How does Databricks compare to Snowflake?

Today, Snowflake is a “data cloud,” with support for a wide variety of data science and other functions, while Databricks has its own data warehouse product competitive with Snowflake  called Delta Lake.

Snowflake was one of the biggest successes and most richly valued companies of the 2020-21 IPO window. Their IPO raised $3.36B, valuing the company at $30B+, a high premium on their last private fundraising valuation of $12B, and a nearly 150x price-to-sales valuation.

Databricks, on the other hand, has been one of the highest-valued private companies in the world since they raised $1.6B from Counterpoint Global in the August 2021 Series H that set their valuation at $38B.

Snowflake, like virtually every software company that went public in 2020-21, saw its valuation drop precipitously through 2022. By the end of the year, Snowflake’s stock price was down more than 57%. 

Over the same timeframe, private company valuations dropped as well. 

We saw Databricks itself lower its internal implied valuation to $31B in October 2022, while Fidelity marked its stake down to $24B. 

Snowflake is currently trading at 14.9x their expected sales for the next twelve months, which analysts put at $2.96B.

As of this June, we estimate Databricks’s forward revenue at $1.55B. Databricks is smaller, but growing faster—Snowflake recently revised its growth forecast for the next fiscal year down from 45% year-over-year growth to 34%, while we estimate that Databricks is growing around 50% year-over-year.

If Databricks were to trade at Snowflake’s 14.9x NTM multiple, their enterprise value would be roughly $23B. That’s close to Databricks’s median secondary price on Caplight of $48.06 per share, which implies a valuation of $24.9B.

Public companies indexed on AI in one way or another have seen the best results in the stock market in 2023—like NVIDIA (NASDAQ: NVDA) which is up 172% year-to-date, or Crowdstrike (NASDAQ: CRWD)  which is up 49%, or Microsoft (NASDAQ: MSFT) which is up 39%.

While Snowflake is incorporating machine learning and AI features into their products, and in May acquired AI search company Neeva, they have not been as front-and-center in the explosion of artificial intelligence that started last year as Databricks have been.

Despite Snowflake’s changes over the last few years, Databricks’s infrastructure is still considered to be more suitable for machine learning and AI than Snowflake’s.

AT&T used Databricks to train their own AI models for stopping fraud, while Shell uses it to run simulations to determine how many spare parts they should store in inventory across thousands of warehouses.

In March, Databricks open sourced an AI model—Dolly—that it said was comparable to ChatGPT and designed to be used by companies internally.

---

In 2020-1, Databricks, like their closest comp in the market Snowflake, traded at a high valuation driven less by the intrinsic value of their earnings and more by the speculative option value of their potential.

Today, while a fundamental valuation of Databricks with respect to Snowflake leads to a valuation of about $23B vs. their last primary round valuation of $38B, Databricks may be the company that is better positioned to enjoy the “AI premium” of a $NVDA or $MSFT when it does enter the public markets.