27 Sep 2026 · 7 min readData

Spreadsheets and Data Without the Pain

Most spreadsheet problems are not exotic. They are inconsistent dates, numbers stored as text, merged cells, and a column called 'notes' that is now load-bearing. This guide covers the structure that survives real use, the cleaning steps that fix most problems, and the mistakes that quietly corrupt everything downstream.

Structure that survives

  • One row per thing, one column per attribute. No merged cells anywhere — they break every filter, sort and pivot.
  • One value per cell. A column of 'Mon, Tue, Wed' in one cell cannot be sorted or grouped, ever.
  • Never leave a blank cell where a value belongs. Use an explicit 'unknown' or 'N/A', so blanks mean something.
  • Keep dates as real dates, in one format, not as text in four formats.
  • Do not merge data into a presentation sheet. Keep the raw data on its own and summarise separately.
  • Keep a lookup sheet for repeated lists, so 'Design' is never typed as 'design' in one row and 'Design ' in another.
⚠️ Note: Numbers stored as text are the quiet one. They look right, they total wrong, and charts based on them quietly ignore them.

Cleaning most messy data

  • Find duplicates before anything else. They inflate every total and skew every average.
  • Standardise the text: trim the spaces, fix the capitalisation, decide how missing values are written.
  • Convert the date formats to one, and check whether any are ambiguous — 01/02/2026 is not a date in most of the world.
  • Check the types. Anything summing to zero is usually text pretending to be a number.
  • Decide what to do with blanks before you start, and write it down, rather than improvising per row.

When the format stops fitting

The moment you start writing the same formula with a small change in twelve rows, you have outgrown the spreadsheet. That is the signal to move the data into JSON and work with it properly — and most of that work is now one copy and paste away.

Explore the Full SlashAI Library

Every prompt in our guides is part of our offline-ready vault of verified commands and instant browser tools. Free forever, no account required.

Browse All Commands