At first glance, Couchbase appears to be stuck in the middle of the cloud database market, flanked by competitors with more traction and buzz. But fresh off a $200 million IPO Thursday, CEO Matt Cain relished the opportunity ahead to prove why his company can beat out rivals the market considers more valuable.
The NoSQL database provider's public offering helped propel Couchbase to a $1.2 billion valuation. But unlike one of the last big data-related IPOs, market leader Snowflake's historic debut on the public markets last December, Couchbase has some work to do to differentiate itself.
In terms of rivals, on one side it has open-source tool PostgreSQL and companies like MongoDB. MongoDB Atlas, its now-signature product, pulled in $93.5 million in revenue in the three months through April, a 73% year-over-year increase. Overall, the company reported subscription revenue of $174 million for the quarter, almost double the $96 million in subscription sales that Couchbase reported for all of FY 2021.
On the other side, there's a cadre of other startups like Redis Labs, which private investors value higher than Couchbase at $2 billion with revenue roughly in line with Couchbase's total. And then, of course, there's Oracle, as well as the cloud hyperscalers that are perennial rivals to almost every enterprise tech vendor selling today: Google, AWS and Microsoft.
None of that bothers Cain, who claims that Couchbase's competitors are unable to address the needs of the whole enterprise — both application developers and enterprise architects. It's that differentiation he believes will ultimately help the company break free from the middle of the pack.
"We are focused on building the next great enterprise software company and much more concerned about the technology we offer … than short-term valuations," Cain told Protocol in an interview Thursday. "We are much more focused on the relative value proposition that we offer. We will monetize that over time, but we have long-term views, long-term aspirations."
In reality, the area Couchbase is trying to tackle is not much different than its competitors'. Enterprises are building more applications that require data to move at speeds that can't be met by the architectures of the past. It's a trend that is spurring what Couchbase calls a "generational transition" and underscoring what some industry insiders say is the "golden age" of databases.
"Efforts to accommodate the limitations of legacy databases through ad-hoc temporary fixes and stopgap methods are no longer sufficient given the increasing impact and urgency of digital transformation initiatives," the company wrote in its S-1 filing. "A move to the next generation of modern operational databases is required to provide the performance, reliability, scalability and agility needed for enterprise applications."
Couchbase celebrates its IPOPhoto: Nasdaq
Separating from the pack: Version 2.0
Couchbase is one of the lone independent holdouts from the NoSQL — which stands for "not only SQL" — movement that started around the late 2000s.
The tech is not based on rows and columns like traditional relational databases. Instead, all the data is stored in one entity, like a JSON document, for example. That non-tabular structure gives it the processing speed that NoSQL advocates say is necessary for modern applications.
Many of Couchbase's former rivals, like FoundationDB or Tokutek, were acquired several years ago. The consolidation within the sector was so swift that former CouchBase CEO Bob Wiederhold wrote a blog in 2015 assuring customers that it was "not bad for an industry."
"During the next stage, the leaders separate themselves from the rest of the pack. Their products are a better fit for the market and more users gravitate to them," he said at the time. "Couchbase, DataStax and MongoDB are the three NoSQL players that have clearly separated themselves from the rest of the pack."
Now, it's on Cain to separate Couchbase from the pack yet again. The company launched its managed service Couchbase Cloud in 2020, well behind MongoDB, which released Atlas in 2016. It's currently offered on AWS and Azure; Google Cloud is on the product roadmap, per Cain.
"There's a lot of attention in the market on the way in which companies are offering their customers the way to consume their technology and we couldn't be more excited about our managed cloud roadmap," he said. "We're going to study the success and failures of other companies … but we will do so through the lens of supporting our enterprise customers."
The differences that Cain claims exist between its own product and rival offerings are deeply technical. At a high level, Cain says Couchbase is superior because of its speed and scalability — a common talking point as well among rival database executives.
Among the other differentiators, according to Cain, are its hybrid cloud compatibility, ability to support multiple data models — something competitors like Redis Lab also say they offer — and "high-performance architecture that is memory-first, shared-nothing," an over 30-year-old setup that allows the database to scale faster while also reducing (or eliminating) downtimes. If you need to build a website like Amazon, for example, that has to handle millions of queries, you would build it on a shared-nothing system.
"If you look at legacy relational technologies, modern databases, there is no solution that brings all those capabilities together in a single platform," said Cain. "The modalities that a database chooses to support are important, but I would not underestimate the underlying architecture."
Database companies are in a fierce marketing battle at the moment as they all seek to capitalize on the very real needs of enterprise customers for help managing increasing demands on data. As companies begin to better understand what their modern architectures will look like to support a quickly evolving application suite, those claims are going to be put to the test.
"In terms of how often enterprises are thinking about this, it's a daily exercise. And that is driven by all things digital transformation," he said. "Good enough is not a viable strategy when it comes to databases in the enterprise. With our core architecture, we can find compelling opportunities to add value."
