Fivetran and dbt Merger: What It Means for Your Data Stack in 2026

On June 1, 2026, the Fivetran and dbt merger was confirmed. The merger was first announced on October 13, 2025. It brings together the two most widely used tools in the modern data stack, automated ingestion and SQL transformation, under one company.

The timing makes sense. Companies are spending heavily on AI right now, and they are learning that the hard part is no longer the AI models; it is the data feeding them. That is where ETL pipelines come in: they move and prepare the data. If those pipelines are unreliable, everything downstream does not give the expected results.

Gartner predicts that through 2026, organizations will abandon 60% of AI projects that lack AI-ready data. The Fivetran dbt merger is a bet that the fix is a single, trusted foundation from ingestion to transformation.

This page breaks down what actually changed, the new dbt Fusion engine and the AI-focused launches that shipped with the merger, and what it all means for how you build, run, and pay for data pipelines, plus where a managed alternative fits.

The Fivetran Migration Guide Data Teams Actually Need

Learn how to migrate from Fivetran to Hevo with zero downstream changes to your dashboards or data models.

Download for Free

The Deal in Brief

The combination is an all-stock merger. George Fraser, Fivetran’s CEO, leads the unified company; Tristan Handy, dbt Labs’ founder, is co-founder and President. The combined business is approaching $600 million in annual recurring revenue, and dbt alone is used by more than 80,000 data teams worldwide.

The merger is one of the biggest decisions by Fivetran. In 2025, it bought Census (a reverse ETL tool) and Tobiko Data (the makers of SQLMesh). Adding dbt’s transformation layer on top gives the combined company coverage across ingestion, transformation, reverse ETL, and metadata.

Why the Fivetran and dbt Merger Happened

The goal is to build a data infrastructure that aids AI. The logic is simple: AI is only as good as the data behind it. To trust an AI agent, you need data that is fresh, clean, and well-defined, and that takes two jobs done well: getting raw data in reliably and shaping it into business-ready models.

Fivetran has always owned the first job: moving data from hundreds of sources into a data warehouse. dbt has owned the second: transforming that raw data into clean, tested models. Together, they create one path from source to analytics-ready data, for human analysts and AI agents alike. It also blurs the line between ETL and ELT, since ingestion and transformation now live under one roof.

There is a competitive reason too. The all-in-one platforms, Snowflake, Databricks, and Microsoft Fabric, have been pulling ingestion and transformation into their own ecosystems. The merger is a bet that teams would rather have an open layer that works across any warehouse than get locked into one vendor’s stack.

Hevo already cracked the code for AI-ready data. While the merger works toward a trusted data foundation for AI, Hevo delivers it today: fresh, clean, governed data into your warehouse, transformed with dbt or Python, in one managed platform. See how Hevo works →

The Result of the Fivetran dbt Merger

The bigger news for practitioners is the set of products that shipped alongside the close. Five launches stand out.

dbt Core v2.0 and the Fusion engine

Fusion is a full rewrite of dbt’s core in Rust that turns dbt into a true SQL compiler. dbt Labs reports parsing up to 30 times faster on large projects (10,000+ models), and it can validate your SQL locally, catching errors without a warehouse run. It ships in dbt Core v2.0 under the Apache 2.0 license, so the core stays open-source. One caveat: v2.0 launched in alpha, so it is still early.

dbt State

This makes runs smarter by tracking what has actually changed and skipping models that have not. That is where the cost savings come from. dbt Labs cites 30% or more off warehouse compute, and early customers back it up: Obie Insurance saw 30%+, and EQT cut costs by 50% while running 60% faster. It is in preview.

dbt Wizard

An AI assistant for building and debugging models in plain language, aimed at making transformation work easier for less technical users. It is in beta.

Agents Schema

An open standard for a dedicated schema that gives AI agents clean, governed business context, so they can query data reliably without a human in the loop.

AI Connector Builder

A tool that generates Fivetran connectors from API documentation in minutes, which speeds up pulling data from new or niche sources.

How Does the Fivetran and dbt Merger Affect the Data Teams

If you use Fivetran or dbt today, you probably have a few questions. Here are the ones that matter, answered simply.

Do I need to do anything right now?

No. Both tools keep working exactly as they do today. Your pipelines and schedules are safe, and no one is forcing you to move or upgrade. You can carry on as normal and watch how things develop.

Will I be pushed to use both tools?

Over time, likely yes. The two products will get more connected, shared setup, bundled plans, and features that work best when you run both. If you already use Fivetran and dbt together, that is good news. If you use just one, expect more nudges to add the other.

Am I now tied to one vendor for too much?

This is a fair concern. Ingestion and transformation, two of the most important parts of your stack, now belong to one company. That can mean less room to negotiate on price and less freedom to pick the best tool for each job. If you have been mixing tools, it is worth keeping an eye on.

Could this lower my bill?

Possibly, yes. If you run big dbt jobs often, you are likely paying to rebuild data that has not changed. A new feature skips that work, and dbt reports 30%+ savings on warehouse costs. It is worth testing once it is past the early alpha stage.

Why wait for a unified stack? Hevo already is one. Hevo already combines no-code data movement with built-in dbt and Python transformation in one managed platform, no alpha releases, no second vendor.  See how Hevo works →

What Actual Data Engineers Are Saying About This Merger

The community had plenty to say when the merger broke. Here is the sentiment, in their own words, from the Hacker News discussion on the day it was announced.

1. Concerns about dbt Core’s open-source future

The biggest worry is what happens to dbt Core. Even with the open-source commitment, many expect development to slow. As one engineer put it, “the concern is that dbt-core will become stagnant.” Another was blunter about the motive: “dbt Cloud is an uncompelling product. Fivetran is convenient but absurdly expensive. Now dbt core development is just another marketing cost for Fivetran.”

There is a real counter-argument. dbt Core has around 2,400 forks on GitHub and an Apache 2.0 license that cannot be revoked. If the community dislikes where things go, developers can fork the project and keep it alive on their own. The license is the safety net.

2. Pricing worries

Fivetran already had a reputation for high bills before this deal, so consolidation makes people nervous. One commenter summed up the fear plainly: “this type of consolidation will very likely bring increased prices.” The worry isn’t abstract, one engineer shared a real number: “They charged us over $140k / year to basically stream ~1TB of production data (50m rows) from our Postgres into a redshift database.”

The history behind the nerves: when Salesforce bought Tableau and Google bought Looker, both drew complaints about slower innovation and rising prices over time. The community worries this follows the same pattern.

3. Questions about strategy

The sharpest question is why Fivetran bought dbt at all when it already owns SQLMesh, a dbt competitor, through its 2025 Tobiko Data acquisition. One widely-shared comment laid out the fear: “I can’t imagine Fivetran spent a ton of money just to have 2 products that do very similar things… With this announcement, I’m worried we’re going to end up being forced to migrate to dbt.”

The read from the thread is that Fivetran bought SQLMesh for the technology and dbt for the customer base. Fivetran’s own estimate is that 80% to 90% of its customers already use dbt, so this looks less like cross-selling and more like locking in both user bases. As one engineer framed it, the move is “likely just a move to lock-in customers of both companies by upselling an ingestion/transformation layer to existing customers.”

Not everyone is negative. A Fivetran product manager joined the thread to push back: “We are fully committed to open source dbt and don’t want to build a ‘walled garden.'” Tighter integration between ingestion and transformation is genuinely useful, and if the merger pushes more of the stack toward open standards, that could be good for everyone.

Hevo as an Alternative to the Fivetran and dbt Stack

The merger is built on an idea Hevo has followed for years: ingestion and transformation belong together, in one place. Hevo delivers that through three principles, simplicity, reliability, and transparency, which happen to answer the exact concerns the merger raises.

Unified already, nothing to stitch

Hevo combines no-code data movement from 150+ sources with built-in dbt and Python transformation in one platform. You get the unified outcome the merger is promising, without running two products.

Production-ready, not alpha

The Fusion engine and several launches are still early-stage. Hevo’s ingestion and transformation are battle-tested and live now, so you are not piloting an alpha release on production data.

No engine to run

Everything is fully managed: infrastructure, schema changes, and failures are handled for you. There is nothing to host, tune, or maintain.

Predictable pricing, no lock-in worry

Flat, transparent pricing means no surprise bills, and you are not betting on how one vendor bundles or reprices two critical layers over time.

What Teams Say After Moving from Fivetran to Hevo

This is not hypothetical. Some teams hit the same walls, pricing, support, and reliability, and they already switched. Ebury is one of them.

Ebury, a global fintech, ran on Fivetran until bugs in its History Mode, slow support, and high pricing pushed them to look elsewhere. They evaluated Hevo against Stitch, Fivetran, and Matillion, and chose Hevo for better reliability, 15-minute load frequency, built-in reverse ETL, and pricing that worked. The result: 100% data accuracy and 50 million events a month syncing with Salesforce.

Not sure the Fivetran + dbt path is right for you? Test the alternative. Run ingestion and transformation in one platform, free, and see if a managed, no-lock-in setup fits your team better. Try Hevo free →