.png)
17 hours ago
3
13 Jun, 2025
About a week ago, I posted on LinkedIn about how a Snowflake product announcement during the Snowflake Summit 2025 explained some of dbt Labs’ recent decisions that caused controversy.
For me, the connection felt obvious, and I was excited to share it. But what followed was even more fascinating than my pride in connecting the dots of a five-dimensional corporate chess game.
The post drew a good amount of attention and sparked discussion. The result was a handful of lessons about human nature.
First, we humans are psychological, not purely logical, beings. Perhaps unsurprisingly, we go to great lengths to bend logic and rationality to justify our feelings. Occam’s Razor is useful to brandish in debate when you want to sound intellectual, but applying it to interrogate your own thoughts? Not necessarily.
A vivid example emerged in the thread about dbt Fusion licensing and Snowflake’s announcement: one commenter claimed that Snowflake had already bought dbt Labs and that the announcement merely made this secret deal “official.”
Others argued that dbt Fusion is already available on Snowflake.
Still others said there was an agreement for Snowflake to license the new version of dbt and distribute it through the platform, an argument they felt contradicted dbt Labs’ decision to change dbt Fusion’s license to gain more leverage over its IP.
My original thesis still stands, and I want to add a bit more color to continue the conversation.
In short: Snowflake’s decision to offer dbt Core as a native Snowflake feature explains and, in my view, justifies, dbt Labs’ move to change dbt Fusion’s license.
More speculatively, the announcement also sheds light on the SDF acquisition, its price, and how quickly dbt Labs integrated it into the product line. The deal bought dbt Labs time: it let the company keep growing, limit distribution disruption inside Snowflake, and avoid changing dbt Core’s license, something that would have been far more painful for both the company and the community.
In other words, dbt Labs acted rationally in response to market conditions.
People have long said that dbt Core is “a weekend project” and that dbt Labs has no real moat. The moat was never the tech; it was the community and the distribution the company built on top of it.
dbt Labs created a tight collaboration with Snowflake. Using dbt Core was good for Snowflake: it boosted warehouse consumption and opened the platform to teams that previously lacked data engineering resources, evidence of which follows:
By 2025, dbt Labs reported serving 60,000+ data teams globally and tens of thousands of weekly active companies using dbt . A scale achieved largely by riding Snowflake’s cloud-warehouse wave.
Analysts note a “huge overlap of the community” between the two companies, and dbt’s leadership hopes deep integration will convert more Snowflake users into dbt Cloud customers .
The partnership is mutually beneficial: Snowflake encourages tools like dbt because they drive workloads inside Snowflake, while dbt Labs gains a steady stream of users and revenue.
As Snowflake SVP Christian Kleinerman said, “We see organizations unlock the power of their data when more people are able to participate in analytics processes,” and partnering with dbt Labs lets Snowflake deliver that participation “cost-effectively, scalably, and securely” on its Data Cloud .
dbt Labs has raised more than $400 million in venture funding, much of it on the promise that dbt would become the transformation layer for Snowflake and other clouds.
In short, Snowflake has been critical to dbt Labs’ growth, providing customers, co-developing products, investing capital, and publicly validating dbt’s value. That alignment propelled dbt Labs from open-source project to $4 billion-plus company at the center of the modern data stack.
Why does dbt Labs have to buy time?
dbt Core is a great tool. It effectively created a new breed of developer, the analytics engineer, turning a traditionally zero-sum game into a positive-sum one. But a tool is not a product. The most obvious commercial wrapper around dbt Core is essentially an orchestrator, a category defined by Airflow. That’s not especially glamorous, and dbt Labs has long tried to frame dbt Cloud as a productivity platform rather than an orchestrator.
Here's some evidence on this based on the major product announcements through the years.
2017–2021. Early announcements pushed productivity messaging hard, and even the pricing model smelled of productivity.
2021–2023. dbt expanded into Python and data engineering workflows via a Databricks partnership. Adoption lagged: Python data engineers already had mature tooling, and few wanted to abandon Airflow.
During the same period, “headless BI” morphed into the semantic layer, and dbt Labs acquired a company to pursue that strategy and still pursues today.
2024–2025. Enter the SDF acquisition, dbt Fusion’s new license, and AI features like dbt Copilot.
Leadership turnover
Product leadership has been a roller-coaster ride:
| 2016–2022 | Drew Banin | Co-founder, acting CPO |
| 2022–2023 | Margaret Francis | CPO (industry veteran) |
| Jul 2023–Jul 2024 | Luis Maldonado | VP Product (ex-AWS) |
| 2025–present | Ryan Segar | CPO (promoted from CCO) |
The high turnover underscores how hard the product puzzle remains.
dbt Labs built an amazing tool that transformed analytics engineering, but no one has yet forged a product that fully justifies the company’s valuation, hence the continual push into new categories.
It’s inspiring to watch Tristan Handy and the dbt Labs team grow the business while chasing a generational vision.
It takes time, but time eventually runs out, unless you have enough cash to buy a little more.
Takeaways
- dbt Labs did the right thing. They figured out a way to buy time and do it in a way that minimizes disruption to the dbt Core community.
- From co-selling to upselling. Now that Snowflake bundles dbt Core, dbt Labs must upsell those users to Fusion. The impact on growth remains to be seen.
- dbt Fusion white-labeling. If Snowflake white-labels dbt Fusion, upselling becomes trickier. dbt Labs will need revenue-share terms or new platforms to OEM the product.
- What's next for the dbt Labs product? This might be the most exciting part: Ryan is smart, ambitious, and knows the customer better than anyone else. I would definitely keep an eye on what dbt Labs will be announcing for the product in the next couple months.
#SaaS strategy #Snowflake #analytics engineering #cloud data warehouse #data engineering #dbt Core #dbt Fusion #dbt Labs #licensing change #modern data stack #open-source licensing #product strategy #startup growth
