Data infrastructure used to be like Victorian plumbing: every source and transformation required its own pipe, coupler, and patch. Teams stitched together connectors, wrote brittle scripts, and built home-grown orchestration just to move data from SaaS applications into a warehouse. That complexity spawned an entire generation of analytics engineers and a cottage industry of ETL tools.
In October 2025, that landscape shifted when Fivetran and dbt Labs announced an all-stock merger worth nearly $600 million in annual revenue, the modern data stack’s first true megamerger.
The Backstory: Built to Be Together
Fivetran’s push-button connectors made extraction and loading effortless: flip a switch, and Salesforce or Postgres data appears in Snowflake. A few years later, dbt introduced the software engineering discipline to analytics with version-controlled SQL and testable models. Their products were so intertwined that dbt’s founder, Tristan Handy, once said they were “literally built to be used together. That interdependence laid the groundwork for a merger that now feels almost inevitable.
When the merger was announced, Handy noted that the combined company would serve more than 10,000 customers and generate approximately $600 million in recurring revenue, formalizing years of collaboration that had already blurred the line between the two brands.
He also promised continuity: dbt will remain dbt and Fivetran will remain Fivetran, and the open‑source dbt Core project will keep its current licence. Years of collaboration – from shared booths at conferences to joint support tickets – made the deal feel less like a takeover and more like a formalisation of a long relationship. Even Andreessen Horowitz board members urged the founders to merge because the products already overlapped.
Why Merge Now? AI and Competitive Pressures
Several factors explain why the tie‑up is happening in 2025. First, there’s the AI boom. Fraser told Reuters that the combined company aims to build open, interoperable infrastructure to support enterprises racing to use their business data for artificial‑intelligence applications. Running AI models at scale demands clean, reliable pipelines and consistent metadata; fragmented tools slow data teams down. A unified platform promises to reduce friction and accelerate AI readiness.
Second, there’s an economic logic. Extraction and loading account for perhaps a quarter of an organisation’s analytical compute; transformations consume another big slice. Tower.dev notes that by owning both E/L and T, and through its acquisition of Tobiko Data, developer of dbt competitor SQLMesh, even more, Fivetran doubles the market it can address. The company has already been on an acquisition spree, buying the reverse‑ETL pioneer Census and the transformation start‑up Tobiko Data. Owning more of the pipeline means capturing more consumption‑based revenue and insulating the business from threats like open table formats (Iceberg, Delta Lake) that allow companies to keep data in their own files. Without expanding into transformation and reverse-ETL, Fivetran risked watching its core extraction business shrink as open formats proliferated.
Third, competitive dynamics in the broader ecosystem encourage consolidation. Cloud warehouse providers such as Snowflake, Databricks, and Microsoft Fabric ship native connectors, built‑in transformation frameworks, and even end‑to‑end platform bundles. To remain relevant, smaller vendors are merging. SiliconANGLE reported that Fivetran’s talks with dbt were part of a multibillion‑dollar acquisition race to assemble a complete platform. Analyst Jennifer Stirrup warns that this consolidation signals “dramatic shifts ahead in vendor risk”.
The Promise of an Open Platform
The merged entity argues that it is building something different from the monolithic suites of the past. Handy calls the vision ‘open data infrastructure’, a pluggable, standards-based platform that lets users run pipelines on Snowflake, BigQuery, Databricks, or their own lake with push-button simplicity. Fraser emphasises that dbt Core will remain available under its current licence and that the combined company will continue to contribute to open source. This stance could reassure organisations wary of being locked in.
If the vision holds, the unified platform offers tangible benefits:
Push‑button ingestion and modelling. Teams can configure sources once, build transformations in dbt, and let the system orchestrate incremental loads and dependency graphs, reducing setup time and engineering toil.
Unified metadata and lineage. Consolidating extraction and transformation allows the platform to capture lineage, test results, and freshness metrics in one place, improving governance and troubleshooting.
Simplified operations. One vendor for ingestion, transformation, and (potentially) reverse‑ETL reduces the number of contracts, connectors, and systems to manage. Support and billing become more straightforward.
Commitment to open source. dbt Core and its successor, dbt Fusion, will continue to be shipped under open-source licenses, offering a safety valve for teams that need to run pipelines outside the managed service.
For B2B data teams stretched thin, these benefits could translate into a faster time to insight and fewer middle-of-the-night on-call pages.
The Dark Side: Vendor Lock‑In and Innovation Risks
Every consolidation has drawbacks. Stirrup cautions that a unified Fivetran–dbt platform could dominate data movement and transformation, increasing the risk of vendor lock‑in and reducing buyer power. If Fivetran builds a proprietary warehouse layer, truly neutral ETL partners may vanish. Reduced competition can slow innovation and lead to price increases. Larger platforms also tend to move slower; Stirrup notes that while deep integrations are nice, leaders should watch for nimble upstarts and evolving standards such as DuckDB and Iceberg that could leapfrog incumbents. Tower.dev notes that when one dominant vendor is absorbed, new challengers inevitably emerge, and customers resist the concentration of power.
There are other concerns. Consolidation does not magically solve technical complexity. Customers will demand robust SLAs, transparent data‑ownership terms, and strong security as ingestion and transformation converge. They will also need to scrutinise pricing models. Consumption‑based billing can be attractive, but bundling more services may make it harder to compare costs across vendors and could result in expensive renewals. As the open‑table‑format movement gains steam, enterprises should ensure they are not locked into proprietary storage or connectors.
Blueprint for B2B Leaders
So how should technology leaders respond? Here’s an expert-informed playbook:
Map your dependencies. Audit your data pipelines and identify where you rely on Fivetran, dbt, or other single vendors. Understand where overlapping features lead to double payments and where you could swap in an open‑source component.
Prioritise open standards. Invest in tools and formats that avoid lock‑in. Lake‑house table formats like Iceberg, Hudi, and Delta Lake separate storage from compute. Open‑source ingestion frameworks such as Debezium or dltHub and transformation engines like dbt Core or SQLMesh can coexist with managed services.
Negotiate contracts carefully. Consolidation changes pricing. Ensure your agreements include transparent consumption metrics, fair renewal clauses, and exit options. Monitor how bundling affects your total cost of ownership and benchmark against alternatives.
Build a multi‑vendor strategy. Avoid putting all your eggs in one basket. Use Fivetran for some sources and open‑source tools for others; run dbt Core in‑house for sensitive workloads and use the managed service for the rest. Multi‑vendor setups increase resilience.
Invest in skills and community. Encourage your team to keep learning. Python‑based data‑engineering stacks (Polars, Apache DataFusion) are improving and could form the basis for future open alternatives. Participate in the dbt community and other forums to stay abreast of best practices.
Looking Beyond the Merger
The Fivetran–dbt deal doesn’t close the book; it opens a new chapter. Consolidation today will likely spur another innovation cycle as open-source projects and Python-native stacks mature. AI’s rise will only heighten the demand for seamless, governed data pipelines.
For B2B leaders, the message is clear: reevaluate your vendor exposure and prioritize open standards. You can build flexibility into your data architecture. Consolidation may simplify operations, but openness and resilience will decide who thrives in the next wave of data innovation.
