A prioritization framework to grade and filter AI or automation ideas based on pain level, task frequency, data availability, risk, and build difficulty before writing code.
The Scoring Rubric
| Criteria | Score (1-5) | Description / Guidelines |
|---|---|---|
| Pain Level | How painful is the current manual process? (1 = minor annoyance, 5 = severe bottleneck causing lost revenue or high error rates). | |
| Frequency | How often does this task occur? (1 = quarterly/annually, 5 = multiple times per day per user). | |
| Data Availability | Is the required context clean and accessible? (1 = siloed in emails/Slack, 5 = structured in a clean CRM/database). | |
| Risk / Governance | What is the impact of an AI hallucination or error? (1 = high risk/sensitive customer-facing, 5 = low risk/internal draft). Note: Inverse scale. | |
| Build Difficulty | How complex is the setup? (1 = custom agent with multiple API loops, 5 = simple static prompt or n8n workflow). Note: Inverse scale. | |
| Adoption Likelihood | Will the team actually use it? (1 = requires massive change management, 5 = fits seamlessly into their existing UI/Slack). |
Calculating the Priority Score
Add the scores together to get a total out of 30:
- 24 - 30 (Quick Wins): High impact, low risk, clean data, and easy to build. Automate these immediately.
- 18 - 23 (Planned Projects): Worth building but requires planning (e.g. data cleanup or user onboarding).
- Below 18 (Backlog/Discard): Too risky, too complex, or solves a problem that happens too rarely. Keep manual or defer.
Downloadable Template
You can copy this table directly into your team's Notion or Google Sheets database to triage incoming automation requests.