In an interview with CRN, dbt Labs founder and CEO Tristan Handy discusses how the dbt Labs-Fivetran merger, announced last week, came about and how the blending of the companies' data movement and data transformation platforms creates opportunities for channel partners looking to spend less time on data plumbing and more on high-value data services around AI workflows.
Fivetran, a leading provider of automated data movement and connectivity software, and dbt Labs, a developer of breakout data transformation technology, last week announced a plan to merge and create a powerhouse player in the open data infrastructure space.
The merger will unify the companies' complementary data movement, transformation, metadata, and activation capabilities within a single company.
The move comes at a time when the demand for data management tools is surging thanks to the data needs of AI systems, including AI applications and AI agents. Businesses and organizations are seeking ways to automate the data workflows that support AI and are demanding systems with integrated capabilities -- rather than point products -- to make that happen.
That trend is driving some measure of consolidation across the big data arena with Salesforce's pending $8-billion acquisition of Informatica being a major example.
Oakland, Calif.-based Fivetran will merge with dbt Labs, headquartered in Philadelphia, with the resulting unified company approaching $600 million in annual recurring revenue. Fivetran co-founder and CEO George Fraser will be CEO of the new company while dbt Labs founder and CEO Tristan Handy (pictured) will serve as co-founder and president.
Fivetran and dbt Labs are both privately held companies and share some investors including private equity heavyweight Andreessen Horowitz. Dbt Labs investors also include Altimeter Capital and Sequoia Capital while Fivetran investors include General Catalyst and Matrix Partners.
The all-stock deal is based on an agreed upon ratio tied to revenue and growth rates, according to a Reuters report, with the combined entity worth more than its last private valuation. The merger has been approved by each company's board of directors and shareholders, but the deal is still subject to customary closing conditions and regulatory approvals.
In the merger announcement, the two companies stressed that after the merger is completed, the new company will remain committed to keeping dbt Core, the central technology that powers the dbt Labs platform, an open-source technology under its current license and maintaining it "with and for" the technology's community.
This week CRN spoke with Handy about the pending merger, including how it came together and how the merger aligns with current trends in the big data and AI space. The following Q&A has been edited for clarity.
First, I want to ask a couple of questions about the mechanics of the merger deal. Is one company actually buying the other?
The agreement that we signed is actually called a merger agreement. Nobody is buying anybody. It is actually a combination of the two [ownership] share pools. In the process there's no assigned value to the equity. It's not like we assigned a value to Fivetran's shares and then dbt got, you know, a dollar-value on the [shares]. We literally agreed to a relative percentage [of ownership] and we put the companies together.
What will the new entity's name be and where will its headquarters be?
So we actually don't have a name for the combined business yet. On our side we've been calling it 'NewCo.' When we're referring to it in public, we are actually just saying 'DBT and Fivetran.' It's interesting, both of the companies' brands have a lot of brand equity. They're widely known and they're widely known for very specific and different things.
And so there's kind of three options on the table. We could call the combined company 'Fivetran.' We could call the combined company 'dbt.' Or we could call the combined company some third thing that we would make up. And, actually, we haven't had the time to even answer that question yet. So that is a thing that I imagine -- I don't know whether it's over the next two months or the next six months -- it's something that we need to figure out.
Who are the owners of the two companies and who will be the owners of the new company?
So both companies are classic venture-backed companies. All of this is publicly available. Fivetran owners [include] Andreessen Horowitz, Matrix Partners, General Catalyst. We actually share a board member from Andreessen Horowitz, so Andreesen Horowitz has big positions in both companies. And on [the dbt Labs] side we have Sequoia and Altimeter and some others.
So Andreessen Horowitz has a position in both companies?
Yes, right.
Were they a catalyst in this merger? I'm thinking of the case where Qlik and Talend were both owned by private equity firm Thoma Bravo, which engineered Qlik's 2023 acquisition of Talend.
Venture capitalists are almost always minority investors in any company they invest in. A venture capitalist almost never takes a controlling stake of any company. So, no, it's not like Andreessen Horowitz could make this happen.
Really, the two co-founders of Fivetran, [CEO] George [Fraser] and [COO] Taylor [Brown] and I have worked side by side for over a decade at this point, and, honestly, a lot of other people in the companies just know each other well, there are employees that have moved back and forth between the companies, there's a bunch of relationships.
We have 1,500 shared customers. We have more shared customers than many companies of our scale have customers at all. So the connective tissue between the companies is just very deep.
Honestly, sometimes when venture capitalists get involved in trying to suggest things like this, it creates resistance from the respective principles.
Why do this from a technology perspective, and why do this from a business perspective?
Oh, gosh, I keep joking that at some point I want to be an adjunct at Harvard Business School and write the Harvard Business School case study about this because I think it's so interesting.
What I said on stage at Coalesce [the recent Coalesce 2025 analytics conference] is that the industry dynamics right now really prioritize scale. Customers have become impatient with the idea of having to buy 12 different products to create a data ecosystem, and they've become impatient with duct taping a bunch of stuff together. End to end [data management] solutions are more and more valuable.
Our biggest partners, what we call the 'Big Five:' AWS, GCP, Azure, Databricks and Snowflake -- they all started out as compute platforms, whether that's in Databricks' case a spark-based compute platform, or whether, for the other four, a kind of SQL-based compute platform. But they've all started to broaden their offerings to become what we're calling, quote, unquote, all-in-one data platforms.
And so our belief is that we needed to provide an end-to-end solution that didn't lock customers into using any particular compute platform. All of our products work with any compute platforms from the Big Five, but also with newer compute offerings like DuckDB, [Apache] DataFusion and ClickHouse. Those engines are much more tuned for the type of workloads that AI really cares about. And so we think about this as open data infrastructure for both analytics and AI.
Can you drill down a bit as to what dbt Labs and Fivetran each bring to the merger from a technology perspective? What are the benefits of each technology portfolio and what's the one-plus-one-equals-three results?
Fivetran is data connectors. They do data ingestion. They have 700 connectors and the best, most robust enterprise grade connectors in the world that bring data into your data lake or data warehouse. And then we take it from there, we do the [data] transformation on top of that.
Our two products were literally designed to be used together. Fivetran, I think they started working on Fivetran in either 2013 or 2014, I forget. [Fivetran was founded in 2012 and launched its automated data integration platform in 2013.] But seeing Fivetran back in 2016 was part of my mental model and was what made me realize I needed to build dbt on top of that data, because at the time it was increasingly clear that there were going to be solutions to get data into your data warehouse, but there was no transformation tool. So I literally built dbt to be used with Fivetran.
So how will customers and channel partners benefit from this combination?
Let me [talk about] channel partners first, because I think that's an interesting one. Customers don't have a problem of, 'How do I do data transformation?' or 'How do I ingest data?' or 'How do I do data observability?' They have problems like, 'How do I build a customer 360 [system]?' or, 'How do I make sure that I have my regulatory reporting on top of my SOC [System and Organized Control standard] compliance?'
And so when a channel partner comes in, they come in to help solve one of those types of business problems. And they have a strong preference to bring fewer [IT] vendors into the mix, because every vendor that you bring in as a channel partner -- and I know this as [having worked] at a very small-scale system integrator back in the day -- every technology vendor you bring in is another procurement hurdle. It slows things down. You have to convince the customer that they need to buy this [additional] thing.
So if you can do more with fewer vendors, that's better. And we validated that with our Partner Day, at the beginning of Coalesce this year. We had, like, 400 people from partners in the room and everybody was very excited about what we could do together.
Of course there is also this big upgrade cycle happening in the dbt world where we're moving from dbt Core to dbt Fusion. And that is another big opportunity for channel partners because there will be some level of work required for that upgrade.
So channel partners right now see this as a big opportunity to do a lot of work. Fusion also runs much more efficiently, so it saves customers a bunch of money. What that means -- I actually asked folks in the room [at Coalesce], 'What happens if you can save customers 33 percent on their data platform costs? Where does that money go?' And their answer was [that] there's all these other AI-related projects. And that's great for [partners] because [they] would much prefer to be helping [customers] get that stuff off the ground.
So providing services for the higher-value AI projects rather than the lower value IT integration kind of stuff?
Correct. And ultimately, I think that's where the big data platforms want to go to, whether it's Databricks and [AI/BI] Genie or Snowflake and Cortex, everybody wants to be moving in the direction of powering more AI workloads. And so we hope that the integration between dbt and Fivetran -- plus Fusion, with a bunch of cost savings -- is going to help people focus more on the new and less on the pipelining.
The interesting thing from a customer perspective, for our direct customers, is that the two products are already well-integrated today, we've been partners forever. And George [Fraser] was very upfront about this in front of our Executive Summit [at Coalesce]. He said, 'Look, yes, I'm sure that the two products will become more well-integrated over time. But really, dbt is still dbt, Fivetran is still Fivetran, and this is non-disruptive for customers.'
