It's been a tough week for investors trying to get their hands on Databricks' record $10 billion funding, one of the VCs who led the deal told TechCrunch.
George Mathews, managing director of Insight Partners, said with a laugh: “I've been getting calls late at night, but that's okay. That creates good opportunities.” Insight was one of six companies that led the deal, along with new investor Joshua Kushner's company Thrive. All but Thrive were existing investors.
“We worked hard to be co-leads even though we were already investors on the cap table,” Matthew said. Insight first invested in Databricks in 2021. But to participate in this huge deal, the company needed to use the Insight Partners public equity fund, which was created under managing director John Wolfe to buy public stocks.
There was so much interest that allocations and valuations rose rapidly. As of mid-November, the deal was expected to be worth about $8 billion, Reuters reported at the time. A few days later, it was valued at $9.5 billion with a valuation of $60 billion, and by Tuesday it closed at $10 billion with a valuation of $62 billion.
For perspective, this is bigger than the $6.6 billion OpenAI raised in October, its largest venture round in history.
“There has been tremendous institutional demand and interest in intergenerational companies,” Matthew said. “I've been investing in Insight for the past four years on everything related to data, AI, and ML. This is what I live for.”
The investment included a large secondary tender offer in which Databricks employees and other existing investors could sell their shares. New preferred shares have been issued to new investors. Databricks did not disclose what is secondary to the amount raised, other than to call the $10 billion “non-dilutive,” which means a significant portion.
Interestingly, Databricks, founded in 2013, could have been a tragic story. Ten years ago, its founders created Spark, a technology that was key to the “big data” trend of yesteryear. Spark has helped companies analyze their internal big data at lightning speed.
The rise of cloud-hosted data has led the company to process data before handing it over to other players. Gradually, it may have been relegated to unrelated big data functions.
Databricks co-founder and CEO Ali Ghodsi (pictured) sought advice from Mathew, who ran big data company Alteryx as COO before becoming a VC. The two have been friends since the early days of Databricks.
“A few years ago, Ari called me and said, “I'm thinking of getting into the data warehousing market.'' And I just said, “That's the best thing I've ever heard. 'And I couldn't have been more wrong,' Matthew laughed, noting that Godi didn't listen to him or force his bad advice on him. He added that he was happy about that.
At the time, traditional data warehouse vendors that housed vast amounts of corporate data used for analysis were also struggling against emerging cloud stars such as Snowflake and cloud vendor-owned products such as AWS's Redshift.
But in late 2020, Databricks introduced Databricks SQL, a data warehousing product anyway, and quickly became a big competitor to Snowflake.
Then along came large-scale language models (LLMs), which are always hungry for high-quality enterprise data. “Where does this high-quality data come from? For businesses, it's going to come from places like Databricks,” Matthew said.
Fast forward to the end of 2024, and the IPO market remains locked, with investors looking to acquire AI infrastructure products such as data warehouses that can service LLMs.
Databricks says that by the end of its fiscal fourth quarter, Databricks SQL's revenue run rate will reach $3 billion, an annual increase of 150% to $600 million.