Clean Up a Messy ATS Export
When to use it
Make a raw ATS export usable before you report on it, without pasting candidates' personal data into an AI tool.
The prompt
Here is a sample of a raw export from our ATS. Help me get it into a shape I can analyse. WHAT I'M TRYING TO REPORT ON: {{e.g. time to hire by department, source of hire, pass-through by stage}} SAMPLE ROWS: {{paste 10-20 representative rows, including the header row and any rows that look wrong — with names, emails and phone numbers replaced by placeholders like "Candidate 1"}} Tell me: 1. What's structurally wrong with this data — inconsistent date formats, duplicated candidates, free-text fields that should be categories, blank cells that mean different things. 2. A step-by-step cleanup in {{Google Sheets / Excel}}, with the formulas or Find-and-Replace patterns for each step, in the order to apply them. 3. Which columns I actually need for the report I described, and which I can drop. 4. Which problems I should fix at the source in the ATS instead of patching every export, and roughly what that would involve. 5. Any column where the data looks plausible but is probably unreliable — a source field recruiters fill in inconsistently, a stage date that only records the latest move — and how I'd check. 6. A short checklist to rerun on next month's export, so the cleanup is repeatable.
Tip
Strip names and contact details before you paste, but keep the ugly rows. A tidied sample gets you a cleanup plan for data you don't have; pseudonymised real rows keep every problem and none of the personal data.
How to use it
Paste the prompt into ChatGPT, Claude or whichever assistant you use, then replace every {{bracketed}} part with your own detail. The more specific and messier your input, the better the output — a model given raw notes has more to work with than one given a tidy summary you wrote first.
More Reporting & Metrics prompts
Open the full tool for all 87 prompts, the glossary, and the JD decoder.