Salesforce has agreed to buy data management firm Informatica for $8 billion in a highly strategic deal that gives it end-to-end data capabilities.

The deal, announced today, represents Salesforce’s second attempt to buy Informatica, after advanced conversations collapsed over terms in 2024. 

Informatica’s data catalog, data integration, governance, quality and privacy, metadata and master data management combined with Salesforce will “unlock new capabilities for Salesforce’s enterprise data stack,” it said. 

The cash and debt acquisition at $25/share comes 20 days after Informatica reported Q1 ARR of $1.7 billion with cloud subscriptions up to 50% of ARR.

(It stopped reporting cloud renewals however amid some churn. CFO Mike McLoughlin would only tell analysts that "we're very much on track with the operational and systematic changes we made to make sure that that cloud renewal rate is solid and gets even better in '26 and beyond...")

Informatica has been nudging customers from on-premise deployments to SaaS; – those making that shift have included insurance giant Zurich – and boasted of processing “over 109 trillion cloud transactions” in March. 

It also recently announced the general availability of its cloud data governance and catalog service natively on Google Cloud, letting customers deploy committed GCP spend for its data management capabilities.

Why make the move?

A big focus of the announcement was AI: Salesforce said its ownership of Informatica would allow it to “establish a unified architecture for agentic AI.” 

Forrester VP Kate Leggett told The Stack: “Salesforce has made a hard pivot away from its traditional CRM products into Agentforce [its platform to build AI agents], and adoption of Agentforce has not met expectations due to missing features, inconsistent results, and inability to understand ROI.”

She added: “This is why Salesforce launched new pricing last week…  This deal would give Salesforce more leverage in multi-cloud and hybrid cloud deals — which is especially important in customer operations.” 

Satyen Sangani, CEO of Informatica rival Alation, noted: "Enterprises want and need to invest in AI. But it’s hard and often inaccurate. Most stop at customer support agents and chatbots that help customer support reps. To go higher up the value chain, you need accuracy.

Error-prone, hallucinating, stochastic LLMs have a hard time reading from deterministic, precise, structured data systems. LLMs... Fumble when interpreting 𝐬𝐭𝐫𝐮𝐜𝐭𝐮𝐫𝐞𝐝 𝐝𝐚𝐭𝐚; struggle to query 𝐜𝐨𝐦𝐩𝐥𝐞𝐱 𝐭𝐚𝐛𝐥𝐞𝐬; break when 𝐦𝐞𝐭𝐚𝐝𝐚𝐭𝐚 is missing or inconsistent... Well-governed metadata is the answer. And it’s been the answer to trust since the days of BI. Only, with humans in the loop, you could hack it. Autonomous AI means you can’t hack it anymore.

"The first rule of metadata is that you can’t talk about metadata. The best way to improve metadata is to actually do stuff with your data. It’s not to build a “better” catalog, but rather to just make metadata happen on the way to solving a business problem" Sangani concluded.

Mark Starobinsky, a former global lead on metadata and data lineage at Accenture, added: "Metadata is no longer just about lineage diagrams or cataloging at scale. It’s now an existential infrastructure layer for reasoning-capable AI... The next metadata company that wins will not merely index data—it will understand it. We’re talking about real-time trust arbitration, symbolic reasoning overlays, and AI-native compliance scaffolding. Not static catalogs. Living graphs."

Salesforce expects the deal to close in early 2027. 

The link has been copied!