JW Soft insight
Six signs your operation has outgrown its spreadsheet
Spreadsheets fail gradually and then all at once. Six concrete signals that a process has passed the point where a spreadsheet is the right container — and what to do next.
By Peter Basl ·
Spreadsheets are excellent, and most criticism of them is unfair. They are fast to create, universally understood, infinitely flexible, and require no procurement process. A large share of the world's important work runs on them, correctly.
But spreadsheets fail in a characteristic way: gradually, then all at once. The failure is rarely dramatic. It shows up as a slow accumulation of manual steps, tribal knowledge, and near-misses that everyone treats as normal.
Here are six signals that the container has stopped fitting the contents.
1. There is a person whose job is the spreadsheet
Not "someone who uses it" — someone who maintains it. They fix the formulas when a column shifts, they re-share it when permissions break, they merge the copy that someone edited offline. Nobody assigned them this role.
That person's time is the real cost of the spreadsheet, and it is invisible on any budget line. When you can name them, the spreadsheet has become infrastructure without ever being treated as infrastructure.
2. Multiple people need it simultaneously, and coordination has become a protocol
"Don't edit it, I'm in it." "I'll send you the updated version." "Use the one in the Drive folder, not the email attachment."
Concurrency is where spreadsheets structurally struggle. Cloud spreadsheets handle simultaneous editing better than the desktop generation did, but they do not handle simultaneous decision-making — two people can both mark the same slot as filled, and the sheet will happily let them.
3. The rules live in someone's head, not in the file
Ask what happens when a value is blank, or negative, or arrives late. If the answer is "well, you just know to..." then the business logic is not in the system. It is in a person, and it leaves when they do.
A spreadsheet cannot enforce a rule it doesn't know about. Software can refuse invalid input, require a reason for an exception, and record who approved it.
4. History is gone
Spreadsheets overwrite. When a cell changes, the previous value is typically unrecoverable in any practical sense. For anything where you might later need to answer "what did we know on the 14th, and who changed it?" — scheduling, finances, assessment, compliance — that is a serious limitation.
Version history in cloud spreadsheets helps, but reconstructing a specific record's state from document-level version snapshots is archaeology, not reporting.
5. Reality has stopped matching the sheet
The clearest signal. Someone updates the sheet after the fact, from memory, at the end of the week. Or people have started keeping their own side-copies because the shared one is never current.
When the sheet has drifted from being where the work happens to being a record of where the work happened, it has already been replaced — by group chats, by email threads, and by memory. It just hasn't been formally replaced by anything reliable.
6. The workarounds have workarounds
A second tab that reformats the first tab for a report. A monthly export that gets pasted into a different sheet. A conditional-formatting scheme that encodes status in colour, which nobody can filter on.
Each of these was a reasonable local fix. Collectively they are a system with no documentation, no tests, and no owner.
What to do about it
Separate what the spreadsheet does well from what it doesn't.
Spreadsheets are excellent at analysis, modelling, and one-off calculation. They are poor at concurrent multi-user state, enforcing rules, preserving history, and mobile access. Frequently, the right answer is software for the state and the rules, with an export that lets people keep analyzing in a spreadsheet — because that part was never broken.
Build for the point of capture, not the point of reporting.
The reason data drifts is that entering it is inconvenient at the moment it's generated. Ladders came out of exactly this: recurring pickleball groups were tracking sessions, results, and standings across group chats and sheets, and the data was always slightly wrong because nobody wanted to open a spreadsheet mid-session on a phone at a court. Fixing capture fixed reporting for free.
If several of these signals sound familiar, the next step is not to choose a tool. It is to write down what actually happens — see the five tests — and then decide.
A workflow worth examining?
