When two of the most widely used tools in modern analytics, Fivetran (for ingestion) and dbt Labs (for transformation), decide to merge, it’s more than industry news. For small and mid-sized companies, it raises important questions about how to build scalable, maintainable, and cost-efficient data stacks.
What parts of your stack are now redundant? Which areas become more efficient, or more risky? And how should teams think about their next move in a post-merger landscape?
What Happened
On October 13, 2025, Fivetran announced an all-stock merger with dbt Labs, uniting two of the most popular components of the modern data stack. The combined company is projected to reach approximately $600 million in annual recurring revenue and will operate under a joint leadership structure: George Fraser (Fivetran) as CEO and Tristan Handy (dbt) as President.
The companies describe their shared vision as “open data infrastructure,” a unified ecosystem that simplifies data movement, transformation, and activation without locking users into a single compute engine. dbt Core will remain open source, preserving flexibility and community contributions, while both platforms continue to operate independently until the deal officially closes.
Two main factors drove this merger:
- Consolidation pressure. Companies are demanding simplified, end-to-end data workflows with fewer integration points and less maintenance.
- AI readiness. Reliable, low-latency pipelines are becoming mission-critical as organizations look to operationalize AI and analytics.
The merger also reflects a broader shift in the data tooling market. As competition has intensified, with lakehouse platforms expanding upstream and open-source tools gaining traction, Fivetran and dbt have moved to strengthen their position by uniting. It’s a strategic decision that consolidates two category leaders into a single, more comprehensive offering designed to hold the middle ground in an increasingly crowded ecosystem.
What This Means for Growing Data Teams
The merger promises efficiency and simplification. Tighter integration could reduce the need for “glue code” and manual orchestration between ingestion and transformation layers. But there are tradeoffs to consider.
| Potential Advantages | Possible Drawbacks |
| Seamless data flow between ingestion and transformation | Increased vendor lock-in and consolidation of control |
| Less operational overhead and faster implementation | Reduced flexibility for unique or complex workflows |
| Shorter time to insights for standard use cases | Potential deprioritization of advanced or custom features |
For example, a small team that currently uses Fivetran for EL, dbt for T, and Airflow to coordinate dependencies could see that middle layer shrink. The ingestion-to-transformation path may become almost plug-and-play, freeing up time to focus on modeling and analytics rather than pipeline maintenance. However, custom orchestration (such as multi-domain triggers or mixed workloads) will still require tools like Airflow or Dagster.
Should Teams Still Write End-to-End DAGs?
The short answer: only when they add real value.
The Fivetran–dbt merger doesn’t eliminate the need for orchestration, but it does shift when and why it’s necessary. Here’s how the main approaches compare:
1. Managed Stack (Fivetran + dbt Cloud)
Best for: Lean teams prioritizing speed and simplicity.
- Pros: Quick setup, minimal maintenance, strong reliability.
- Cons: Less flexibility for custom logic or cross-system orchestration.
2. Modular / Open Stack (Airbyte + dbt Core + Airflow/Dagster)
Best for: Teams needing control, cost flexibility, or hybrid integrations.
- Pros: High configurability, open ecosystem, no single-vendor dependence.
- Cons: More assembly required and higher operational overhead.
3. Lakehouse-Native Stack (Databricks DLT, Snowflake pipelines)
Best for: Companies already invested in a unified platform.
- Pros: Simplified governance, integrated workflows, fewer moving parts.
- Cons: Platform dependence and limited portability.
Guiding principle: If a team spends more time maintaining pipelines than building value from data, it’s time to simplify.
How This Shifts the Industry
This merger is likely to trigger ripple effects across the broader data ecosystem:
- Open-source EL tools (like Airbyte) will double down on flexibility and cost transparency.
- End-to-end ETL vendors (like Matillion or Stitch) will position themselves around simplicity and one-platform convenience.
- Lakehouse platforms (Databricks, Snowflake) will continue bundling more native pipeline functionality, creating competition for independent ELT providers.
This consolidation trend mirrors others in the tech stack: once integration becomes effortless, the next differentiator is governance, AI integration, and cost efficiency.
For other players in the ecosystem, the merger raises the bar for integration and completeness. Competing vendors may respond by specializing further, consolidating, or rethinking pricing and interoperability to stay competitive. The result is likely a period of renewed focus on efficiency and clarity of value across the data stack.
Choosing the Right Stack Pattern
Depending on team size, data maturity, and risk tolerance, most companies fall into one of three stack archetypes:
A. Managed / “No-Ops” ELT
Best for: Small teams and fast-moving startups.
- Stack: Fivetran → dbt Cloud → Snowflake/BigQuery/Databricks
- Pros: Easiest to implement and maintain.
- Cons: Vendor-dependent and higher cost at scale.
B. Modular / Open Source
Best for: Teams wanting flexibility or mixed data environments.
- Stack: Airbyte (EL) → dbt Core (T) → Airflow/Dagster (Orchestration)
- Pros: Open ecosystem, customizable.
- Cons: Requires engineering resources for setup and upkeep.
C. Platform-Native / Lakehouse
Best for: Teams already using Databricks or Snowflake as their primary environment.
- Stack: Native ingestion + transformation (e.g., DLT or Snowflake pipelines)
- Pros: Unified governance and automation.
- Cons: Less flexibility outside the platform ecosystem.
Some companies adopt hybrid models, combining managed tools for standard workloads and open-source orchestration for complex or specialized jobs.
A Simple Evaluation Checklist
To determine which approach fits best, evaluate the following:
- Team capacity: Do you have engineering bandwidth to maintain DAGs and pipelines?
- Data gravity: Where does most of your data live today, warehouse, lakehouse, or hybrid?
- Change velocity: How frequently do your sources and transformations evolve?
- Compliance needs: Are lineage and governance top priorities?
- Lock-in tolerance: How critical is tool flexibility to your roadmap?
If your answers lean toward speed, standardization, and minimal maintenance, a managed ELT stack will likely deliver the best ROI. If flexibility and control rank higher, an open or lakehouse-native setup may be a better long-term investment.
To Conclude
For most small and mid-sized companies, the Fivetran + dbt merger represents good news: a faster path to reliable, scalable analytics without overextending internal teams. But no single stack fits every situation.
Some organizations benefit most from a managed, ready-to-go ELT stack, while others need open or platform-native architectures to support complex integrations or governance models. The key is balancing simplicity with sustainability.
Distillery helps teams design and optimize modern data ecosystems, whether that means standing up a lightweight Fivetran + dbt implementation, modernizing open-source pipelines, or building lakehouse-native patterns on Databricks or Snowflake.
As data vendors continue to consolidate, it’s important for companies to evaluate how these shifts could impact long-term flexibility, scalability, and budget planning. Distillery helps teams stay ahead of these changes, designing data architectures that perform efficiently today while remaining adaptable for what’s next.
Ready to evaluate your next stack move?
Distillery’s data experts can help assess your current architecture, model future-state options, and create a roadmap that fits your team’s goals, scale, and budget. Ready to modernize your data stack? Get in touch with our team.