How to Sell Data as a Business: Routes, Rules, Pricing

To sell data as a business, choose one of three routes (a direct deal, a sale to a broker or other buyer, or a marketplace listing), confirm you have the right to license the data, and settle privacy duties before you sign anything. If the data includes personal information about California consumers, you must notify them that it may be sold and that they can opt out, and you must stop selling a consumer's data no later than 15 business days after an opt-out request.

This is general information, not legal advice. Have a lawyer read your specific facts before you sign.

What are the three ways to sell data as a business?

You can sell through direct deals, through brokers and other buyers, or through data marketplaces. Each route changes who does the work of finding the buyer, who writes the contract, and how much you can negotiate.

1. Direct deals

You find a buyer, agree on scope and price, and sign a license or sale agreement. You control the terms, but you also carry the sales effort and the legal drafting.

If you do not want to build a buyer list yourself, FileYield lets you list a dataset for free in about two minutes with its AI listing assistant. It needs no file upload, so the data stays with you. FileYield runs buyer outreach for listings, and buyer and seller negotiate price, license, exclusivity and delivery directly.

2. Selling to data brokers or other buyers

Here you hand data to a company that resells it or uses it in its own products. Start with our list of companies that buy data for AI and ask each one what it requires from sellers. Onboarding steps vary by buyer, so get them in writing.

Be aware that California treats some sellers as data brokers. We cover that test below.

3. Data marketplaces

A marketplace lists your dataset where buyers already shop. Fees, packaging rules and license templates differ from one marketplace to the next, so read each set of seller terms before you apply. Our AI data marketplace guide compares nine of them, and how we pick the marketplaces explains our criteria.

What should you verify before you sell data?

Run three checks before any buyer sees a sample: rights, consent and de-identification. Skipping one can stall a deal late or expose you after it closes.

1. Rights. List who created or collected each part of the dataset: employees, contractors, customers, partners, or public sources. Then read every contract and terms of service attached to that data for limits on resale or reuse. Whether you can license it depends on those documents, so have counsel confirm.
2. Consent and notice. Check what you told people when you collected their information, and whether a sale fits that notice. California adds specific notice and opt-out duties, covered below.
3. De-identification. Decide which fields the buyer needs, and remove or transform the rest.

Write down the result of each check and keep it in a file you can hand to a buyer's legal team. Buyers often ask for it.

Is it legal to sell personal data in the United States?

We cannot give you one nationwide answer. The primary sources we rely on here cover California, and they do not establish a single federal rule for every data sale.

Whether a transfer counts as a "sale" or "sharing," and which notices and contracts apply, depends on the data, the recipient, the purpose, your relationship with the people involved, and the state. If your data covers people in several states, check each state's rules.

What does California require if you sell personal information?

California gives consumers a right to direct a business not to sell or share their personal information. A business that sells or shares it must tell consumers so and tell them they can opt out.

The main duties for a seller are below.

One draft CPPA text says businesses need two or more designated methods for opt-out requests, and describes opt-in for sale or sharing as a two-step process. It is a draft, so check the final regulation and its effective date before you build around it.

Are you a data broker under California law?

You may be. California defines a data broker for registration purposes as a business that knowingly collects and sells a consumer's personal information to third parties when it has no direct relationship with that consumer.

If you collect data directly from your own customers, you may not fit that definition for that data. If you resell data about people you have never dealt with, you may. Confirm with counsel.

If you do qualify, the rules add up quickly:

• You must register annually, starting in the year after you began brokering California residents' data.
• Registration requires business information and details about the data you collect and who receives it.
• California residents can use DROP to send one deletion request to all active data brokers beginning January 1, 2026.
• Beginning August 1, 2026, brokers must access the state's deletion mechanism at least once every 45 days and process covered deletion requests, subject to limited exceptions.

The registry notes that brokers may handle highly sensitive categories such as Social Security numbers, precise geolocation, health-related information and browsing history. If your dataset includes any of these, flag it early to every buyer and to your lawyer.

How do you handle consent and opt-outs in practice?

Treat opt-outs as an operating process, not a one-time form. A workable setup looks like this:

1. Add the required homepage link and notice language to your site.
2. Route every opt-out request to one owner, and log the date you received it.
3. Remove that person's records from future deliveries within 15 business days.
4. Tell buyers in your contract how you will pass opt-out and deletion requests to them. The contract terms are up to you and the buyer.
5. Do not add the person back unless they later authorize it.

If you sell updates on a schedule, build the opt-out check into each delivery.

How do you de-identify data before selling it?

Start by cutting the dataset down to the fields the buyer's use requires. Then document what you removed or changed, and keep that record with your rights file.

Whether a particular method counts as de-identified depends on the law that applies to your data, so do not assume one method fits every case. Our data anonymization guide walks through the options, and health data holders can read whether providers need consent to license de-identified data.

If you list on FileYield, its listing flow lets you flag personal data (PII) and compliance requirements early, so buyers see them before talks begin.

How do pricing models and licensing terms work?

Price and license terms are negotiated per deal. We cannot cite a standard price list, per-record rate or typical earnings figure, and any source that quotes one without context deserves caution.

What moves the price is the dataset itself and the deal around it:

• Quality, uniqueness and freshness of the data.
• Volume.
• How clean your rights are.
• Which uses the buyer is permitted.
• Whether the buyer gets exclusivity.
• The support you provide.
• How many buyers want it.

The license should spell out four things:

1. Price and payment structure. How much, and when it is paid.
2. License scope. What the buyer may do with the data, and what it may not.
3. Exclusivity. Whether you can sell the same data to others.
4. Delivery. How and when the data reaches the buyer.

To size your ask, read our data valuation methods and real deal prices, then compare with AI data licensing deals and what they paid. For clause-level detail, see how AI data licensing deals and terms work.

What are the steps to sell your data?

1. Inventory the dataset and where each part came from.
2. Confirm your right to license it, and write down the basis.
3. Decide whether it includes personal information, and which rules apply.
4. Set up notices, the homepage links and opt-out handling if California consumers are involved.
5. De-identify or trim fields the buyer does not need, and record what you changed.
6. Choose your route: direct, broker or buyer, or marketplace.
7. Negotiate price, scope, exclusivity and delivery, and get them in a signed contract.

When your paperwork is ready, browse the directory of active buyers on our homepage: sell data to AI companies.