Measuring artisan livelihood impact without a separate reporting exercise

The rule: every reported figure links to a row

If a number in a funder report cannot be clicked through to an order, a ledger entry or a verified product, it is an assertion. The design goal is that the report is a query over the operating system, not a document assembled beside it.

The five metrics that carry the most weight

These are derivable directly from transaction data, with no survey instrument required.

• Makers with at least one verified product published
• Makers with at least one settled payable in the period
• Median earnings per active maker, and the distribution — not just the mean
• Repeat production rate: makers who submitted again after being paid
• Techniques kept in active production, including endangered ones

What surveys are still needed for

Transaction data cannot tell you about household decision-making, dignity, or whether income displaced other work. Keep the qualitative instrument — but stop using it to produce numbers the ledger already holds more accurately.

Comparability across a portfolio

Funders comparing grantees need a shared data model, not a shared template. When every programme runs on the same rails, a weaving cooperative and a refugee livelihoods programme produce structurally identical evidence, and portfolio-level analysis stops being a guess.

What happens after the grant

The strongest signal in artisan livelihood investment is whether the earning channel continues without the grant. Measure it directly: production and settlement volume in the two quarters after funding ends.