"It was the best of times, it was the worst of times, it was the age of wisdom, it was the age of foolishness” - Charles Dickens

It was an interesting experience attending the Snowflake and Databricks Summits back-to-back, as both of these giants are going toe-to-toe in the ring, or so people would have you think. Standing in Moscone Centre, you learn quickly how segmented these platforms are. Snowflake is the de facto warehousing tool for BI and reporting, while Databricks is a specialized platform built for serious data engineers focused on AI models and more technical workloads. Yet the internet seems hellbent on making these two stand off against one another.

For the purpose of this article, I will compare the two as they both chase the same target—owning a piece of the storage layer. However, as you dive deep, you learn how differentiated their methods and products are.

A few weeks back, the headline was Databricks acquiring Tabular, the company built by the creators of Apache Iceberg, for ~$1B. That being said, the rumour at Snowflake Summit was that Databricks paid closer to $2B and Snowflake themselves fell short of bidding on the company at $1.5B. Who really knows?

Before we get into it, let’s quickly review Tabular and Apache Iceberg. Apache Iceberg is an open-source table format that allows teams to work simultaneously with the same data sets from different compute engines. Tabular, founded by the creators of Iceberg, is a hosted (paid) version of the open-source platform designed to help companies store and manage their data more efficiently.

Ok, back to the story…Although these companies address different problems and have distinct approaches, this acquisition was a big move for Databricks, and Snowflake should be kicking themselves for failing to close the deal.

On Monday, the day before Databricks announced their acquisition, the headline was about Snowflake’s new tool Polaris Catalog, an open-source catalog built directly for Apache Iceberg. This purchase came as a bit of a surprise for a few reasons:

1. Leaning into Open Source: Historically, Snowflake has been a closed-source platform that tends not to play nice with open-source tech. They originally built their product as a closed platform that prided itself on speed and cost.
2. Acknowledging their weakness: For data nerds like me, listening to Snowflake's recent quarterly earning calls feels like Groundhog Day. They all share a similar tune. Snowflake’s CFO said this in the last call, “We have invested in Iceberg because we expect it to increase our future revenue opportunity. However, for the purpose of guidance, we continue to model revenue headwinds associated with the movement of data out of Snowflake and into Iceberg storage.” Snowflake storage accounts for roughly 10% of revenue. Watching customers churn from this and go to an open standard can’t be easy for Snowflake to digest.