A spreadsheet is the right tool for a young business. It costs nothing, everybody already knows how to use it, and it bends to whatever you need that week. Nothing here is an argument against it.
The trouble is that a spreadsheet never tells you when you have outgrown it. It keeps opening, and the work around it quietly gets heavier. These are the five signs we would look for. None of them needs an accountant to spot.
Signs in the files themselves
The first sign is two versions of the same file. Sales keeps a customer list, accounts keeps another, and they disagree about who owes what. Somebody then spends a morning working out which one is right, and the answer is usually neither. When a file has to be emailed round to be updated, copies are not an accident. They are how the method works.
The second sign is the one person who understands the workbook. Every business on spreadsheets has one: the formulas are theirs, the hidden columns are theirs, and nobody else dares to touch a cell. It works until they are on leave during month end, or until they resign. At that point the business discovers that part of its accounting system was a person.
Signs in the numbers
The third sign is a month end that takes two weeks. If the figures for March arrive in the middle of April, most of that time was not spent on accounting. It was spent collecting files, retyping them into one place and chasing the differences. A number that arrives that late is a record of what happened. It is too old to help with a decision.
The fourth sign is a stock figure nobody trusts. The sheet says forty units, the shelf holds thirty-one, and the storekeeper has learnt to go and look before promising anything to a customer. Once people walk to the shelf rather than open the file, the file has stopped being a system. You are paying for the count twice and believing it once.
The sign that arrives from outside
The fifth sign is a question you cannot answer quickly. A bank asks for receivables by age. A supplier wants to agree a balance. A partner asks which branch actually makes money. None of these is an unreasonable question, and each of them turns into days of work. How long a plain question takes to answer is a fair measure of whether your records are still serving you.
One sign on its own is a nuisance. Three together mean the business is being run from memory, with spreadsheets as the paperwork.
What to do first, and it is not buying software
The tempting next step is to go shopping for an ERP. We would hold off. Software bought before anyone has written down how the business works tends to be configured around guesses, and then the team quietly goes back to the spreadsheets. Start with paper. For each process that matters — selling, buying, receiving stock, collecting money, paying suppliers — write down:
- Who does it, and who covers when they are away
- Which document starts it: an order, a delivery note, an invoice
- Which file it is recorded in, and who else keeps a copy
- Who checks it, and when
- Where it goes wrong most often
This takes a few sittings, not a project. It also earns its keep whatever you decide afterwards, because some of what it reveals can be fixed without any software at all: a missing approval, a duplicated list, a step that nobody owns.
Then be honest about the size of the answer
With the processes written down, the real question becomes visible: does this business need a full ERP, or a focused tool? If several departments must work from the same data — sales, stock, purchasing and accounts all depending on one another — a full ERP is the honest answer, and it comes with an implementation project. If the business runs one kind of operation and the pain sits in one place, a focused application for that job may be everything it needs. Both are good outcomes. The poor outcome is buying the large answer for a small problem, or patching a large problem with one more spreadsheet.