Turn a messy CSV into a report that says what to fix first

How this helps you

Messy data becomes a short list of what to fix first, with the cleanup shown so you can trust it.

✦ Try it

Here is a CSV of [90 days of leads]. Clean it first and list every change you made. Then build a report Page: where we lose [time or money], the evidence, and each opportunity scored on impact and ease as do now, do later or leave it.

What they did

Using a made-up HVAC business with invented numbers, Kacper Rutkiewicz exports 90 days of messy lead data as a CSV (mixed date formats, duplicates, inconsistent source names) and asks his Bot for an opportunity report: clean it up, show where the business loses money, and say what to fix first. The report includes cleaning notes (rows removed, dates normalised, casing fixed), a headline number, a summary of findings such as faster replies closing more often, and an impact versus ease matrix sorting each opportunity into do now, do later or leave it. He then asked for edits, like making the headline number bigger.

Who it's for: Owners, consultants and analysts with spreadsheet data.

How to try it

Export the data as CSV (or share the sheet link).
Ask for a report that cleans the data first and lists what it changed.
Ask it to score opportunities on impact and ease and sort them.
Edit the page in chat until it reads clearly.

What to expect

A report Page that starts with cleaning notes and ends with a prioritised list.

Common mistakes

Skipping the cleaning notes. They show whether the numbers can be trusted.

Related tips

Analyse your own usage CSV: Export usage from the Usage page and ask your Chief of Staff Bot how to cut it.
Give deep research a finish line: "Research this deeply" gives no finish line and no reason to stop browsing.
Build a weekly before-and-after scoreboard from a Google Sheet: Point a Page at a sheet, compare baseline weeks with recent weeks, and refresh it every Monday.
Find what your audience keeps asking for in 28 days of comments: Connect YouTube through Zapier's MCP server, then have your Bot read recent comments and rank the most requested topics with real quotes and like counts.