Should Snowflake Buy dbt or Let it Die
Where Innovation Meets Community Expectations
You love the modern data stack. You love its accessibility and scalability. But above everything, you love its community!
But let's talk about something that's been on my mind and, apparently, on
’s too. His thought-provoking poll last week caught my eye, echoing questions I've been pondering recently.What if dbt Labs are at a crossroads? What if they need to choose between soaring high or facing a downturn?
It would be a critical moment, not just for them but for the entire data community.
And what if Snowflake have the opportunity to buy dbt Labs?
I dive into this scenario, examining the potential outcomes and implications for the data community. It's a journey through possibilities and speculations to uncover what the future might hold for dbt.
Ready to explore this with me?
Let's delve into the future of dbt and discover its potential paths!
The Spark that Started it All
We had cloud warehouses and lakehouses way before dbt. We even had other transformation tools before dbt. But the vision of the team behind dbt and the community around it were the flint that sparked the movement we now call Modern Data Stack.
dbt started as a tiny tool built by yet another consultancy. However, it was precisely what analytics engineers (did we even have that role back then?) needed. The word about dbt spread swiftly, and the community grew blazingly fast.
Fishtown Analytics doubled down on the hype train and secured a massive amount of money. They morphed into dbt Labs and transformed the product into a powerhouse.
It's your dream, too, right?
Building something everyone loves. Securing funding that lets you focus on your passion. Being at the core of an entirely new movement in the data world. It just can't be better!
But, as they say, with great power (and investments) comes great responsibility. Now, dbt Labs has to navigate the tricky waters of turning a profit for their investors. And that's where things get interesting.
This journey from a small tool to a major player in the data industry is more than just a success story. It's a testament to the power of vision, community, and suitable investment. But it's also a prelude to the challenges that dbt Labs now faces – challenges that could shape the future of the modern data stack.
The Gold Mining Journey
Let's talk about dbt Cloud. It's a compelling offer, sure. Its de facto is the default way to start with dbt. But here's the thing:
It's essentially an orchestrator with an editor. And honestly, that's a problem.
Why?
Because there are alternatives out there, some even free and local, that can do the job just as well, if not better.
The truth is there are equally good dbt editors out there. And I'm not talking about team-oriented cloud workspaces like Paradime. I'm talking about those which you can run for free, locally, with no Internet connection.l
If you have ever seen Deep Channel, you know what I am talking about. It has it all — autocompletion, expanding, lineage, DB explore, query runner, and so much more.
And if you are a seasoned engineer, you might prefer to use something even more potent like the dbt Power User VS Code plugin or dbtpal by
(I use Neovim, btw!). No matter how good the dbt Cloud editor is, it just can't compete with such versatile tools.But here's my take:
The real issue isn't the editor but the orchestrator. From my experience, it doesn't offer too much beyond a basic cron job functionality.
In fact, that's the main reason we moved away from dbt Cloud. It was pretty hard to use in more complex use cases like ours. We had multiple related projects that required notifications in different Slack channels. We also needed to run various linting and testing processes outside of dbt. None of that was possible back then.
However, in fairness, dbt Cloud has seen improvements over the years. New features have been added, catching up with customer needs. Even the pricing model has changed, sparking debate. And contrary to popular opinion, I don't find the pricing too wild. In fact, I'd argue it makes more sense than the original one.
So, all this leads to a crucial point:
The survival of dbt Labs and the community depends on converting more dbt Core users to dbt Cloud.
But can dbt Mesh and MetricFlow be the catalysts for this shift?
I'm sceptical. The team needs something groundbreaking, something that redefines our expectations. Something we don't even know we need.
I don't know what that thing is. But I hope dbt Labs will come up with something that will make us beg to throw money at them. And that's the big question:
What if dbt Labs doesn't find that 'something'?
What if they don't innovate quickly enough?
The Failure Scenario
Right now, it seems like they're shooting in the dark, trying to hit a target that isn't clearly defined. To be fair, this isn't uncommon for startups, but time is ticking, and the stakes are high. Sure, they've got the cash from their last funding round and a good number of paying customers, but that's not an endless runway.
Now, here's the truth:
dbt Labs might be in a bind because dbt Core is too good and, well, free. It's a paradox.
dbt Labs could fail because the people behind it are doing too much good for the data world.
But that's only part of the story:
If dbt Labs struggles, it wouldn't necessarily spell doom for the community, as most businesses use dbt Core. And remember, once a project goes open-source, it stays that way. There are over a thousand forks of dbt on GitHub, any of which could become the new community standard.
In essence, if dbt Labs wins, we all win, but if they lose, well, life continues for all of us.
But let's pivot to another scenario. Imagine if dbt Labs is up for grabs.
The Acquisition Scenario
I think Snowflake might be interested if dbt Labs goes for sale.
Why?
Snowflake is a leader in innovation, constantly pushing the boundaries of the data industry. Snowflake partners with all major BI tools. They partner with all major EL tools and reverse ETL companies. They literally have a stake in every single aspect of the data space.
On top of that, Snowflake made some exciting investments in the last few years. I'm talking about the acquisitions of Streamlit, Leapyear, Neeva, and so many more.
Snowflake is forward-thinking, and once they identify a niche, they enter early and fast. They are obviously very bullish about the future of data.
So, should Snowflake buy dbt Labs?
It's a tantalising thought, but it's not straightforward. There are pros and cons to such a move, and the impact on the community and the industry would be significant.
The Pros
The benefits can be quite exciting when we think about Snowflake acquiring dbt Labs. Given dbt's influence and popularity in the data community, this move could be a strategic masterstroke for Snowflake.
Expanding the User Base
One of the most significant advantages lies in the user base. dbt boasts a robust and dedicated community with over 90k Slack users who love dbt unconditionally. If Snowflake were to acquire dbt Labs, it could tap into its user base, converting them into Snowflake customers. This conversion wouldn't just be about numbers; it's about integrating a community deeply invested in dbt into the Snowflake ecosystem.
Enhancing Snowflake's Reputation
Snowflake's reputation in the data community is already strong, but saving a beloved tool like dbt from death could take it to new heights. This would be more than just another product to Snowflake's portfolio. Snowflake would show commitment to the broader needs and values of the data community. Such a move could strengthen the trust and loyalty of current and potential Snowflake users.
Creating a More Cohesive Data Workflow
Owning dbt could allow Snowflake to integrate it more deeply with its existing services. This integration of dbt could streamline the data workflow, making it more efficient and user-friendly.
A perfect example is what happened with Streamlit. Even after that acquisition, you can still use Streamlit for free, build your apps, and self-host those apps. Hell, Streamlit still lives in its original GitHub space!
Building a Full-Stack Data Process
Snowflake already has all the infrastructure needed to support any dbt use case.
Just picture this:
You could get most of your data through Snowflake sharing and dynamic tables, orchestrate dbt jobs with native Snowflake tasks, and then share Streamlit-powered dashboards directly from Snowflake.
I'm talking about the full-stack data process here!
Now, like most other things, a potential acquisition wouldn't be just flowers and roses. There are some potential pitfalls we need to mention.
The Cons
Snowflake is not Google. They still don't have a history of killing services shortly after launch/acquisition. Yet some people would expect Snowflake to close the source code of dbt or make dbt Snowflake-exclusive. And to be fair, those concerns are always valid.
The Challenge of Supporting Competitors
By acquiring dbt Labs, Snowflake would find itself in a tricky position:
Investing in a tool that also benefits its direct competitors. This situation could lead to conflicts of interest and strategic dilemmas. Buying dbt Labs would mean that Snowflake should spend money on a product that directly benefits its competitors.
Facing Complexity and Redundancy
Every new feature in dbt increases complexity significantly. Snowflake must produce code, documentation and testing for all officially (and potentially unofficially) supported backends.
At the same time, a juggernaut like Snowflake can develop a transformation tool explicitly tailored to its platform. Such a tool would be much easier to maintain and have significantly faster run times.
Risk of Alienating Locked-In Customers
Snowflake is not famous for its low prices (whether that's on them or us is another topic). Yet, there are companies stuck with the platform because of dbt. Those companies have invested a lot of cash into vendor-locked models. They are unwilling to risk rebuilding those models, not just because of the high costs but also because of the risk.
Think about it:
Facing an unfavourable change in dbt's functionality or integration, combined with the dream for cost savings, could form the perfect storm. It might catalyse data teams who have postponed migration to another platform for years.
The High Cost of Acquisition
Finally, there's the matter of dbt Labs' price. Last year, dbt Labs raised $222 Million with a more than $4 Billion valuation. Acquiring dbt Labs might be a substantial financial undertaking for Snowflake. This expense would need to be justified by the expected benefits and integration potential of dbt within Snowflake's suite of services. And I don't think that's true at this point.
Final Thoughts
When weighing the pros and cons, the future of dbt Labs and its relationship with Snowflake becomes a complex equation.
I hope, no, I believe, dbt Labs will continue to innovate and pivot as necessary, maintaining its relevance and success in the modern data stack landscape. They have the talent, the community support, and the vision to keep pushing forward. They deserve success!
But in light of the discussion, here's my final take:
If the opportunity arises and the price is right, Snowflake should seriously consider acquiring dbt Labs. It's a strategic move that goes beyond financial calculations. At the end of the day, it's better to have a stake in the most beloved tool in the data community and steer the wheel than let somebody else do it.
Not taking bold steps can be riskier than cautious inaction. Acquiring dbt Labs could be a decisive move for Snowflake, ensuring a pivotal role in shaping the modern data stack landscape.
As the data world watches and waits, the decision will undoubtedly have far-reaching implications for the future of the modern data stack.
What are your thoughts?
Do you agree that Snowflake should seize the opportunity to acquire dbt Labs if it becomes available?
Did you enjoy that piece? Follow me on LinkedIn for daily updates.
