Digging into the Databricks "bargain"
Add Axios as your preferred source to
see more of our stories on Google.
Illustration: Allie Carl/Axios
"We got a bargain" is not usually what you hear from a venture capitalist who just committed to a $10 billion round at a $62 billion valuation.
- But it's what one told me yesterday as news broke of the big Databricks round. And I heard similar sentiments throughout the day, leading up to my Axios AI+ Summit interview with company co-founder and CEO Ali Ghodsi.
The bull case is logical.
- Databricks is growing more than 60% year-over-year, topping $3 billion in revenue run rate with expectations of being cash-flow positive this quarter.
- Rival Snowflake has more top-line revenue, but the gap keeps shrinking and its growth is significantly slower (some believe Snowflake missed the first AI boat and is trying to catch up in a canoe).
- In short, this was a deal that many investors felt they could not miss out on, which is why there are now around two dozen new institutional investors on Databricks' cap table — some which never even met with Ghodsi — with some participants writing the largest single checks in their firms' histories.
But there are three issues with the "bargain."
1. Everyone could be wrong about the company's future. This is the same risk as in any VC round, although this quantum of capital inflates the consequences. Moreover, Ghodsi himself said on stage that we're in a "peak AI bubble," which suggests that this round could look expensive in hindsight, even if Databricks maintains its growth trajectory.
2. If the VCs are correct, it means that employees are selling low. Almost all of the $10 billion is going toward a nondilutive tender. Of course Databricks employees will take some money off the table — it's the only responsible move for those without existing wealth — but they could have sold much higher had Databricks gone public at a market price (possibly $100 billion or more).
3. Certain LPs will look askance at what was essentially a party round, given that different partygoers charge higher fees than others. A bit reminiscent of LBO club deals before the Great Financial Crisis.
The bottom line: Ghodsi said that Databricks will eventually go public, possibly as early as next year, but gave no definitive calendar. Whenever that time does come, we'll all find out if this week's VC math added up.
Watch the full interview with Ali Ghodsi, including discussion of the financing, how his refugee journey impacted company culture, and what he wants David Sacks to do as "AI czar.".
