A spreadsheet is not a failure of planning. For most small businesses it is the correct first tool. It is cheap, everybody already knows how to open one, and you can change it on a Tuesday afternoon without asking permission from anyone. Plenty of profitable companies run on a spreadsheet for years, and they are right to.
The trouble is that a spreadsheet never tells you when it has outgrown itself. It just gets slower, riskier, and more dependent on one person, and nobody notices until something goes wrong in front of a customer. Here are the four signals worth taking seriously.
Only one person really understands it
Ask who could explain the pricing tab from memory. If the honest answer is one name, you do not have a tool, you have a dependency. When that person is on holiday, changes wait. When that person leaves, part of how your business works leaves with them, because it was never written down anywhere except inside formulas.
This is the signal that matters most, and it has nothing to do with how big the file is. A two tab spreadsheet with one owner is more fragile than a twenty tab one that three people maintain together.
People are nervous about editing it
A healthy tool invites changes. Watch what your team actually does. If they open the file read only, save a copy before touching anything, or send you a message asking whether it is safe to change a number, then the file is already known to be fragile.
The nervousness is rational. Nothing in a spreadsheet stops a mistyped cell from changing a price for every customer, and nothing tells you afterwards that it happened. There is usually no record of who changed what, so a mistake found in March cannot be traced back to February.
Everybody keeps their own copy
Copies are the clearest symptom, and they nearly always have the same three explanations:
- A copy on somebody's desktop, because the shared one was locked when they needed it.
- A copy emailed to a customer or a supplier, now several price changes out of date.
- A copy somebody made to try something, which quietly became the one they use every day.
Each copy is a competing version of the truth, and reconciling them costs real hours. Nobody set out to create them. They appear because the original was not available at the moment somebody needed it, which is a design problem rather than a discipline problem.
It cannot be used where the work happens
This one is easy to test. If the work happens on a warehouse floor, in a van, or standing in a customer kitchen, and answering a simple question means walking back to a desk, the tool is in the wrong place. Spreadsheets on a phone are technically possible and practically miserable. Columns run off the screen, a thumb edits the wrong cell, and nobody wants to search a file on a device where searching means pinching and zooming.
The realistic alternatives
There are three, and they are nowhere near equally expensive. Consider them in this order.
- Fix the spreadsheet. Lock the formula cells, give it one named owner, delete the tabs nobody uses, and move it somewhere with version history such as SharePoint or Google Drive. This takes a day or two and often buys another year. It is the right answer more often than software vendors will tell you.
- Move the data into a shared list with a simple front end. Microsoft Lists, Airtable, and tools like them give you a proper record per row, an edit history, and a usable phone view, without a development project. Good when the data is simple and the rules are few.
- Build a real application. Microsoft Power Apps is usually the sensible choice if your team is already on Microsoft 365, because the sign in and most of the licensing are handled. This is the right step when the rules are genuinely complicated, when different people need different permissions, or when the spreadsheet has become the record you could be audited on.
The decision is rarely about the size of the file. It is about how much the business would hurt if the file were wrong for a week and nobody noticed.
If you want somewhere to start, open the file this week and count the people who could confidently change it. If the number is one, fix that first. Everything else on this list gets easier once it is no longer true.