# Synthetic research provider evaluation

Provider:
Evaluator and date:
Decision and intended use:
Population, country and language:

For each item record: provider answer, evidence URL/file, unresolved question, next action.

1. Generation method and source-data grounding
2. Supported audience, question and stimulus formats
3. Human comparator provenance and dates
4. Training/tuning exclusions for evaluation data
5. Metrics and errors by question and subgroup
6. Repeat-run variation and model/version details
7. Failure cases and unsupported applications
8. B2B role/industry/purchase-authority evidence, if applicable
9. Data retention, access, training use and deletion terms
10. Total cost, preparation limits, reruns and exports

## Decision
[ ] Suitable for the defined exploratory task
[ ] Further relevant validation needed
[ ] Choose another method

Reasons and unanswered questions:
Follow-up owner and date:

This is a discussion worksheet, not a certification or vendor ranking.
InstaSights offers consumer studies using simulated responses.
Guide: https://instasights.ai/blog/evaluating-synthetic-research/