Skip to content

How much is my company's data worth?

Short answer

It depends on the record, not the size of the files. Years of continuous history, records that tie decisions to outcomes and clean rights raise the value; short history and data you don't own cap it. The only number worth trusting comes from buyers who have reviewed your anonymized summary and made an offer.

Get your data pricedBook a call

Updated

Why file size doesn't set the price

Storage size says very little about value. A drive full of duplicate PDFs and a smaller set of linked records are two different assets, even if the drive is ten times the size. What makes the smaller set worth more is the links: each ticket tied to the change that closed it, each invoice to its approval and payment, each estimate to its photos, calls and final invoice.

Buyers count things that map to real work: how many jobs, tickets, matters or closes, how many years, how complete each record is, and how many have photos, recordings or linked documents attached. That's why our intake asks for counts and dates, not gigabytes.

What raises the value

Nobody can hand you a formula, but these are the things that shape it.

  • Long, continuous history. Ten years in the same system with no gap shows how your business behaves through busy seasons, slow ones and price swings. A record that starts last spring can't.
  • Records that show decisions and outcomes. A quote, the revision and the approval or the loss. A claim, the supplement and the adjuster's answer. A ticket, the fix and whether it held. Outcomes are what let a model learn what worked.
  • Specialized work. Month-end closes, contract redlines, freight exceptions, insurance submissions, roofing claims and solar permitting are hard to find outside the companies that do the work.
  • Linked evidence. A ticket tied to the commit that fixed it, a call tied to the booking it produced, an estimate tied to the final cost, a photo tied to the job it documents.
  • Clean rights. You own the records, your contracts allow the use, your recordings were made with proper consent, and clients have consented wherever professional rules require it.
  • An ongoing feed. Some buyers want the record to keep growing, which a one-time export can't offer.
  • Exclusivity. A buyer who gets the record to themselves has a reason to pay more. See exclusive vs. non-exclusive licenses.

What holds it down

  • Short history. One or two years shows a moment, not a pattern.
  • Data you don't own. Measurement reports, supplier price lists, a client's books, code written for a client, an architect's drawings and MLS listings often belong to someone else or come with licenses that don't let you pass them on. We'd leave them out and license your own records and the fields you derived from them.
  • Personal information that can't be removed. If a record only makes sense with a customer's identity attached, it isn't something we can license. Identifiers have to come out before delivery.
  • Material that's already public. Permits on file with the city, public court filings, open-source code, your website, your published reviews. Buyers can get these without you.
  • Gaps and scatter. Records split across three systems with years missing in between are worth less than one complete run.

Most of these change what goes into the license rather than ending the conversation. The rights side is covered in is it legal to sell company data.

Why competition matters for price

Ask one buyer what your record is worth and you get the number that works for them. You have nothing to compare it to, and they know it.

Buyers also value different parts of the same record. A voice AI company cares most about calls. An insurtech company cares about claim files and outcomes. A legal AI company wants redlines, a coding-tools company wants code linked to tickets, and a data lab and an AI lab each look for something else. When they all see the same anonymized summary at the same time, they price against each other, and each one shows you what it values by what it offers.

A marketplace listing waits for a buyer to find it and usually leaves you to set the ask. We take the summary to buyers and ask them to name theirs.

Why we don't quote a number first

After your intake we can tell you what looks strong and what looks thin. We won't tell you what it will fetch. That would be a guess dressed up as an estimate, and an early number tends to anchor everyone, you included.

Pricing happens when qualified buyers, under NDA, review the anonymized summary and put offers in writing. Until then, the honest answer is that the price isn't set. We are paid a small commission out of a deal that closes. If nothing closes, you owe nothing.

What the market has shown

The four licenses closed by buyers in our network that we show on the case studies page fall in the $110K–$450K range. All four were exclusive licenses to the company's operating record: one perpetual, two for 24 months and one for 18 months.

Those figures describe the market. They are not deals we brokered, and they are not a forecast of what your record will bring. Every record is priced on its own.

Follow-up questions

No. After your intake we can say what looks strong and what looks thin, but we don't put a price on a record before qualified buyers have reviewed the anonymized summary and made offers. Anything earlier would be a guess.

No. More years of continuous history and more fully documented jobs, tickets or matters help. More copies of the same files, or more records from one short window, add little. Buyers look at counts of real work, years and how complete each record is, not gigabytes.

No. Length and continuity of history matter more than company size. A small firm with a decade of clean records can have a stronger record than a larger company with two years of patchy data.

This is general information about how data licensing works. It is not legal advice, and we are not your lawyer. Have your own counsel review any license before you sign.

Keep reading on licensing your data.

What happens in a data licensing deal, step by step?

You give a buyer permission to use a defined slice of your records, for a defined purpose and term, and you keep ownership. With us it runs in seven stages: intake, review, market, offers, contract, delivery and paid. You decide at the offers stage, and you owe nothing if no deal closes.

What's the difference between exclusive and non-exclusive data licenses?

An exclusive license gives one buyer your record to themselves for a defined scope and time, and usually pays more. A non-exclusive license pays less per deal but leaves you free to license again. We collect bids both ways so you can compare before you decide.

What industries do AI companies buy data from?

Most industries, if the company has years of records from real work done in business software. The kind of work matters more than the sector: finance and accounting, legal and contracts, software and IT, support, sales and operations. Some industries need consent or compliance work before anything is licensed, and a few aren't a fit at all.

Who buys company data for AI?

Five kinds of buyer: data labs, AI labs building agents and evaluations, vertical AI companies building for one industry, insurtech and claims companies, and voice AI companies. Each wants real operating records for a different reason, and finding and vetting individual companies is hard for all of them. We qualify buyers first, and they see an anonymized summary, never your name, until you sign.

Is it legal to sell company data?

Often yes, for a company's own operating records, when three things line up: the records are yours to license, your customer agreements and privacy notice allow de-identified use, and the people in them can be taken out first. This is general information, not legal advice, and your own counsel should read any license before you sign.