Listen Labs Sold for 67x Revenue. Airtable Sold for 2.7x.

Salesforce just agreed to buy Listen Labs for a reported $2 billion. The company has roughly $30 million in annual recurring revenue. That's about 67 times revenue. Two months ago, Bending Spoons agreed to buy Airtable at about 2.7 times revenue. I sit on the board of Vurvey Labs, which works in the same space as Listen Labs, so I've been thinking about what the gap between those two numbers means.

My answer: AI-native companies are now valued on their data and their models more than their revenue.

The Listen Labs deal

On September 29, Salesforce signed a definitive agreement to acquire Listen Labs, an AI customer research and human simulation platform. Its AI agents design studies, interview customers, and synthesize what they hear, turning months of research into days. Salesforce plans to fold it into Agentforce, Marketing Cloud, and Service Cloud.

• Price: about $2 billion as reported (Salesforce didn't disclose it), on about $30 million in annualized revenue, according to TechCrunch.
• Multiple: roughly 67 times revenue.
• Background: Listen Labs was valued at $500 million in January 2026, then walked away from a signed $1.5 billion funding round to take the Salesforce deal.

The Airtable deal

On August 4, Bending Spoons agreed to acquire Airtable in an all-cash deal at an enterprise value of $1.285 billion. Airtable has about $480 million in annual recurring revenue, growing over 20% a year, and more than 500,000 organizations use it, including 80% of the Fortune 100.

That works out to about 2.7 times revenue. Airtable was valued at $11 billion in 2021.

Models and data vs. a database

Airtable is a good business. It's also a database with an interface designed for people clicking around a grid. As agents do more of the work, that interface matters less every quarter, and buyers price it that way.

Listen Labs is the opposite. Every interview its agents run adds to a body of data no one else has, and that data makes its models better at simulating how customers think. Revenue is a lagging signal. The data and the models are what Salesforce wanted.

If you're building now, that's the question to ask about your own company: what do you own that gets better every time a customer uses it? I wrote more about how buyers price companies in valuation is a range, not a number, and about the software side in Business.

Why this matters for Vurvey Labs

Vurvey Labs was founded in 2021 in Cincinnati. Its People Model is built on millions of real consumer interviews, and its AI agents let brands run research and test ideas against that human foundation, at a speed traditional market research can't match.

That's why I joined the board. The Listen Labs deal is another proof point that this category is valued on the data and the models.

• In November 2025, Havas led Vurvey's $8.5 million Series A, with committed client spend, a board seat, and integration across its Health and Creative networks.
• In April, TechCrunch named Vurvey among the most interesting startups showcased at Google Cloud Next 2026.
• In June, at Cannes Lions, Havas announced an increased investment in Vurvey Labs, just over €20 million, as part of its Converged.AI push.

The part I like most: Havas isn't just funding Vurvey, it's using the product. Havas's own Science of Desire study included 1,000 AI-powered interviews. An investor that is also a customer is the strongest signal a young company can get.

What I'd take from this

• If your product is mostly an interface for humans, expect buyers to price you on revenue, and expect that multiple to shrink.
• If every customer interaction makes your data and your models better, you're building the kind of asset buyers are paying 60x for.
• Revenue still matters. It just isn't the whole story anymore in AI-native categories.

To see how a buyer might price your company, try the free valuation tool. For more on the board and investing work I do, read why Outperform Ventures exists.