Data (Quality) Is the New Oil

The acquisition announcement probably didn’t seem like much to most investors.

All things considered, acquiring Informatica (INFA) for $8 billion is a small deal for a company the size of Salesforce (CRM).

Salesforce is worth $245 billion and is sitting on $17.4 billion in cash. It will generate $41 billion in revenue this fiscal year… and about $13.7 billion in free cash flow during the same period.

An $8 billion acquisition is breakfast.

But what makes the acquisition interesting is the underlying technology involved, the history of Informatica’s software sector, and what the acquisition represents…

The Push for Cloud Subscriptions

A decade ago, companies that developed software for data integration, data quality, and data governance were about as boring as they come. Important, but difficult for the market to get excited about.

They were largely perpetual licensing business models, which meant an up-front fee with ongoing maintenance and support fees. These companies were slow to make the migration to cloud-based subscription models, and their valuations suffered because of that.

Cloud-based subscription business models are more desirable due to their much-improved gross margins and free cash flow, resulting in higher valuation multiples.

And that’s what made these data software companies a particularly attractive target for private equity.

The list is too long… the number of software-as-a-service (SaaS) companies that have made or are in the process of making that transition. From a perpetual licensing model to a recurring cloud-based subscription model. It’s painful, takes years, and – given enough time and good execution – it always pays off.

That’s why private equity likes it. Simple to understand. Produces consistent returns. Always works when executed well.

I remember in 2015 when Permira stepped up to acquire Informatica, in partnership with the Canada Pension Plan, for $5.3 billion. The deal went through at roughly 4.9 times Informatica’s forecasted 2015 sales.

Again, it wasn’t an interesting acquisition at all. It was clear what Permira’s game would be: to push the migration to recurring cloud subscription revenues… and either bring the company public again or sell it off at a higher valuation.

So it was no surprise when Thoma Bravo stepped up to make another boring acquisition of a data quality and data integration company, Qlik, in 2016 for about $3 billion.

And very similar to the Permira/Informatica deal, Thoma Bravo picked up Qlik for about 4.2 times Qlik’s forecasted 2016 sales.

If that wasn’t enough, several years later, Bain Capital stepped up and made a $150 million investment into the private data management and data quality company Ataccama in 2022. I don’t know what the valuation metric was for the deal, but I’d be surprised if it wasn’t too far off the first two deals, as 2022 was a big down year in valuations for private tech companies.

At face value, these three deals don’t look very interesting. And they actually weren’t.

The investment thesis behind the deals above was all about changing the business model so that the asset could be sold at a higher valuation.

And then, artificial intelligence came along…

The Big AI “Data Problem”

The boom in AI – and training frontier AI models – has changed everything.

But it didn’t happen overnight…

In 2023 and 2024, the major players building frontier models were pumping as much data as they could get their hands on into the training of their AI models.

It was the brute force method. It was effective, which is why they just kept pushing for “more” data… but it had a downside: Low-quality data resulted in bias, incorrect answers, less accurate results, and the dreaded hallucinations.

It wasn’t until late 2024 and into this year that data quality became a point of differentiation in the training of AI models. I would argue (in early 2024) that it became a necessity.

xAI and its sudden ascent to the top of the AI leaderboards were in part due to its fanaticism on data quality as a key means to creating a “maximum truth-seeking AI.” It worked, and it produced a higher-performing artificial intelligence.

xAI’s success forced the rest of the industry to focus more on data quality.

It’s no longer just about volume – it’s about the quality of the data, and whether it is real-world data or synthetic data, it is critically important to improve AI model performance and eventually achieve artificial general intelligence.

As a result, a lot more money will be spent on the collection of high-quality real-world data, the production of high-quality synthetic data, and, of course, the data software to govern and manage that data.

So now we understand why Salesforce was interested in acquiring Informatica.

After all, Salesforce was fairly early into leveraging artificial intelligence in enterprise software. And Salesforce maintains what is probably the world’s largest collective enterprise database. But it was missing the data quality and data governance software component that Informatica provides.

And Salesforce is getting Informatica at an attractive price, representing about 4.7 times forecasted 2025 sales. That’s very close to a multiple of a decade ago.

What is more interesting, however, is that Salesforce made a run at Informatica last April… when the stock was trading around $38 a share.

Permira Got Greedy

Stupidly, Permira got greedy last April and wanted too much for Informatica. What happened next was ugly.

Informatica (INFA) shares plummeted almost 60% between April 2024 through April 2025.

The drop, and the fact that Informatica has only been able to migrate about 50% of its revenues over to recurring cloud subscription revenue (shown in light blue in the chart below), is what enabled Salesforce to get such a great price for Informatica.

Informatica’s Quarterly Revenues | Source: Informatica

Permira stuffed this one up, assuming the deal goes through.

Permira has been invested in Informatica for a decade, had to pay interest on the debt that it used to fund the deal, and also suffered dilution from a $408 million secondary offering late last year to raise cash.

Had it sold at the highs after the IPO in 2021 or last April, it probably would have gotten out of the asset in the black, but not now.

And that’s great news for Salesforce…

The big story, and the one that’s interesting, is the sudden shift by the industry to focus on data quality and governance.

Not only will AI firms need “more” high-quality data, they’ll also need the tools to manage and govern that data.

I do expect to see more mergers and acquisition action like this one as a result…

Because Salesforce is making a big push into agentic AI… and they’re not the only ones.

That’s what this deal is all about.

Salesforce’s Agentforce product is designed to automate complex workflows with a contextual understanding. And the end goal is for agentic AIs to be able to make real-time, accurate decisions in a completely autonomous way.

They can’t do that without accurate data.

Data (quality) is the new oil.

Jeff