. Cube offers a “Cube Cloud” product which is a hosted version of this, further lowering the work that a developer needs to do to start using the product.
KEY HIGHLIGHTS
- Cube.js is one of the fastest growing open source out there. The company has also attracted dozens of contributors.
- The company has launched a Cube Cloud hosted offering and onboarded strong design partners.
TEAM OVERVIEW
The company was founded by Artyom Keydunov (CEO) and Pavel Tiunov (CTO), as a spin out of their previous venture, Statsbot. Artyom was previously a senior Rails engineer at Happy Numbers (top 200k Alexa). Pavel was a freelancer through TopTal, one of the best freelancing networks, and then co-founded Allcountjs, a node.js development framework. He holds a PhD in computer science and a master in physics from Bauman Moscow State Technical University.
Both of them are very strong technical founders. Since starting the company, they have been able to surround themselves with talent that focuses on the business/growth side. One example is Ray Paik, who spent more than 2 years as a community manager at GitLab ($6B company), after spending time at Medallia as a Client Services Manager and almost 11 years at Intel in various roles in the developer relations and marketing team.
INDUSTRY OVERVIEW
Alessio wrote a blog post on the history of business intelligence, which is available on our website, that goes in depth on how companies in this category have evolved. Nowadays, there’s two main ways that people solve this problem:
- For business intelligence, companies might use Tableau or Looker. While powerful, these tools are very hard to integrate within users’ workflows as they are not embeddable. They are also somewhat of a black box, and require users to have tool-specific knowledge on top of SQL knowledge in order to operate them. Those tools also don’t support customer-facing experiences, restricting them to internal use only.
- For more flexible use cases, developers within organizations might build tailored experiences using open source frameworks such as Chart.js, D3, etc. This is the case here at 645 Ventures. The great thing about this is that the visualizations can be embedded right in the user workflow, rather than redirecting them to an external application. For customer-facing experiences this is the easiest way, as it allows to fully control the design in order to match your brand. As the application grows, implementations start to hit multiple limitations:
- Querying performance: as the amount of data you are analyzing grows, it becomes harder and harder to deliver results in a timely manner. This usually leads to a degradation of user experience, which negatively hurts your business. A frontend/full stack engineer might not be proficient enough in SQL to rewrite their ORM queries in a more efficient manner.
- Caching complexity: some of these queries start to get pretty expensive and the data isn’t very dynamic, so you resort to caching. As the application grows, managing cache expiration becomes challenging.
- Multi-data store analysis: when running queries, you might need to join data from multiple databases or data warehouses. This isn’t trivial to do in a performant way, and it might not be supported at all by the ORM you are using. This limitation also applies to traditional BI tools. For example, Tableau didn’t have support for hosted MySQL and PostgreSQL databases until 2015.
THE NEW AGE
At this point in the business intelligence market life cycle, we don’t believe there is an opportunity for new billion dollar companies to be created just by offering data analysis and visualization tools. This is a space that has been commoditized by the vast amount of options available to the customer. The biggest opportunity we now see in the market is business intelligence platforms that allow companies to power both internal analysis and embedded intelligence. Take Stripe for example; their core offering is a payment API, which is what the developer is exposed to. On the other hand, Stripe’s web application is basically a business intelligence dashboard.
Salesforce is another good example; a CRM tool is great for managing your customers, but what’s even more important is the data underlying the CRM: for example, the average close rate of your sales team, your company’s average contract value, or your pipeline for the quarter. Salesforce allows you to see this with their “Dashboard” and “Reports” feature.
If you think about all the products that you use on a daily basis as a knowledge worker, you will notice that charts and graphs are a core part of a lot of them. This is due to the fact that the insights gathered from the data your usage generates are just as important as the product itself. Offering analytics and reporting with software products is becoming table stakes.
We believe a product like Cube can be widely adopted by developers working at startups as well as large enterprises. In Stripe’s “Developer Coefficient” report, executives listed four major pain points that developers can address: “Bringing products to market faster” and “Internal reporting / visibility” were two of them. Another large trend is the increase of custom apps and tools being built inside organizations, as displayed in the chart below. The adoption of Cube can help companies cut down on development time by abstracting complexity away, and then use that same infrastructure to also spin up internal reporting. We believe developer productivity will be a big focus in the next decade (and why we are so excited about also being investors in DeepSource) and it will generate multiple billion dollar companies.
While the core product is developer focused, Cube also has a “Playground” where non-technical users can experiment with creating queries and building dashboards without the need for engineering support by leveraging Templates. We believe that in the long run, Cube has the potential to displace existing BI tools by using their developer experience as a wedge, a product offering incumbents can’t compete with. This further increases the opportunity for Cube to build an iconic company.
COMPETITIVE LANDSCAPE
As a preface, none of these companies are direct competitors to Cube.js. There’s no other product in the market today that focuses on empowering developers to build embedded intelligence experiences. Most of the companies listed below are traditional business intelligence tools, some of which are open source. The ones in the latter categories are the closest to directly competing with Cube, as their product could be more easily re-architected to support this use case.
Preset
Total Fundraising
- $12.5M from Andreessen Horowitz
Product
- Preset is the commercial version of Apache Superset, one of the most popular, open source business intelligence tools with over 31,000 stars on GitHub. The company has yet to launch a product available to the general public.
Founding Team
- Max Beauchemin has worked at the leading edge of data and analytics his entire career at data-dependent companies like Yahoo!, Facebook, Airbnb, and Lyft. A leader in the open- source community, Max created both Apache Airflow and Apache Superset.
Metabase
Total Fundraising
- $21M from NEA over two rounds
Product
- Metabase is an older company (started 2014), using open source as a distribution model. They are also a business intelligence product. Their open source repo has more than 2000 open issues on GitHub, which isn't a good sign. They are mostly focused on analysts/non technical users.
Founding Team
- Sameer Al-Sakran served as the CTO of Expa and Blackjet. He also spent time at General Catalyst as a data scientist.
Toucan Toco
Total Fundraising
- $14M from Balderton Capital
Product
- Toucan Toco offers a more full-fledged product, which supports both embedded analytics and traditional business intelligence reporting. The company is focused on “Data Stories for non-technicals”, and their open source projects have no adoption. Their main focus is on enterprise clients.
Founding Team
- The company is based in France; the founders have studied and taught at top technical schools in France, but don’t have any outstanding business experience in the US market.
Below you can see how we think about this overall market in terms of current options, as well as historical companies that got to large scale.
MARKET SIZING
The closest proxy for Cube’s market sizing would be looking at Tableau’s market share. Any company sophisticated enough to use Tableau for reporting also likely has a need for a product like Cube to power their internal products. On top of that, there’s an array of companies already rolling out custom business intelligence tools rather than using an off the shelf one, who could also be customers. A product like Cube will actually expand the latter market, as it makes it easier to build reporting capabilities instead of buying a tool.
In 2018, Tableau had 86,000 customers and generated ~$1.2B of revenue, for an ACV of $13,488. It is estimated that Tableau has penetrated 13% of their target market, which would equal to a TAM of $8.8B. Gartner sizes the business intelligence and analytics market at $24.8B, so our estimate is much more conservative.
COMMUNITY & MARKETING
Cube has one of the most active open source communities out there. The open source repo has attracted more than 9,300 stars, as well as dozens of contributors. Their Slack is very active with more than 2,300 members.