7 Data Monetization Strategies: How to Sell Data
AITrainingDataBroker.com
7 data monetization
strategies
From the least effort to the newest buyers
Use it, sell insights, marketplaces, data products, AI labs, closing sales, made to order
Jimmy Lin
7 data monetization strategies
A company has 7 main ways to make money from its data, which is what people call data monetization.
Some take a week, and some take a whole team.
So who buys data, and how do you sell it?
The video
Full script
A company has 7 main ways to make money from its data, which is what people call data monetization.
Some take a week, and some take a whole team.
So who buys data, and how do you sell it?
To answer that, start with Barbara Wixom and Jeanne Ross, 2 researchers at MIT.
In January 2017, they sorted the ways a company earns from data into 3 roads.
A company can use data to improve its own work, add data to what it already sells, or sell the data itself.
Since then, the selling road has grown, and it has new buyers, such as AI labs.
So I put 7 strategies on one staircase, from the least effort at the bottom to the newest kind of deal at the top.
This is general information, not legal advice.
Most owners think monetizing data means selling a customer list.
It does not.
In fact, step 1 sells no data at all.
Then what does step 1 do?
You use the data inside your company, to make better decisions.
Take a delivery company, just as an example, that studies a year of its own trip logs.
It finds that 1 route is always late, so it moves 2 stops, and the drivers save time every day.
No one buys anything, but the company still earns from its data, in hours saved and fewer mistakes.
That is the base of every other step.
If your own team can not find the data or trust it, no one outside will pay for it.
Once it is clean, you can sell what it shows.
That is the second strategy: you sell insights or reports, which means you sell what the data shows, not the data itself.
For example, a chain of hardware stores could sell a monthly report on which tools sell best in each region.
Suppliers, investors and trade groups often pay for that kind of view.
But sell the pattern, not the people.
In February 2024, the US Federal Trade Commission ordered Avast, an antivirus maker, to pay $16.5 million.
The agency said Avast sold people's web browsing data after promising to protect it, and the order stopped those sales for ads.
So a report should hold totals and trends, with no personal details in it.
Next, the third strategy takes more work, because you list a dataset on a data marketplace, an online store where companies shop for data.
The store brings the companies that buy data to you, and it usually keeps a share of each sale.
In return, your dataset needs a clear description, a sample, a price and regular updates.
Europe has a rule for these stores, in its Data Governance Act, which applies from September 24, 2023.
Its Article 12 says such a store may use your data only to pass it on to its users, not for its own goals.
Still, this will not fit every company.
A marketplace works best for data that many companies want in the same shape, like prices or locations.
So what if your data is one of a kind?
Then the fourth strategy fits better: you turn it into a data product.
Often that is an API, which is a live feed that other software connects to.
For example, a freight company could sell live wait times at its docks to shipping apps.
Customers pay every month, so the money keeps coming in, as long as the feed keeps running.
But a product needs people behind it, to fix the feed at night, update it and answer customers.
That is real work, so many companies start lower on the staircase.
And the biggest change is on the next step.
That change is the fifth strategy: you license company data to AI labs.
AI labs use it to train their models.
They look for the work your team already does, like support tickets, team chat and design files, made by real people.
In May 2025, the US Copyright Office reported that licensing deals for AI training are already being made.
It advised letting that market keep growing on its own.
Price follows supply and demand.
The kinds of data AI labs need most change during the year, and data that is in short supply gets a better price.
Before anything goes out, your data is de-identified, which means the names are removed, so it stays within the privacy rules that apply to it.
A license also lets you keep using your own data.
But what happens to the data when a company closes?
It can be sold as an asset, like the desks and the trucks, and that is the sixth strategy.
In August 2026, Reuters reported that Google was the winning bidder, at $10 million, for the business data of Spirit Airlines in its bankruptcy.
According to The Register, the package held more than 30 million recorded customer service calls and 600,000 IT tickets.
A small company that closes will not see a number like that, but its emails, chats and tickets can still have value.
The catch is timing, because the data has to be exported before the accounts are shut down.
Finally, the seventh strategy is the newest: you make data to order.
An AI lab asks for something that does not exist yet, such as an RL environment.
That is a practice world where AI learns a job by trial and error, like a delivery robot learning a city.
It can also be a task set: hundreds of real jobs, each with the right answer.
A model is then tested on every one of them.
The best ones are seeded with real data, like a delivery company's real routes and delays.
It is the most work on the staircase, but it pays for what your team knows, not only what it stored.
So which step fits your own company?
To help you answer that, let me tell you who I am.
I'm Jimmy. I have worked in data at Fortune 500 companies, as a data analyst, a data engineer and a software engineer, and I am trained in data science. I am also a Canadian military veteran, specialized in data analysis and reporting.
Steps 5, 6 and 7 are where I come in, as your broker and technical advisor.
I take a de-identified sample to several AI labs and bring back their offers, so you can compare them side by side.
I work through the terms with you, and you make the final call.
I also guide your team through the export, the format, the de-identification and compliance.
AI companies pay my fee.
Already shopping your data around? Contact us anyway, and I can still help you negotiate a better price.
If you are buying data for an AI lab, I find the right data and see it through to delivery.
In short, start on the step your data is ready for, and climb once it is clean and you know who wants it.
More on selling data to AI labs at AITrainingDataBroker.com.
Thank you for watching.
I will see you next time.
More slides
AITrainingDataBroker.com
Should I use
a data broker?
Pros & cons
From recent deals, names removed
Inside the deal, the honest cons, when to go alone
Jimmy Lin
AITrainingDataBroker.com
How to sell
company data
to AI companies
A guide for business owners
What they buy · who buys · direct or broker · what to check
Jimmy Lin
AITrainingDataBroker.com
Buy/Sell/License
AI Data Market
Update, Nov 5
Who is buying, what is urgent, and what to do today
Jimmy Lin
