See what your data could be worth
See what your data could be worth. Most companies are sitting on $25K to $5M+.
Find out in 60 seconds. Nine questions, no commitment, and the estimate stays private unless you choose to send it to us.
Half of what makes data valuable is what your people did with it: the steps they took, the judgment calls, the outcome. Workflow data is worth more than raw records, and almost every operating company has it.
Messages
Support threads, chats, call transcripts
Codebases
Repos, reviews, incident write-ups
Documents
Contracts, reports, email, filings
Raw records
Tables, events, outcomes, telemetry
Task logs
Who did what, in what order, and why
Describe the dataset
No free text, nothing is sent until you choose to.
Ballpark license value
$100K – $500K
Licensable
Clear buyer demand from data vendors and labs building niche eval or fine-tuning sets.
Rubric score
69 / 100
Modality multiplier
×1.0
Why this number
- Uniqueness / scarcity×214/20
Semi-proprietary; a few comparable sources likely exist.
- Rights cleanliness×218/20
Clear ownership — this is what buyers want to see.
- Volume / scale6/10
Meaningful for fine-tuning and evals.
- Freshness / recurrence10/10
Could be sold as a recurring feed, not just an archive.
- Structure / readiness5/10
Some packaging work before buyers can evaluate it.
- Exclusivity available4/10
Non-exclusive is fine, but caps the top of the range.
- Buyer substitutes (inverse)7/10
Substitutable, but at several times the cost.
- Cost to replicate5/10
Expensive but possible to recreate.
Bands are deliberately conservative floors. Competitive processes with lab buyers have cleared 2–5× the rubric band. This is not an offer or appraisal.
How the estimate works
Each dataset is scored out of 100. Uniqueness and rights cleanliness are double-weighted because they decide whether a deal happens at all; the remaining six factors set the price within the band. Scores under 40 are usually a rights problem, not a value problem, and are often fixable.
The band is then adjusted by modality: images are abundant (×0.8), code and media command a premium (×1.2–1.5), and expert workflows with outcomes attached are the scarcest data labs are buying today (×2.5).
The rubric is calibrated to be a conservative floor. Publicly reported deals — airline data archives, publisher back catalogs, UGC platform licenses — have cleared well above their rubric band when more than one buyer was at the table. We price the floor with the rubric and let the process find the ceiling.
We only take clean-rights data to market. If your rights basis is unclear, the first thing we do together under the partnership is a rights audit — before any buyer hears your name.