There comes a point, in any growing company, when someone says "we can't keep running this on Excel." Usually right after an expensive mistake. Usually the next step is buying software, and usually the problem stays exactly where it was.
The spreadsheet is not the problem. The problem is the moment it becomes the company's only source of truth and depends on one person to stay correct. This article shows how to recognise that moment, and what to do once it arrives.
Excel is very good at what it was built for
A spreadsheet is unbeatable for modelling a decision: running scenarios, testing a price list, drafting a budget, doing a calculation you'll only ever do once. It is flexible, it is instant, and everyone knows how to use it.
What a spreadsheet does badly is holding the official record of an operation that several people change over time. It has no history of who changed what, nothing stops someone deleting a formula without noticing, it validates nothing you type in, and it cannot guarantee that the version open in front of you is the right one. None of this is a flaw — it is the consequence of being a calculation tool, not a database.
The signs the spreadsheet has stopped being enough
There is more than one version of the same file
"Client_list_v3_final_revised.xlsx" is a complete diagnosis in one line. Once the filename starts carrying the history, it is because there is no history inside the file itself. From here, two people can be working off different numbers, both convinced they are right.
Only one person knows where the formulas are
If the cash-flow sheet is only ever updated by one person, and nobody else understands how it works, the company no longer has a process — it has a dependency. It is the same risk described in when the operation lives in one head, made worse by being invisible — it looks like the knowledge is in the file, when it's actually in the person.
Excel's numbers don't match the accounting
And the argument about which of the two is right has already happened more than once. When the sheet and the trial balance diverge, the company does not have two sources: it has none, because it has stopped trusting either.
Information gets copied from a system into the sheet, by hand
Exporting invoicing data and pasting it into a spreadsheet is where mistakes creep in and where time gets lost. If this happens every month, the annual cost of that copy-paste often already pays for the tool that would make it unnecessary.
Nobody can answer a simple question without opening the file
"How much does this client owe us today?" should not require opening a file, filtering a column, and trusting the last update was actually done. If it does, the information is not accessible — it is filed away.
What to replace first
The transition happens function by function, never all at once. This is the order that tends to work best at an SME:
| Function | Sign it's breaking down | What it should move to |
|---|---|---|
| Invoicing | Manual numbering, VAT errors | Tax-authority-certified invoicing software |
| Accounts receivable ledger | Nobody knows what's outstanding | The same invoicing software |
| Expenses and documents | Documents lost between email and paper | A digital archive with a naming rule |
| Payroll processing | Manual calculation of bonuses and absences | Payroll software or the accountant |
| Cash flow forecast | — | Can stay in a spreadsheet |
| Annual budget | — | Can stay in a spreadsheet |
Issuing invoices through tax-authority-certified software is now mandatory for most companies, which makes this the first replacement to make — and also the least debatable.
What is not worth replacing
The cash flow forecast and the annual budget should stay in a spreadsheet for much longer than most software vendors would like. They are modelling exercises, they change structure often, and they benefit from flexibility. A 90-day cash flow forecast built in Excel beats no forecast at all, built in an expensive system nobody opens.
How to make the switch without stopping the operation
- Replace one function at a time, three months apart.
- Run the new system and the old sheet in parallel for a month, and compare the results at the end.
- Write the process down before choosing the tool — a new tool laid over an undefined process produces more expensive chaos.
- Always check how you get data back out. The criteria are in how to choose invoicing software.
- Decide who becomes responsible for each function in the new system, before switching it on.
Check your own case
- There is a single file, with no parallel versions, for each critical function.
- At least two people know how to update each important sheet.
- The sheet's numbers and the accounting's numbers were checked against each other last quarter.
- Invoicing is issued through certified software, not a sheet or a text template.
- No information is manually retyped between systems every month.
Two or more "no" answers mean Excel stopped being enough a while ago — and the cost is already being paid, just spread thin enough not to notice.
Frequently asked questions
What's the right moment to switch?
It is not a question of size, it is a question of how many hands touch it. As long as a sheet is maintained by one person and read by that same person, it works fine even at companies with dozens of employees. From the moment two people change it, or a real decision depends on it, the risk stops paying off.
Does buying an ERP solve the problem?
It solves it if the process already exists; it makes it worse if it doesn't. An ERP imposes a way of working, and a company that hasn't yet decided who approves what will spend months configuring decisions it never actually made. Writing the process down first is always cheaper.
What if the company has nobody to manage the transition?
This is the most common case, and the reason so many migrations stall halfway. The transition needs someone with allocated time, not someone doing it "whenever they can" — whether that's an internal person with freed-up hours, or external support for the months it takes.
In summary
You don't replace Excel: you replace the role it ended up taking on without anyone deciding it should. Invoicing, the accounts receivable ledger and payroll should come off the sheet. Forecasting and budgeting should stay on it. And no new tool makes up for a process that still isn't written down.