[§1] Qualitative checklist, eight drivers

Evidence before an estimate

The Data Value Scorecard records readiness across eight drivers. It does not price a dataset or verify licensing rights.

[§1.1] Can you document permission to license this data for AI training?

[§1.2] Does collection or removal of personal data support the proposed use?

[§1.3] How difficult would it be to source similar data elsewhere?

[§1.4] How much specialist domain content does it contain?

[§1.5] Does it cover under-served languages or dialects?

[§1.6] Is it structured, deduplicated, and labelled or aligned?

[§1.7] Is its usable volume meaningful for the data type?

[§1.8] Will new data keep arriving?

How to interpret it

Each answer is worth 0, 1, or 2 points. The overall score is the total divided by 16, multiplied by 100, and rounded. Driver scores are points multiplied by 5. All drivers carry equal, heuristic weights. A driver below 5 flags an evidence gap; it is not a monetary discount.

Rights, consent, uniqueness, domain depth, language coverage, quality, scale, and refresh all deserve documentation. A high total cannot compensate for missing rights. No numeric score determines whether a license is lawful.

Older research references to a scorecard or weighting matrix describe a qualitative framework. This tool provides a readiness checklist only. Use the valuation workbench for arithmetic based on separately entered prices.

Read the methodology or request a scenario review.

Evidence before an estimate

Readiness result

How to interpret it