Feedback prioritization

# Feedback prioritization

Use when you want the customer ideas in a filtered list report ranked by likely business impact.

## Workflow

1. Confirm the input is a filtered list report of ideas
2. Read the full report data. Do not rank from the report configuration alone.
3. If the report is unfiltered or too broad to produce an actionable result, ask the user to narrow it by workspace, category, theme, status, segment, or time frame
4. Evaluate every idea in the report. Do not sample.
5. Assess likely business impact using these default criteria unless the user overrides them: number of votes, product value score, and recency
6. Use requesting organizations as an additional demand-strength signal when available, and say when it materially changes the ordering
7. Group every idea into high, medium, or low impact. Keep the high tier small enough to act on, usually five to eight ideas.
8. Order ideas within each tier by the strongest available signals
9. Recommend a path forward for each tier: promote now, watch, or close and defer
10. Save the analysis as an artifact so the team has a reusable version

## Rules

- State which signals were populated and which were missing
- Do not treat a blank product value score as a low score
- When the user changes the criteria or weighting, restate that logic before presenting the ranking
- When signal quality is weak or conflicting, say so and explain how that uncertainty affects the ranking
- Every idea appears in exactly one tier
- Reference every idea by reference number and name
- Do not update records, change statuses, or promote ideas unless the user asks
- Headers use sentence case. No em dashes, en dashes, or curly quotes.

## Output

```
### Summary  
- Top ideas by likely business impact:  
- Main pattern across the set:  
- Biggest ranking uncertainty:  
  
### Signals used  
- Populated:  
- Missing:  
- Ranking logic:  
  
### High impact  
| Idea | Votes | Product value score | Recency signal | Why it ranks here |  
  
### Medium impact  
| Idea | Votes | Product value score | Recency signal | Why it ranks here |  
  
### Low impact  
| Idea | Votes | Product value score | Recency signal | Why it ranks here |  
  
### Recommended path forward  
Short narrative sub-sections, not terse bullets.  
- Promote now: name the ideas, then 1 to 2 sentences on why they are the clearest near-term moves.  
- Watch: name the ideas, then 1 to 2 sentences on what validation they still need.  
- Close or defer: name the ideas, then 1 sentence on what evidence would justify revisiting them.  
- Low-confidence placements: include whenever an idea could move materially on missing evidence. Name the ideas and what would change their position.  
  
### Takeaway  
One short paragraph. Summarize the overall signal quality in the backlog and the best immediate decision pattern for the team.
```

## Quality checks

- Every idea in the report appears in exactly one tier
- The signals section names both what was populated and what was missing
- The takeaway is a paragraph, not a list