The Data Monetisation Guide· Part 1
Who captures the value
Owning data and being paid for what it is worth are two different things. What the deals show about who ends up with the money.
By Gianluca Carrera · 4 min read · Last reviewed 23 October 2026
Holding valuable data does not decide who is paid what it is worth. The party that turns the data into an answer a buyer pays for usually ends up with most of the money, whoever owns the records.
In short. Two businesses can hold equally valuable data and keep very different shares of what it earns. The deals in the register show the pattern: some data owners do the work on their data themselves and sell the answer; others hand the work to a partner and are paid for their records. The share a business keeps as the value is created has a name, the rake, and the gap between what it could keep and what it does keep has another, under-capture.
Who captures the value of data?
Usually the party that does the enrichment: the work, described on the previous page, that produces something the data does not state on its own. Owning the data decides who may sell it. It does not decide who learns what it is worth.
What does it look like when the data owner does the work?
The register's clearest examples are businesses that pulled the work inside rather than handing their data to a partner. Citi launched its commerce media business in September 2026 on 6.5 billion card transactions a year, and in August agreed to acquire Kard, a commerce media platform, to bring the targeting and measurement in-house. Brands buy placements, not the transaction file. McDonald's began piloting its own media network in August 2026 in 450 company-owned restaurants, with a stated ambition of a $1 billion business built on a customer base it puts at about 70 million a day; advertisers buy reach, and the visitor and transaction data stays with McDonald's.
Circana shows the same thing from a different angle. Its Liquid Mix product, launched in September 2026, runs Google's marketing-mix code, free since February 2025, on Circana's own point-of-sale data and household panel. Google contributes the code and takes no disclosed share. Circana does the work on its own data and sells advertisers the answer.
What happens when someone else does the work?
The value tends to move to them, often without anyone deciding that it should.
Customers Bank, a $25.9 billion US lender, signed a multiyear partnership with OpenAI in April 2026 to automate lending, deposits and payments, with OpenAI engineers working inside the bank. The agreement also has the two co-creating tools OpenAI could sell to other banks. Universal Music licensed its catalogue to ElevenLabs in September 2026 for artist tools and a fan platform that ElevenLabs launches and runs. PitchBook became a data partner for OpenAI's assistant for financial services in September 2026: its records are queryable inside the assistant, while the analyst's workspace and the enterprise contract sit with OpenAI. TriNetX licensed Regeneron access to de-identified health data on about 300 million people in April 2026, and Regeneron does the discovery work on the result.
None of these deals is a mistake on its face. Each data holder gets something real, or has been promised it. The question each of them answers, often without asking it, is who will know what the answer is worth in five years.
What is the rake?
The rake is the share of the value created that a business keeps as that value is created, rather than once at signature. It is the measure that tells two data businesses with the same data apart.
What is under-capture?
Under-capture is the gap between what a data business could keep and what it does keep, deal after deal. It is rarely lost in the negotiation. It is usually settled before anyone sits down to negotiate.
Why that happens, and the decisions that settle it, are the subject of my book, Value Capture: The Hidden Economics of Data Businesses, out in 2027. More about the book.
Questions people ask
Who owns the value of data? Owning the data decides who may sell it. The party that turns it into an answer a buyer pays for is usually the one that learns what it is worth and keeps most of the money.
What is the rake in data monetisation? The share of the value created that a business keeps as the value is created, rather than once at signature.
What is under-capture? The gap between what a data business could keep from its data and what it does keep, usually settled before the negotiation starts.
Where can I see how my business stands? The Data Monetisation Readiness Assessment on Monetaise is a structured first look.
Next in the guide: How data is priced, the pricing models data sellers use and what moves the price. Every deal cited here is in the register, each with its reading.