How this estimate works
A transparent scenario model — not an offer.
The estimate prices what it might cost a buyer to recreate your usable records, then applies hypothetical discounts and adjustments. It is not machine learning and has not been checked against real sale prices.
Model v0.1.0 · as of October 15, 2026 · evidence library 2026-10-15.2 · assumptions revision 0 (built-in defaults)
Saved assumptions unavailable; showing built-in defaults.
The math
- raw records = your record count × count-uncertainty (or, if unknown: employees × years × records per employee-year — low confidence)
- usable examples = raw records × usable fraction (by quality) × traceability adjustment [capped at 5,000,000]
- cost to recreate one example = BLS mean hourly wage (by category) × hours per example × compensation load × curation premium
- modeled price = usable examples × cost per example × buyer willingness-to-pay fraction × category × quality × distinctiveness × location × terms factors [capped at $5,000,000]
- 2,000 seeded scenarios sample every range uniformly; we show the 10th, 50th and 90th scenario percentiles, rounded to $500 / $1,000 / $5,000.
Evidence (quoted from sources)
Only the two BLS figures feed the calculation. The rest is context.
- Government statisticModel input
Occupational Employment and Wages, May 2025 — Table 1
U.S. Bureau of Labor Statistics · May 2025 estimates
Mean hourly wage: all occupations $33.54; customer service representatives $22.40; accountants and auditors $45.56; software developers $71.20; business and financial operations occupations $45.78.
How we use it: U.S. national means. Used as the hourly cost of a specialist who would recreate an example. Not a price for data.
- Government statisticModel input
Employer Costs for Employee Compensation — December 2025
U.S. Bureau of Labor Statistics · December 2025 (released March 2026)
Private industry: wages and salaries were 70.1% of compensation costs; benefits were 29.9%.
How we use it: Broad private-industry average, not specific to any profession. Used as a compensation-load factor (1 ÷ 0.701 ≈ 1.43).
- Broker-reported anecdoteContext only
AI's new training data: your old work Slacks and emails
Forbes · April 16, 2026
Forbes reports a broker's statement that startup wind-down archive deals typically pay $10,000–$100,000 per company, and a reported cielo24 deal described as worth “hundreds of thousands” of dollars in an unusual shutdown case.
How we use it: Broker-reported, shutdown archives only. Not transferable to an operating company and not used in the calculation.
- Vendor advertisementContext only
micro1 — Company referral
micro1 (provider website) · Reviewed October 15, 2026
micro1 advertises “$100K–$2M+ for approved data packages.”
How we use it: A vendor's own marketing range for approved enterprise packages. Not a verified payout or market average; not used in the calculation.
- Vendor advertisementContext only
Polyshares — Data licensing
Polyshares (provider website) · Reviewed October 15, 2026; publication date not stated
Polyshares advertises $100,000–$2 million as its typical licensing range. This is the provider’s own characterization, not an independently verified payout range or an industry average.
How we use it: Marketing claim, not an offer or a prediction of what your company would receive. Not used in the calculation.
- Vendor advertisementContext only
Polyshares — Selected track record
Polyshares (provider website) · Reviewed October 15, 2026; deal dates not stated
Polyshares lists selected, anonymized perpetual-exclusive licenses: roofing $450,000; general contractor $300,000; solar $275,000; marketing agency $110,000.
How we use it: Self-reported by a seller/buyer platform; no company names or contracts independently confirmed. Selected rather than representative, not market averages or automatic benchmarks. Not used in the calculation.
| Category | BLS occupation used | Mean $/hr |
|---|---|---|
| B2B workflows (CRM, projects, internal docs) | Business and financial operations occupations | $45.78 |
| Customer support | Customer service representatives | $22.40 |
| Software engineering | Software developers | $71.20 |
| Finance & accounting | Accountants and auditors | $45.56 |
| Physical trades / field operations | All occupations (broad fallback — no trade-specific figure used) | $33.54 |
| Other | All occupations (broad fallback) | $33.54 |
Compensation load: 1.427 (1 ÷ 70.1% wage share, BLS private industry).
Our assumptions (hypotheses)
Every range below is a Data Deal Guide modeling assumption, unvalidated against realized sale prices. No source supports these specific numbers.
| Assumption | Applies to | Range |
|---|---|---|
| Records per employee per year (fallback when count unknown) | all | 200 – 2000 |
| Uncertainty on a supplied record count | all | 0.8 – 1.2 |
| Usable fraction of raw records | low | 0.02 – 0.08 |
| Usable fraction of raw records | medium | 0.05 – 0.15 |
| Usable fraction of raw records | high | 0.1 – 0.3 |
| Traceability / linked-outcome adjustment | none | 0.2 – 0.4 |
| Traceability / linked-outcome adjustment | partial | 0.5 – 0.8 |
| Traceability / linked-outcome adjustment | strong | 0.85 – 1 |
| Specialist hours to recreate one example | b2b | 0.1 – 0.4 |
| Specialist hours to recreate one example | support | 0.05 – 0.2 |
| Specialist hours to recreate one example | engineering | 0.25 – 1 |
| Specialist hours to recreate one example | finance | 0.15 – 0.5 |
| Specialist hours to recreate one example | field | 0.1 – 0.4 |
| Specialist hours to recreate one example | other | 0.1 – 0.4 |
| Curation / QA premium | all | 1 – 1.3 |
| Buyer willingness-to-pay fraction of replacement cost | all | 0.05 – 0.25 |
| Category demand factor | b2b | 0.9 – 1.1 |
| Category demand factor | support | 0.9 – 1.1 |
| Category demand factor | engineering | 1.05 – 1.35 |
| Category demand factor | finance | 1 – 1.25 |
| Category demand factor | field | 0.95 – 1.25 |
| Category demand factor | other | 0.6 – 0.9 |
| Structure / quality factor | low | 0.75 – 0.9 |
| Structure / quality factor | medium | 0.95 – 1.05 |
| Structure / quality factor | high | 1.1 – 1.25 |
| Distinctiveness factor | common | 0.7 – 0.9 |
| Distinctiveness factor | some | 0.95 – 1.05 |
| Distinctiveness factor | rare | 1.2 – 1.6 |
| Language / location factor | us | 1 – 1 |
| Language / location factor | other | 0.6 – 0.9 |
| License terms factor | nonexclusive | 1 – 1 |
| License terms factor | exclusive | 1.5 – 2.5 |
Limitations
- Not calibrated against realized sale prices: no reliable public dataset of active-company data licensing prices exists.
- Scenario percentiles describe spread across our own assumptions, not a statistical confidence interval or a predicted offer.
- Vendor-advertised ranges and broker-reported wind-down deals are shown for context only and are not used in the calculation; results may differ from them.
- No estimate is given unless you say licensing rights are confirmed and sensitive or third-party material is limited.
Machine-readable: valuation-methodology.json
Try the estimate