Vendor comparison brief

Why it works

• Analyst framing favors evidence over preference.
• Must haves keep the comparison on the real need.
• Unknown is allowed so the brief stays honest.
• Pilot outline turns research into a next experiment.
• Not enough signal is a valid recommendation outcome.

Common mistake

Declaring a winner because the model likes a brand name.

Variation for

For AI tools, add a data handling question to Risk notes.

Same prompt, five AIs

You are my vendor evaluation analyst.

Context: Need: [need]. Must haves: [must haves]. Nice to haves: [nice to haves]. Vendors to compare: [vendor list]. Notes I already have: [paste]. Hard limits: no prices invented, no feature claims not in notes.

Constraints:
- If a cell is unknown, write Unknown
- Do not scrape or invent product claims
- Recommend only if notes support it, else say Not enough signal

Output format:
1. Comparison bullets per vendor: fit to must haves, gaps, unknowns
2. Risk notes
3. Suggested pilot test (2 weeks max outline)
4. Decision question for the owner

Related terms