Analyze Your Data · Lesson 1Free~12 min readNo code requiredFive questions answered by the end

From spreadsheet to answers, just by asking.

Your spreadsheets are full of answers you don’t have time to dig out — and the formulas to get them are exactly what most people never learned. This lesson removes that wall: you’ll upload real data, ask it questions in plain English, and learn to verify the answers so you can trust them.

The mental model

You don’t need formulas or a stats class. You ask your data questions in plain English, and AI does the work.

Spreadsheets hide their answers behind formulas most people never learned. AI removes that wall: you upload your data and ask, “which month was best?” or “what stands out here?” and it analyzes and explains. The skill isn’t spreadsheet syntax anymore — it’s asking good questions and checking the answers.

The Reframe

Interview your data. Treat your spreadsheet like an expert you can question. Start broad (“what’s interesting here?”), then drill into whatever it surfaces.

Step 01 Get your data ready

AI reads tidy data far better. Quick prep:

Headers on top — one clear label per column.
One record per row — no merged cells or stacked tables.
Consistent formats — dates as dates, numbers as numbers.

Step 02 Upload and ask

Upload the file to a tool that reads data (ChatGPT or Claude with file upload, or the AI built into Sheets/Excel) and ask in plain language.

Starter questionsHere’s my [sales] data. Give me a quick overview: totals, the trend over time, the top and bottom performers, and anything that stands out or looks unusual. Explain what each number means in plain English.

Step 03 Ask better questions

The good ones go beyond totals: trends over time, top and bottom, comparisons between segments, and the open-ended “what stands out” that surfaces things you didn’t think to ask.

Step 04 Verify before you trust

Spot-check at least one number against the raw data, and make sure it read the right columns. An answer you haven’t sanity-checked is a guess with confidence.

AI can miscount, misread a column, or quietly choke on messy data — and still answer confidently. Always verify a key figure against the source, and clean obvious mess first. A wrong number that looks authoritative is worse than no answer.

Your challenge: interview a real spreadsheet

Grab a spreadsheet you actually use. Then:

Tidy it: headers on top, one record per row, consistent formats.
Upload it and ask for a plain-English overview.
Ask five real questions, including one open-ended “what stands out.”
Verify at least one number against the raw data.

That’s answers out of a spreadsheet without a single formula. Next, go from answering questions to producing real analysis — cleaning data, finding patterns, and making charts — that’s Lesson 2.

Next: insights, charts & summaries →Back to Analyze Your Data

What you can do now

Ask a spreadsheet questions in plain English
Prepare data so AI reads it correctly
Ask question types that surface real answers
Use open-ended questions to find the unexpected
Verify AI’s numbers before trusting them

Pro

Up next in Analyze Your Data

Lesson 2 · Insights, charts & summaries

Clean messy data, find real patterns, build the right charts, and summarize analysis for people who’ll never open the spreadsheet. Go to Lesson 2 →