Planning & Forecasting
When do spreadsheets stop being enough?
Simplify31 August 20263 min read
In short
The signals are about dependency, repeatability and trust, not about the tool. A spreadsheet becomes a risk when the business depends on one person's memory of how it is put together, and buying software before fixing that just produces an expensive version of the same confusion.
Spreadsheets get more criticism than they deserve. They are flexible, everyone can read one, and they carry a business much further than software vendors would like you to believe. Plenty of profitable companies run their finances out of a well-built file.
The problem is rarely the spreadsheet. It is what has quietly grown around it.
The six signals
- Only one person genuinely understands how the file works, and everyone else is guessing.
- Closing a month takes days of reconciling rather than hours of reviewing.
- Decisions wait on someone being available to update the numbers.
- Errors are found late, usually by accident, usually by someone outside the finance process.
- The same figure exists in three places and does not quite match in any of them.
- Nobody can say who changed a number, or when, or why.
None of those are really about the tool. They are about dependency, repeatability and trust.
The honest test
Why buying software first usually fails
Putting a system on top of an undefined process produces an expensive version of the same confusion, plus an implementation project and a subscription.
The pattern is consistent. The tool is chosen on features, the data going into it is still assembled by hand from the same sources, nobody has agreed what each number means, and six months later the team is maintaining both the system and the old spreadsheet because neither can be trusted alone.
What to fix first
Four decisions, none of which need a purchase.
- 01What gets tracked. A short, agreed set of figures rather than everything that can be produced.
- 02What each one means. One line of definition per number: what is in, what is out, which source, which period.
- 03Who owns each piece, and by when. Most reporting delay turns out to be about ownership rather than process.
- 04What the number is for. A figure nobody acts on does not need automating; it needs deleting.
What a better organised spreadsheet looks like
Sometimes the answer to outgrowing a spreadsheet is a better spreadsheet. Five changes do most of the work, and none of them cost anything.
- 01Separate the tabs by job. Raw data in one place, calculations in another, the view people actually look at in a third. Most broken files mix all three on one sheet, which is why nobody can trace a number back.
- 02Put every assumption in its own labelled cell on one tab, with a note saying where it came from. No figures typed inside formulas.
- 03One source per number. If headcount appears on three tabs, two of them should reference the first rather than repeat it.
- 04Lock everything that is not an input, and colour the inputs so it is obvious where a person is meant to type.
- 05Turn on version history and stop emailing copies. Almost every reconciliation argument traces back to two people working on different versions of the same file.
Once that is settled, the tooling question becomes much smaller, and the honest answer is sometimes still a spreadsheet, just a better organised one.
When a system genuinely is the answer
There are real thresholds, and they are structural rather than about size.
- Multiple entities or currencies that have to consolidate.
- Transaction volume high enough that manual reconciliation cannot keep pace.
- Statutory or investor reporting on a fixed cadence with an audit trail.
- More than two or three people who need to work in the numbers at the same time.
Cross one of those and the case is clear. Short of them, the constraint is almost always the process rather than the file.
About Simplify
Simplify is a finance clarity and investment readiness practice working with founders across India, built on six years inside startups. We write about the questions founders bring before a decision, not after it.
Planning happening after the problem rather than before it?