Analyze this CSAT survey data.
Inputs:
- Survey responses: {{paste — scores, comments, customer metadata, date}}
- Period covered: {{e.g. Q2 2026}}
- Comparison period (optional): {{e.g. Q1 2026}}
Output:
## Distribution
| Score | Count | % | vs. comparison period |
|---|---|---|---|
## What stands out
- Bimodal? (Lots of 5s and 1s, few in the middle = polarized product or service)
- Long left tail? (A handful of very unhappy customers — these are the ones to call)
- Long right tail? (Lots of mild positives, few champions — under-loved product)
## Top positive themes
2–3 themes from comments. Representative quotes (not most extreme).
## Top negative themes
Same structure. Per theme: how many customers, what segment, what specific issue.
## Who I'd call this week
The 3–5 customers whose comments suggest they're at risk but reachable. Include the specific issue + suggested opening line.
## What this trend means
If you have a comparison period: is the average movement real signal or normal variance? If the average is flat but the distribution shifted (more 5s and more 1s), that's a finding even when the headline number didn't move.
Hard rule: don't celebrate "average went up" if the underlying distribution got more polarized. That's not improvement; that's volatility.customer-successcsatanalysis