Founder & operator workflows
AI tool evaluation scorecard workflow
New AI tools ship every week. You can't evaluate them all, and bad tool choices cost months.
Audience
Founder & operator workflows
Problem
New AI tools ship every week. You can't evaluate them all, and bad tool choices cost months.
Outcome
Every tool gets a 30-minute scorecard with a keep/park/kill verdict and a follow-up date — no more open tabs forever.
Tools used
- A blank scorecard template
- Sample inputs from your real workflows
- An LLM only as a comparison aid
- A calendar
Time required
30 minutes per tool
Difficulty
Easy
Steps
- Define 6 fields: real problem it solves, your sample input, output quality, integration cost, privacy posture, switching cost.
- Try the tool on your real (sanitized) sample input — not the vendor's demo data.
- Score each field 1-5. No averages — just look at the worst score.
- Decide: keep (use this month), park (revisit in 90 days), kill (don't reopen). Set the date.
- Publish the scorecard internally so you don't re-evaluate the same tool in 30 days.
Example output
Tool: Vendor X note-taker. Worst score: privacy posture (2/5, no clear data-retention policy). Verdict: park. Revisit: 2026-08-26.
Human-review point. Founder/team lead writes the verdict and date. LLM may compare features but cannot decide adoption.
Privacy notes. Never paste real customer data, employee data, or unreleased product information into vendor tools during evaluation. Use sanitized samples.
Fork and remix ideas
- Adapt for hiring tool evaluation.
- Adapt for community-bot platform comparison.
- Adapt for picking a model provider with clear privacy rules.
related workflows
Keep going.
These pair well with this workflow.
Founder weekly metrics build-in-public workflow
You want to share weekly progress publicly but don't know which numbers honestly reflect momentum versus vanity.
AI meeting notes to action tracker workflowMeeting recordings pile up. Decisions and owners get lost the next morning.
No-code lead qualification workflowInbound leads include great fits, casual browsers, and time-wasters — and you waste hours sorting them.
submit yours
Have a better version of this workflow?
Send the audience, the problem, the steps, and what you'd never automate. We review every submission with a human before publishing.