Fill in these first
- Complaint log: [paste rows: date, product or service, channel, complaint text or category, date closed, outcome]
- Period: [e.g. 2026-07-01 to 2026-09-30]
- Orders or customers in the same period: [total, so complaint rates can be worked out, or unknown]
- Current categories, if any: [list, or none]
- Promised resolution time: [e.g. 48 hours]
Task
Find what customers complain about most, what takes longest to fix, and which fixes would remove the most complaints.
Steps
- Sort every complaint into 6 to 10 issue types written in plain words, such as "late delivery" or "wrong size sent". Merge near-duplicates.
- Work out count, share of total, complaint rate per 1,000 orders if orders are given, median days to close and share closed late.
- Find repeat complaints from the same customer or about the same order.
- Cross the top three issue types with product and channel to find where they concentrate.
Output format
- Five-bullet summary with the biggest numbers first.
- | Issue type | Count | Share | Per 1,000 orders | Median days to close | Closed late |
- Top five products or services by complaint count, each with its main issue.
- Five fixes ranked by complaints removed, each with the owner team and an effort level of low, medium or high.
- Three numbers to track monthly from now on.
Rules
- Quote no customer names, phone numbers, emails or order numbers.
- If a complaint text is unclear, put it in "Unclear" rather than guessing a type.
- Base every rate on the counts pasted and state the sample size.






