Start for $0.99

20 spreadsheet prompts that show the formula, not a guess

Ask a model about data it cannot see and it will invent the numbers. These prompts hand it your columns and a few sample rows first, then make it write the formula or query so you can check every step.

Swap anything in [brackets] for your columns, sample rows or question.

01. Get a formula that fits your real columns

You know what you want, not how to write it.

I use [Excel / Google Sheets]. My data:

Columns: [column letters and headers, e.g. A: Date, B: Region, C: Revenue]
Sample rows:
[paste 5 rows]

I want: [what the formula should calculate]

Write the formula for cell [cell], explain each part in one line, and tell me what it returns for the sample rows so I can check it. Mention any case where it would break, like blanks or text in a number column.

02. Explain a formula you inherited

A spreadsheet full of formulas nobody remembers writing.

Explain this formula from my spreadsheet:

[paste the formula]

The columns it refers to hold: [what each referenced column contains]

Break it down from the inside out, one piece per line, in plain words. Then say what it is trying to do overall, whether there is a simpler way to write it, and any bug you can see.

03. Plan the cleanup before you touch messy data

An export with duplicates, blanks and mixed formats.

Here are the headers and 15 sample rows of a messy dataset:

[paste them]

List every data quality problem you can see: duplicates, inconsistent formats, blanks, typos, mixed units, merged values. For each, give the fix as a formula, a Find and Replace, or a step, in the order I should do them.

Do not clean the data yourself. I will run the steps.

Unlock the other 17 prompts for free

Sign up for the Whizi newsletter to read the rest of this pack.

  1. Match data across two sheets
  2. Design the pivot table for your question
  3. Write the SQL query against your schema
  4. Pick the chart that answers the question
  5. Plan an analysis before you run it
  6. Read your results without fooling yourself
  7. Highlight the rows that need attention
  8. Find and remove duplicates safely
  9. Fix dates that will not behave
  10. Plan a dashboard people will actually check
  11. Automate a spreadsheet task you do every week
  12. Pull a piece out of messy text cells
  13. Build a simple forecast you can defend
  14. Summarize survey responses without cherry picking
  15. Review a spreadsheet model before someone relies on it
  16. Explain what the numbers mean to someone non technical
  17. Have a second model check the first one's analysis