Salesforce Buys Informatica in an $8B AI Move: Where’s Next?

Salesforce acquired data management firm Informatica on May 27, 2025, for a reported $8 billion. The deal was mooted last year but fell through – then the CRM giant decided to go all-in on AI agents. Now Informatica’s expertise in delivering quality data at AI-scale could be the catalyst for Salesforce’s renewal.

As Wall Street frets about the firm’s slowing growth, adding the world’s leading data-management software to its roster is a logical next step in the pivot to agentic AI.

Some analysts think AI agents will make SaaS software redundant. Has Salesforce bought itself time with the buyout? We look at the deal and its implications.

Key Takeaways

Salesforce, the world’s second-largest SaaS brand by market cap, bought data management and governance firm Informatica for $8 billion.
The acquisition fills a much-needed gap in core technical capabilities that Salesforce needs to complete its company-wide shift to AI agents.
Agentic AI’s rapid advance has caught many cloud software companies off guard, and leading VC firms have warned that agents could replace much of what SaaS software does at a much lower cost.
Generative AI tools like agents require high volumes of reliable, high-quality, and trustworthy data to operate effectively.

Table of Contents Table of Contents

Key Takeaways
Agentic AI: SaaS Killer?
The Quest for AI-Ready Data
AI Needs Data. Maybe Data Needs AI
The Bottom Line
FAQs
References

Table of Contents

Key Takeaways
Agentic AI: SaaS Killer?
The Quest for AI-Ready Data

Show Full Guide

AI Needs Data. Maybe Data Needs AI
The Bottom Line
FAQs
References

Agentic AI: SaaS Killer?

At the end of May 2025, cloud software giant Salesforce announced it had reached a deal to buy global data management firm Informatica for $8 billion. That’s about 30% less than what Informatica was worth 13 months ago when rumors of an incipient buyout started appearing in the financial press. Both firms have seen slumping growth in the interim, which likely brought them back to the bargaining table.

A macroeconomic backdrop of trade uncertainty and recession risk has big corporate tech customers rethinking IT spending plans across the board. But Salesforce faces a particular challenge.

The rise of AI agents – mini programs that can handle sophisticated computing tasks on their own – threatens the core SaaS business model.

Experts say agents could deliver a lot of what Salesforce’s CRM and e-commerce platform does for a lower cost, so the company has made a rapid pivot, offering customers easy-to-configure AI agents of its own.

That’s why investors and tech analysts like the Informatica deal. Informatica’s market-leading data-management platform could help Salesforce strengthen its Agentforce agentic AI offering and move its strategic repositioning forward.

The Quest for AI-Ready Data

Software firms selling generative AI tools have seized on agents as a way to improve sales. Their practical utility is obvious, and that makes it easier for corporate IT buyers to make a business case for investment.

But – AI agents need strong data management capabilities to operate effectively.

A note to investors from Deutsche Bank last week concluded that without a robust capability to manage data quality and governance – something Informatica’s Intelligent Data Management Cloud (IDMC) promises – “the utility of what Salesforce AI can deliver will be limited.”

Informatica and Salesforce today announced that the companies have entered into a definitive agreement for Salesforce to acquire Informatica.

Learn more: https://t.co/HJuNWFwN40 pic.twitter.com/AYRuux4Mgd

— Informatica (@Informatica) May 27, 2025

The inexorable link between data and software efficacy has been well understood for decades. The old adage “garbage in, garbage out” applies just as much to ChatGPT as it does to Excel spreadsheets.

Dodgy data causes the hallucinations and other odd behaviors that routinely plague large language models (LLMs). If the data they draw from is flawed, AI agents that rely on those models to operate are just as likely to be affected.

In a recent blog, tech analyst Joseph Ours wrote that there’s a common assumption that corporate data is AI-ready, “but that’s not true. No one has truly AI-ready data – at least not yet.”

Companies have traditionally built up databases to address operational needs or to feed into analytics tools run by humans. “This often leads to limited and gap-filled datasets,” he notes. “They might be rich in specific operational aspects but missing other potential dimensions.”

Corporate data repositories contain duplicates, empty fields, old information, and files in non-standard formats. Sometimes information has an uncertain lineage or might have been gathered without regard for privacy regulations.

In the agentic-AI world Salesforce wants to dominate, bad data escalates quickly from IT nuisance to major business risk.

AI Needs Data. Maybe Data Needs AI

Firms have been talking about making business decisions insight-driven since the dawn of big data, so why is data quality still an issue?

George Johnston, a Partner at Deloitte UK and the consultancy’s Data Risk and Analytics lead, told Techopedia there are four key reasons: surging data volumes, data held in siloes, out of date systems, and lack of data management skill sets.

He said:

“The sheer volume and variety of data formats available today make it difficult to manage, clean, and maintain. Data comes from diverse sources and systems. Managing it often involves resolving inconsistencies, dealing with missing values, and ensuring data compatibility.”

Graph depicting AI data growth from 2021 to a forecast of 180 ZB by 2025, showcasing a steep upward trend. — Data growth is estimated to reach 180 zettabytes by 2025. Source: Tierpoint

Many organizations, Johnston says, rely on outdated systems that were not designed to maintain data quality at the scale and complexity AI demands. They might also lack the in-house expertise to deal with data strategically.

That’s where the Informatica acquisition looks like a game changer for Salesforce. The company’s IDMC platform handles the time- and labor-intensive tasks around data quality like cleaning, extracting, integrating, cataloging, labeling, and securing it.

It has a built-in AI engine of its own called CLAIRE that automates nit-picky data management processes at scale, applying quality and governance rules to billions of data points.

That’s a much better match for the intense data ingestion and computational demands of AI agents:

A study by Google-backed AI startup Vectara found the frequency of GenAI hallucinations ranging from three percent for ChatGPT, 5% for Meta’s AI systems, and 8% for Anthropic’s Claude 2. Google’s PaLM came in highest at almost 30%.
In a 2024 survey of data science professionals, 90% said that their models display signs of hallucination, and that 93% of MLOps professionals experience issues on a daily or weekly basis.

The Bottom Line

It’s been nine months since Salesforce launched Agentforce. In its Q4 2024 earnings call, CEO Marc Benioff said the firm had already closed more than 3,000 “paid deals” for the service, though analysts wonder if the firm has slashed prices in a bid to win more users.

The company reported around $900 million in AI-related annual revenue for fiscal 2024, just under 3% of its total revenue for the period.

Data quality determines the reliability of AI. If you can’t trust the content of all the strings, floats, bools, chars, enums, and arrays you’re feeding into a large language model, you can’t be certain about the actions an AI agent might take on your behalf. For now, Informatica’s data-management software promises to be a key ingredient in Salesforce’s strategic AI pivot.