Sell Data to AI Companies: Customer Data or Operational Data?
Sell Data to AI Companies: Customer Data or Operational Data?
An owner proudly shows an AI company a list of 50,000 customers.
The AI company says thank you, then asks for something else: how does your team get the work done?
So when you sell data to AI companies, which do they want, customer data or operational data?
The video
Sell Data to AI Companies: Customer Data or Operational Data?
Full script
An owner proudly shows an AI company a list of 50,000 customers.
The AI company says thank you, then asks for something else: how does your team get the work done?
So when you sell data to AI companies, which do they want, customer data or operational data?
Your company holds 2 kinds of data, and AI companies mostly pay for only one of them.
Most owners think the customer list is the prize.
It is worth the least.
AI companies pay for operational data, the trail of how your work gets done, day after day.
Lead with the list, and the first call with an AI lab can end right there.
This is general information, not legal advice.
So first, let us sort your data into 2 piles.
The first pile is customer data: names, emails, purchase history and your contact lists.
It tells you who your customers are.
The second pile is operational data: team email, support tickets, quotes, job files, approvals, schedules and chat.
It shows what your team did for those customers, and why.
Both piles live in the same systems, so most companies have never pulled them apart.
Why is a customer list worth so little to an AI lab?
First, an AI cannot learn a skill from a list of names.
A list shows no steps, no choices and no results.
Second, a list is full of personal data, and privacy laws often limit selling it.
One example is the Virginia Consumer Data Protection Act, in force since January 2023.
It says that trading personal data for money is a sale, and each person can opt out of that sale.
In plain words, for the companies it covers, selling your list means the people on it can say no.
Operational data is different, because it shows the steps.
Take one support ticket at a software company.
A customer writes in to say the invoice export is broken.
Your agent asks which version they use, checks the settings, finds the cause and sends the fix.
Then the customer says it works.
That is how the work gets done, from the request to the result.
Thousands of tickets like this one teach an AI model how a good agent thinks.
But your operational data still names customers on almost every page.
Here is an email from a plumbing company about a leaking water heater.
This email should not be licensed unless we de-identify it first.
That means we strip out the names, addresses and phone numbers.
A code like Customer 207 goes in their place.
The customer becomes a code, and the work stays: the leak, the new valve and the Friday visit.
So the email can still be licensed, and your customer stays private, within the privacy rules that apply to it.
There is one more line to draw, between your data and your clients' data.
Some files are in your hands only because a client trusted you with them.
Think of a law firm's case files, or a clinic's patient files.
Those usually cannot be licensed, because they belong to the client, or the law protects them.
Your own operations are different: how you schedule, quote, train staff and run the office.
That will not fit everyone.
A business whose main data is client files may have little of its own to license.
What does operational data look like in a real company?
A construction company has change orders, and the questions builders send in the middle of a job.
An accounting firm has month-end close checklists, with every step ticked and every fix noted.
A call center has call logs that show each question, each answer and how the call ended.
And a training school has its curriculum and how it runs, but not its student files.
In every one, AI labs pay for the work, not the names.
In short, AI companies pay for how your work gets done, not for who your customers are.
Tonight, write down your systems and mark the ones that show the work.
That leaves one question about your own company.
Which of my data could an AI lab license?
Answering that for a real company is my job, so 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.
In a data deal, I come in as your broker and technical advisor.
As your broker, I help you create relationships with AI labs and bring back their offers, so you can compare price, terms and how each AI lab may use your data.
I work through the terms with you, and you make the final call.
As your technical advisor, I guide your team through what to export, the format, the de-identification and compliance.
We move forward only after your team signs off.
I work on commission only, paid by the AI lab.
Already shopping your data around? Contact us anyway. We can still work through the terms with you and 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, so you can focus on AI research.
So the customer list stays with you, and the way your team works is what you can license.
More on licensing your data at AITrainingDataBroker.com.
Thank you for watching.
I will see you next time.