Almost every operation starts in a spreadsheet, and most of them should. It is quick to build, everyone can read it, and it changes the day your process changes. The question is not whether spreadsheets are bad. It is when the one you have stopped being a convenient tool and became something the business depends on.
This article is for someone who runs operations and has a feeling that a particular sheet is "getting out of hand". It gives four signals to check and a simple way to decide. It is about the decision to switch, not about how to do the switch.
Why the decision is hard
A sheet grows slowly. A column is added here, a tab there, and one day it holds orders, prices and deadlines for the whole team. Nobody chose that. Because it never failed in a dramatic way, there is no moment that says "now".
Switching too early is also a cost. A small tool needs a person to look after it, and a sheet that three people use once a week does not need one. So the aim is a clear, repeatable test rather than a general preference.
Four signals
Look at the last month and count how often each of these happened.
1. Errors repeat
A wrong number now and then is normal. Look for the same kind of mistake: a price typed over a formula, a row sorted apart from its neighbours, a date in two formats. When you fix the same sort of error for the third time, the sheet allows what a tool would prevent. A form with required fields and a fixed list of options removes this class of error.
2. People wait
Check whether work stops because the file is locked, because someone must copy the latest version, or because one person has to "send the numbers". Waiting is hard to see because it looks like normal work. Ask the team where they wait for the sheet, and write down each case with roughly how long.
3. More than one person edits it
One editor and many readers is a good fit for a sheet. Three people changing the same area is not. Conflicting copies, "final_v3" files and changes nobody can trace are the typical symptoms. If you cannot answer "who changed this cell and when" for something that matters, you have this signal.
4. One person is the only one who understands it
If the sheet works only because one person knows which tab to touch first and which formula not to break, the sheet is a risk for the company, not a document. A simple test: if that person were away for two weeks, would the work continue? If not, the signal is clear. See also a one-page handover for an internal tool for what to write down in either case.
How to read the count
A simple scoring is enough.
| Signals that happened in the last month | What it suggests |
|---|---|
| None | Keep the sheet. Review in three months |
| One | Clean up the sheet: fixed columns, drop-down lists, one named owner |
| Two | Run a small trial of a tool for one narrow part of the process |
| Three or four | Plan the switch, starting with the part that hurts most |
The numbers are a guide, not a rule. A single signal with a high cost, such as a wrong price reaching a customer, can be enough on its own. Write down the reason when you go against the table.
The first step is cheap
Before anyone builds anything, do these three things in a week:
- Name an owner of the sheet. Not an editor, an owner who decides what changes.
- Fix the structure. One row per item, one meaning per column, lists instead of free text where the values are known.
- Write down what the sheet is for: who puts data in, who reads it, and what decision depends on it.
This costs little, and it is useful in both cases. If you stay with the sheet, it is better. If you move to a tool, you have already done the analysis that most projects begin with. For a longer description of the same idea at larger scale, manufacturing's spreadsheet problem shows how it looks when the sheet has grown into the main system.
What "small tool" means here
It means one narrow function, such as an order intake form with a status list, or a shared tracker with fixed fields and an audit trail. It does not mean a large system. Pick the part of the sheet that triggered most of the signals and leave the rest as it is. If you decide to go ahead, moving from a spreadsheet to a small internal tool in 30 days describes the steps, and build or buy: five questions helps with the choice of approach.
Common mistakes
- Deciding by feeling. Without a count, the loudest complaint wins.
- Switching everything at once. A narrow first step is easier to judge and to reverse.
- Treating the sheet as free. Time spent on corrections and waiting is a real cost, even if no one invoices it.
- No owner. Without one, neither the sheet nor the tool will stay in order.
- Never looking again. Staying with a sheet should be a decision with a review date.
Checklist
- List the sheets your operation depends on
- For each, count the four signals over the last month
- Use the table to choose: keep, clean up, trial or plan
- Name an owner and write down what the sheet is for
- Fix structure before any build decision
- Set a date, ideally within three months, to count again
Sources
- Our related articles linked above.
- The four signals and the scoring table are a working practice, not a standard; adjust them to your context. No client results or market figures are claimed in this article.
