Research framework · October 2026

Industrial Data Valuation

How can a portfolio of industrial datasets be valued as an intangible asset in a way that is transparent, reproducible and honest about its uncertainty?

All values are relative. They rank datasets against each other to decide where to act first; they are not an opposable price.

Limits first

The case is anonymised and every amount is a working assumption, consistent with the others, not accounting data.
The income approach carries most of the value, so results depend on assumptions about future benefits.
Compliance weights are set by judgement; pairwise comparisons (AHP) are used to test them, with an illustrative matrix.
A high compliance score shows that data is ready to become an asset, not that it has become one: rights registration and balance-sheet recognition are outside the grid.

Approach

1 · Compliance: Eight datasets scored on 17 criteria in three blocks: documentation, intrinsic quality, fitness for the model.
2 · Valuation: Cost, market and income approaches, each adjusted by the compliance score, blended by confidence.
3 · Uncertainty and priorities: Sensitivity analysis, 10,000-draw Monte Carlo simulation, AHP weighting, and a value × exposure matrix that sorts datasets into four priority groups.

V = QRU × Σ(Val × ICF) / Vp

Reference formula of the DATAMITE data valuation framework: a quality, relevance or utility score (QRU) times the potential value of each component (Val) weighted by its information cost fraction (ICF), corrected for processing time (Vp).

Findings

2 of 8datasets pass the compliance threshold

−16 % to +12 %correction per dataset, +4.5 % for the portfolio

14 %relative width of the 90 % Monte Carlo interval

Unchangedpriority groups when weights are derived by AHP

The method sorts value rather than inflating it: well-documented datasets gain, the others lose, and the portfolio barely moves. The most profitable improvements are scale completeness and annotation quality on the datasets closest to the threshold.

Global compliance score of each dataset against the two thresholds — Compliance score of each dataset against the two thresholds.

Cost, market and income values of each dataset with the hybrid value — Three approaches and the hybrid value retained for each dataset.

Normalised value against regulatory exposure, with four priority groups — Value against regulatory exposure: four priority groups.

Sensitivity of the portfolio value to each assumption — One assumption at a time: income assumptions dominate.

Monte Carlo ranges per dataset and for the portfolio — Ranges rather than points: 10,000 scenarios, fixed seed.

Regulatory timeline of data as an asset, 2020 to 2026 — Regulatory context of the jurisdiction studied, 2020 to 2026.

Market context

18 indicatorsin Black Book's 2026 European data valorisation index, 700 respondents: value scored before it is priced

$0.7bn → $1.3bnhealthcare data valorisation market, 2026 to 2031, 14.5% a year (MarketsandMarkets)

The European Health Data Space sets March 2027 for key implementing acts, before application phases in 2029 and 2031. These are research-vendor figures for another sector, not audited: an order of magnitude and a check on method, not a price comparable.

Reproducibility

Python 3.13, fixed seed 2026, no external data: Run All recomputes every figure. Code under the MIT License; text, figures and data under CC BY 4.0.