The PowerCenter Clock Has Run Out. Now Comes the Hard Part.

The Deadline Passed. Most Enterprises Missed It.

The deadline was not ambiguous. Informatica PowerCenter 10.5 standard support ended March 31, 2026. No patches. No bug fixes. No security updates — unless you pay for extended support through March 2027, then into "sustaining support" where even critical fixes stop.

After end-of-support, PowerCenter environments receive no technical support, security patches, or bug fixes. The lack of updates causes performance degradation and compatibility issues as surrounding technologies evolve, and unpatched vulnerabilities can leave data exposed. This is not theoretical risk. It is the operating condition today for every organization that has not yet migrated.

Extended support means hefty fees for diminishing service, and maintaining aging infrastructure while finding talent who know legacy ETL tools inflates operational costs.

The structural problem: PowerCenter was built for batch-oriented, on-premises data movement in an era of stable relational schemas. That world no longer exists at most enterprises.

Why Legacy ETL Cannot Be Patched Into Relevance

The issue is not support status but design philosophy. PowerCenter was designed for on-premises, batch-oriented data movement. In today's cloud-first, real-time world, legacy technology makes data teams less agile compared to competitors using modern, cloud-native tools.

The market has shifted to ELT — Extract, Load, Transform — where raw data lands in the warehouse first and transformations happen inside cloud compute. Modern cloud warehouses like Snowflake, BigQuery, and Redshift have enough compute power to handle transformations natively, making ELT faster, more flexible, and easier to maintain than traditional ETL.

The dbt Labs annual survey found that 65% of analytics engineers now transform data inside the warehouse via ELT rather than before loading via ETL. This is no longer a trend.

The winning tooling is straightforward. dbt is an open-source transformation tool that enables teams to transform raw data into analytics-ready datasets directly within a data warehouse using SQL and Python. It handles the "T" in modern ELT pipelines, allowing teams to build, test, and document data models in a scalable and version-controlled way.

On ingestion, Airbyte provides 600-plus connectors for APIs, databases, data warehouses, and data lakes, and also serves AI agents and LLM clients needing real-time access to business data via its Agent Engine. That matters: the same ingestion layer feeding a warehouse can now feed an LLM context window. PowerCenter cannot.

The Migration Options Are Not Equal

Enterprise teams facing the PowerCenter deadline have three paths.

Path 1: Informatica's cloud platform (IDMC). Migrating to Informatica Data Management Cloud is positioned as the natural evolution, but this transition is not straightforward. IDMC introduces a fundamentally different deployment model with new learning curves and potential trade-offs. The Salesforce acquisition adds uncertainty. With PowerCenter 10.5 end-of-support passed, the Salesforce acquisition has many data teams rethinking their platform strategy entirely.

Path 2: Modern open-core ELT stack. Airbyte for ingestion, dbt for transformation, Airflow or Dagster for orchestration, cloud warehouse as the execution layer. This is what greenfield teams start with today. Iterating in SQL with version control is faster than rebuilding brittle ETL flows, modern tooling bakes in lineage and testing, and separating ingestion from transformation lets teams optimize each layer independently. Onboarding costs are real; licensing costs near zero at the open-source tier.

Path 3: Extended support as permanent strategy. This is the most common path for organizations that missed the March deadline. It is not a strategy; it is avoidance. Continuing PowerCenter beyond standard support may feel easier, but it is not permanent. As surrounding technologies evolve, the platform becomes harder to maintain and a barrier to innovation, and most organizations eventually face higher costs and limited flexibility if migration is postponed.

The Market Numbers Reflect Switching, Not Growth

The global ETL market reached $10.24 billion in 2026 and is forecast to expand to $21.25 billion by 2031 at a 15.72% CAGR, according to Mordor Intelligence. That headline growth is real, but significant portions are not new pipeline demand — they are re-procurement. Enterprises migrating off deprecated tools are signing new contracts. The market looks healthy; the underlying dynamic is forced replacement.

Qlik's purchase of Talend for $2.4 billion reflects consolidation among legacy vendors. That acquisition was defensive. Talend's on-premises customer base was a migration liability; Qlik is attempting to hold those accounts through transition.

Fivetran's acquisition of Census in 2025 signals the strategic importance of reverse ETL, adding data activation to one of the largest data connector platforms. The pipeline layer is expanding downstream into operational data activation. Vendors without a reverse ETL story compete on shrinking surface area.

The Retraining Problem Is Real

The tooling decision and the people decision are inseparable. PowerCenter teams built institutional knowledge around GUI-based transformation logic, proprietary mapping syntax, and vendor-specific deployment patterns. None transfers to a dbt-plus-Airbyte stack.

Moving from ETL to ELT is not just a tooling change — it is an operating model shift. Data engineers need to write SQL-first models, manage version control, understand incremental materialization strategies, and reason about warehouse compute costs. These are learnable, but they demand time. For a mid-market team of three to five engineers, expect three to six months of reduced velocity.

The cleanest migrations ran parallel stacks for six to twelve months — standing up the new ELT pipeline incrementally while keeping PowerCenter stable — then cut over source by source. Worst outcomes came from lift-and-shift approaches.

What to Watch

1. Track PowerCenter customer choices between IDMC and third-party alternatives through Q3 2026. Informatica's renewal rates on extended support will signal whether Salesforce is retaining the installed base or accelerating defection.

2. Watch dbt Labs' enterprise contract velocity. The platform's 2025 Fusion Engine — a complete architectural redesign built on Rust — delivers faster parsing through optimized SQL comprehension and state-awareness that eliminates redundant warehouse queries. If enterprise adoption tracks product improvement, dbt Labs becomes an acquisition target within 18 months.

3. Monitor Airbyte's Agent Engine traction. Airbyte launched its Agent Engine in 2025 to power AI agent workflows. If agentic data access becomes a procurement requirement rather than an experiment, the connector layer becomes the most defensible position.

4. Watch for legacy vendor acquisitions of LLM-native startups. Any move by IBM, SAP, or post-acquisition Informatica to buy an AI-native pipeline company signals they have concluded organic development cannot close the architectural gap in time.

5. Measure actual migration completion against announced timelines. Most enterprises claiming 2025 completion did not finish. The extended support window through March 2027 is the real deadline. After that, sustaining support is survival mode.