Compare Rep Performance Across Stages

You are a sales manager doing a fair, data-grounded diagnosis of rep performance at a digital marketing agency. You compare reps stage-by-stage to surface specialization patterns — you do not rank reps overall.
Agency: [AGENCY_NAME] — [SERVICES] | Reps: [REP_LIST] | Period: [PERIOD]
Stage-by-stage conversion data:
[STAGE_PERFORMANCE_TABLE]
Note on volume: [VOLUME_CONTEXT]
For each rep, identify the 1 stage where they outperform the team and the 1 stage where they underperform. Then identify any "specialist" patterns (e.g., a rep elite at top of funnel but weak at close) that suggest role design changes.

- Use stage conversion rates, not absolute deal counts (so a high-volume rep doesn't look better by default).
- Only call out gaps where sample size is meaningful — flag low-volume reps.
- Distinguish a stage weakness from low overall volume.
- Tone is observational, not evaluative. No "best rep" / "worst rep" labels.
- Don't recommend coaching or firing — diagnosis only.

Output:
1. Per-rep card: Strongest stage | Weakest stage | Sample-size caveat
2. Cross-team patterns (e.g., everyone leaks at proposal stage)
3. Specialist candidates (reps whose profile suggests a different role split)
4. Stages with insufficient data to judge

• [AGENCY_NAME] — Agency name
• [SERVICES] — Services sold
• [REP_LIST] — Names of reps being compared (3-8 ideal)
• [PERIOD] — Reporting window
• [STAGE_PERFORMANCE_TABLE] — Rows = reps, columns = conversion rate at each stage
• [VOLUME_CONTEXT] — Note how many opps each rep had so small samples can be flagged

Agency: Northwind — SEO + paid | Reps: Marco, Priya, Dani | Period: Q1 2026
Stage table (% conversion):
| Rep | Lead→Disco | Disco→Prop | Prop→Won |
| Marco | 22% | 40% | 38% |
| Priya | 35% | 55% | 27% |
| Dani | 30% | 48% | 33% |
Volume: Marco 80 leads, Priya 90, Dani 25.

Per-rep cards:
- Marco: Strongest = Proposal→Won (38%). Weakest = Lead→Disco (22%, well below 30% team). Sample size OK.
- Priya: Strongest = Disco→Proposal (55%). Weakest = Proposal→Won (27%). Sample OK.
- Dani: Strongest = appears balanced. Weakest = inconclusive — only 25 leads (low confidence on all stages).

Cross-team patterns: Top of funnel is generally healthy team-wide except Marco. Proposal→Won varies widely (27-38%), suggesting individual selling style matters more here than process.

Specialist candidates: Marco's profile (weak top, strong close) and Priya's (strong middle, weak close) are mirror images. Consider testing a handoff where Priya does discovery+proposal and Marco closes.

Insufficient data: All of Dani's stages — needs another quarter.

• Limit to reps with at least 20 opps in the period — anything less is noise.
• Run this before any major comp plan or territory change.
• Pair with the 'Identify Strong Plays From a Top Rep' prompt to confirm what the strong-stage rep is actually doing.