# How should a club turn viewing into a same-shortlist supplier trial?


**Use the media to formulate tests, not to award purchasing scores.** The following is this desk’s proposed worksheet, rather than a methodology claimed by any individual speaker.

| Trial field | What to record |
|---|---|
| Recruitment decision | The role, squad need, competition-season scope and deadline |
| Existing-stack baseline | What the club can already establish, including current workarounds |
| Supplier contribution | Additional supported findings, separately from reformatted existing findings |
| Evidence trail | Match, player identifier, timestamp, field definition and source where applicable |
| Coverage exceptions | Missing matches, seasons, footage or required fields |
| Integration work | Matching failures, manual steps, engineering dependencies and correction time |
| Intended use | Written clarification for exports, storage, modelling, AI processing and group sharing |
| Decision | Retain, add, replace, consolidate, investigate further or buy nothing |

### What should the evaluation team do differently after watching?

- **Head of scouting:** Specify observable football behaviours before selecting metrics.
- **Recruitment analytics lead:** Reproduce one relevant result using an authorised sample; document exceptions rather than hiding them in an average score.
- **NWSL recruitment lead:** Test women’s competitions and comparison populations directly.
- **Sporting director or general manager:** Ask what decision changes, what work remains and who owns implementation.
- **Group procurement lead:** Request entity-specific scope and written answers on intended workflows before assuming a group-wide solution.

For AI-assisted reports, include deliberate factual verification. NIST’s Generative AI Profile identifies confidently presented false content—and even fabricated supporting logic or citations—as a recognised risk. That is a reason to test evidence tracing, not a finding that any supplier in this roundup has failed. [NIST Generative AI Profile](https://nvlpubs.nist.gov/nistpubs/ai/NIST.AI.600-1.pdf). ([nvlpubs.nist.gov](https://nvlpubs.nist.gov/nistpubs/ai/NIST.AI.600-1.pdf?utm_source=openai))