1 min lesson
A worked severity × impact pass
Use two rows in "A worked severity × impact pass" to state the practical decision rules.
Step 1 of 2
A worked severity × impact passWhat a ranked batch looks like
- Cluster
- Agent silently drops imports
- Severity
- Critical (data loss)
- Impact (vol / ARR)
- Low vol / high ARR (paid teams)
- Score
- High
- Rank
- 1
- Cluster
- Indexing fails on monorepos
- Severity
- High
- Impact (vol / ARR)
- Med vol / high ARR
- Score
- High
- Rank
- 2
- Cluster
- Tab latency spikes after update
- Severity
- Medium
- Impact (vol / ARR)
- High vol / mixed ARR
- Score
- Medium
- Rank
- 3
- Cluster
- ⌘K wording confusing
- Severity
- Low
- Impact (vol / ARR)
- High vol / low ARR
- Score
- Low
- Rank
- 4
| Cluster | Severity | Impact (vol / ARR) | Score | Rank |
|---|---|---|---|---|
| Agent silently drops imports | Critical (data loss) | Low vol / high ARR (paid teams) | High | 1 |
| Indexing fails on monorepos | High | Med vol / high ARR | High | 2 |
| Tab latency spikes after update | Medium | High vol / mixed ARR | Medium | 3 |
| ⌘K wording confusing | Low | High vol / low ARR | Low | 4 |
Note the #1: low volume, but data loss for paying teams beats a loud cosmetic gripe.
Quiet beats loud
The trap this round sets is volume bias. A wave of “the icon is ugly” tickets feels urgent; one report of Agent silently corrupting source is a five-alarm fire. Your ranking is the whole test and a quiet-but-catastrophic issue belongs at the top even with a ticket count of one.