DataStax Raised $190M, Then Stopped. Billy Bosworth Calls That the Strategy
Billy Bosworth joined DataStax as CEO in 2011 when the founder handed him the title. Seven years and $190M later, his answer to every funding question is that he doesn't need to.
On this page
Billy Bosworth was not a founder of DataStax. He was the partner who kept showing up, and in May 2011 the founder who was running the company handed him the CEO title.
For a founder to give up the CEO title is a nontrivial — that is a very, very mature move.
Billy Bosworth, CEO, DataStax
Seven years later the company is past $100 million in ARR with gross margins over 75%, and Bosworth’s answer to every question about raising more money is the same: he does not need to.
The thesis. DataStax raised $190 million across five rounds and then stopped, and Bosworth treats that as the strategic asset rather than the balance-sheet fact. Control of the cash is what lets him refuse to answer the IPO question, resist pressure from a decade-old cap table, and stay ruthlessly narrow about which customers he wants.
How the handover happened
Bosworth ran the database business at Quest Software and could see the relational model flattening while a new class of open-source databases emerged. He partnered with DataStax when it had just formed — Quest built tools for ecosystem technologies, and Apache Cassandra was one worth watching. He was in Dallas, they were in Austin, and the drive was easy.
A technical founder and semi-technical founder who was doing the CEO work — and it was Matt who came and said, you know, I think I could learn all this stuff, but I’m not sure I should learn it as fast as I could, which would be best for the company. And so I think you would be a great fit if you were interested.
Asked whether he was being seduced or doing the seducing, Bosworth calls it a mutual love affair. His first assignment as CEO was raising the Series B in 2011, which he describes as tough — still coming out of the nuclear winter, not a great environment. The rounds after that got easier.
The product and the buyer
DataStax sells DataStax Enterprise on an annual subscription. Customers consume it as a managed service, run it themselves in the cloud, or run it on-premises. At the time of the interview 60–70% of workloads were already in the cloud one way or the other.
The entry point into an account is the part most people get wrong.
Many of our entry points inside of organisations do not start with the centralised technology teams. They actually start with a line of business. The lines of business are the ones who are tasked with moving very fast to either gain market share or stop the erosion of market share.
After two or three projects, DataStax works its way into the central team. The persona is a data architect — the person responsible for fitting together a whole data strategy — and the funnel into that persona runs through open source and free training rather than outbound.
The payback question, answered differently
Latka asks for cost to acquire a customer. Bosworth declines, and the reason is worth reading if you sell alongside an open-source project.
The whole point of open source is a lot of times it’s all about come and get things without having a direct connection. You don’t want to be pestered, Nathan. You want to be an open-source person, you want to be anonymous, you want to go grab the open-source code — and then you find us through many things that we do, like DataStax Academy, which provides free world-class training.
What he tracks instead is how fast the first project succeeds, because that is the gate on everything after it.
Once that first project is successful, the expansions happen much, much faster. The inverse of that is true as well. If we don’t get that first project right, it can really slow the progress on expansion, because a lot of these technologies are still new. There’s a lot of eyeballs in the customer company waiting to look at that first project to say, should we double down or should we stay away a little longer.
The target is net expansion above 110%, which includes churn, and he says enterprise accounts run well beyond it.
Two kinds of project, two expansion curves
Customer 360, recommendation engines, fraud detection. Starts granular; expansion can take a couple of years to get the first project solid and move on.
Supply chain, inventory management, asset management, security access. Often a big project up front with rapid expansion after.
This is why Bosworth resists giving a single ACV growth number. The blend hides two different businesses, and which one an account is depends entirely on the project it starts with.
Why professional services is not a hedge
He is direct that DataStax has a services component and unapologetic about it, and the reasoning comes from having lived through the last platform transition.
I lived through the mainframe to client/server transition. When I came out of school, Oracle was still considered largely a potential toy off to the side. What people forget is how long that transition took in the market. What you require is some initial hand-holding with your customers, because they’re still thinking on old relational paradigms.
Even with the vendor managing operations, someone has to teach the customer how distributed databases change the way they think. That is a people problem, not a software one.
The cap table nobody is agitating on
A company seven years into a decade-old Series A, with $190 million raised and no IPO date, is exactly where founders expect investor pressure. Bosworth says it has not arrived, and gives an unusually honest reason.
That anxiety you mentioned is also directly tied to the success of the funds that you’re already in. Our early investors, their funds are performing extraordinarily well, and so we don’t have that pressure from them.
He is more careful about employees, who took equity instead of salary and are entitled to wonder when it becomes worth something. His answer is communication rather than a date.
You have to really be maniacal about communicating the vision and communicating the company’s progress. If you don’t do that, then you do have employees who are very nervous about the future — is this, did I make the right bet, am I in the right company?
Almost all the capital went to operations; secondaries were “very, very minor rounding error” territory early on. The GetLatka profile carries the $100M figure for the date of this conversation, and Bosworth confirms the band is $100–$200 million rather than above it.
The steel town argument
Bosworth keeps returning to a distinction between how he manages and how he thinks a Silicon Valley CEO would, and he ties it explicitly to where he grew up.
I’m thrilled with the bull market we’ve been experiencing. I’m thrilled with the access to capital we’ve had — we’ve leveraged it, we’ve taken advantage of it. But I’m not going to get silly with it. I’m not going to treat it like it’s free water, because it’s not.
Markets turn, he says, and when they do everyone will suddenly want to know whether you control your cash. DataStax has made two acquisitions — a graph company and a cloud managed-service company — which he frames as building the muscle early rather than buying revenue. On selling the whole thing, he offers the cliché and means it: great companies are never sold, they are bought.
Asked what he wishes his twenty-year-old self had known, he does not soften it.
Don’t be so arrogant and obnoxious, and go learn from your betters.
Sources Billy Bosworth’s interview with Nathan Latka, recorded 16 January 2018; revenue, headcount and funding rows from the GetLatka DataStax profile.
