Why has Snowflake or Databricks not acquired DBT?

Disclaimer: This article is speculative and based on personal opinions. As of the time of writing, there is no indication or news confirming that Snowflake or Databricks plans to acquire DBT Labs.

In this post, I’m trying to understand why given DBT’s and Databricks/Snowflake's shared customer base and the jigsaw-like fit of their capabilities an acquisition has not happened.

DALL*E interpretation of “Marie Curie writing a DBT model on a computer”

DBT’s Pricing Changes: A New Paradigm

DBT Labs recently moved to a consumption-based pricing model, a departure from its seat-based pricing. This shift makes sense for several reasons:

1. Sustainability: We should dispense with the idea that open-source software is some altruistic endeavor that exists outside the realm of economics. Companies need to keep the lights on. And while DBT Labs has garnered a cult-like following for its open-source contributions, goodwill and impractical pricing models don’t make payroll.
2. Value Realization: As DBT becomes the go-to platform for a growing list of data operations — from transforming data to validating it — the disconnect between the value provided and the revenue generated has widened. This change is an attempt to correct that balance, and it’s hard to argue against a company seeking fair compensation for the utility it delivers.

Snowflake and Databricks: The Elephant (or Elephants) in the Room

It’s tough to discuss DBT without mentioning Snowflake or Databricks. The latter have become more than just tools; they’re platforms that underpin entire analytics ecosystems. So how does DBT’s pricing change impact these giants?

It’s a simple but potent change. Raise the cost per query in DBT, and you introduce a pause, a moment of hesitation before users hit ‘Run.’ While this newfound frugality may seem inconsequential, it’s not. Each query not run is a compute cycle not spent, a piece of storage not used. It’s an invisible shrinking of the boundary conditions within which Snowflake and Databricks operate.

If you’re wondering whether this impacts the financial health of Snowflake and Databricks, the answer is an unambiguous yes. These platforms derive value not just from hosting your data but from the myriad ways you interact with it. Fewer queries and reduced storage are not simply less activity; they translate into less revenue. And revenue, as they say, is the lifeblood of any business — be it a start-up or a juggernaut.

The Upside: When DBT Meets Snowflake and Databricks, Synergy isn’t Just a Buzzword

But let’s not paint a one-sided picture here. If DBT’s pricing changes introduce constraints, the tool’s seamless integration with Snowflake and Databricks also uncorks a lot of value. So, what does this synergy — yes, an overused term but genuinely applicable here — really offer?

The Transformation Engine: It’s Mostly DBT

In many organizations, DBT is the go-to workhorse for data transformation. Its unique capabilities are not just an added value; they are often a compelling reason for the adoption of underlying platforms like Snowflake and Databricks.

Data Validation: DBT Adds Depth

Then comes the notion of data quality, an area where DBT excels with its SQL-based checks. With DBT in the mix, you’re not just storing data in Snowflake or processing it with Databricks — you’re also validating its quality. This added layer of interaction increases the platforms’ stickiness and, frankly, makes DBT more indispensable to daily operations.

Metrics and More: Third-Party Leverage

Don’t underestimate the power of third-party tools like Hex or Mode, which leverage DBT’s semantic layer to offer enriched analytics. While they may seem like peripheral players, they often encourage a higher volume of queries. This has a cascade effect: the more queries run, the more Snowflake and Databricks come into play, bolstering their usage and, ultimately, their revenue streams.

When you factor in the three reasons above, it feels like a complete integration would be a step that would benefit either of the big players.

Why No Acquisition?

So, why haven’t Snowflake or Databricks made the move to acquire DBT Labs? So far I can only think of these two reasons:

1. Valuation: DBT Labs has grown in prominence and value, making it a potentially expensive acquisition target. In 2022 it was valued at over $6B. Although the private valuations compressed since then by a lot, it would still be an expensive acquisition.
2. Independent Path: DBT Labs isn’t just sitting by the phone waiting for an offer. They’ve built a moat — a ubiquity in the data world — that suggests they might have grander plans of their own. The idea of charting an independent course and embracing the challenge of monetization could be more enticing than becoming a subsidiary.

Despite these hurdles, the complementary nature of these technologies makes the acquisition question incredibly intriguing. The recent pricing changes by DBT Labs make this all the more relevant, potentially affecting both its and its would-be acquirers’ revenue streams. Whether or not an acquisition is on the horizon, the evolution of DBT as a company will undoubtedly be interesting to watch.