Almost every piece of custom software we've built since 2018 has the same backstory: an Excel sheet that started small and at some point carries half the company. Here are the seven signs that tell you this point has been reached.
Upfront, so this doesn't come across wrong: we like Excel. It's the most successful piece of end-user software of all time, and a well-built spreadsheet is often the right solution. The line "Excel is not a database" is true — but it doesn't help anyone who's currently managing their warehouse with a 40 MB workbook, precisely because it has worked so far.
The problem is never the spreadsheet. The problem is the moment the worksheet has become a system — with multiple users, critical data and processes hanging off it. That moment is easy to miss because it creeps up. Hence the checklist.
The list doesn't come from a textbook, it comes from our first conversations: these are the sentences customers say themselves, shortly before they call us — just put into checklist form. The more of them sound familiar, the further the sheet has already travelled from tool to risk.
The seven signs
- The file exists more than once. "Warehouse_2025_final_v3_NEW.xlsx" isn't a filename, it's a cry for help. Once nobody can say for sure which copy is the truth, you no longer have data — you have versions of data.
- One person is the bottleneck. Only one colleague understands the macros, only one is allowed to touch the pivot. Holiday, illness, resignation — and the process stops. That's not an Excel problem, but Excel keeps it invisible until it blows up.
- The same data is maintained twice. Once in the sheet, once in the ERP or the industry software. Double maintenance means the two states drift apart, guaranteed — and at some point someone decides based on the wrong one.
- Transposed digits cost real money. As long as a typo only annoys people internally: fine. The moment it lands on an invoice, in an order or at a client, the sheet carries a responsibility it was never built for.
- Several people need it at the same time. Locked files, overwritten changes, "can you close it for a second?" shouts across the office. Concurrency is exactly where a worksheet stops being a tool.
- There's no history and no permissions. Who changed the price, when, why? Excel doesn't know. And anyone who can open the file sees everything — including the columns with purchase prices or salaries.
- Reports are projects. If the question "How many open orders do we have per customer?" means half a day of copy-paste, you're already paying for the missing software — in working time, every week again.
Our rule of thumb: one or two signs are everyday life, you can live with that. From three signs at once, the conversation about replacing it is worth having — not because Excel is "bad", but because from that point the sheet costs money weekly.
The cost calculation behind the spreadsheet
Signs are good, numbers are better. The calculation we open in first conversations is deliberately simple: how many hours per week go into maintaining, reconciling and troubleshooting around the sheet — summed across everyone involved? One hour of double maintenance a day sounds harmless, but that's five hours a week and over 200 hours a year. At €40 fully loaded cost per hour, the "free" Excel solution really costs over €8,000 annually — before the first transposed digit lands on an invoice.
That number belongs on the table before anyone talks about software. Because it answers the budget question at the same time: a small patch-sized tool often amortizes against that sum within one to two years — and all the risk the sheet carries beyond that isn't even in the price yet.
Why the moment is so hard to spot
Because every single crutch was rational. The second file came about because the colleague in the field needed a copy. The macro came about because retyping took too long. Each decision was right on its own — the sum is a system nobody planned that way, and that carries the daily business anyway. Which is exactly why it never shows up in any budget until it hurts properly once.
There's a human factor on top: the sheet usually has a proud author. Someone built it over years, refined it, kept it alive — and that someone quickly hears "we're replacing Excel" as criticism of their work. The opposite is true: the sheet proves the process is understood. The best role for its author is therefore not spectator but domain expert of the successor project. In our projects, that exact person regularly ended up as the new tool's strongest advocate.
What comes next — and what doesn't
The reflex "then we need a big system now" is almost always wrong. What replaces the sheet is, in most cases, a small, custom tool: a capture screen with a real database behind it, roles and permissions, history — and interfaces to what's already there. In our sizing logic that's a patch or a small module; the price ranges are public on our studio page.
One item to budget for without fail: importing the legacy data. Ten years of grown spreadsheet means duplicates, format breaks and three columns whose meaning nobody can explain anymore. That archaeology is honest work and belongs in the quote — whoever hides it brings it back later as a surprise.
Time-wise, such a replacement isn't a matter of quarters, by the way: a patch-sized capture tool is live in weeks, not months. In our experience the longest phase isn't the build but the decision before it — the months in which everyone knows the sheet is creaking and nobody wants to be responsible.
How the switch actually goes
The path from spreadsheet to tool isn't a big bang, it's four sober steps:
- Inventory instead of wish list. We look at the real spreadsheet — every column, every macro, every special colour. The sheet is the best requirements documentation there is: it contains what the process really needs, not what it theoretically might need.
- The core workflow first. Version one is deliberately allowed to do less than the sheet. It has to master the one path that runs daily — capture, change, find. The special cases follow once the everyday holds.
- Parallel operation, with an end date. For two to four weeks, sheet and tool run side by side. That takes the fear away and surfaces the cases nobody mentioned in the interviews. The end date matters — eternal parallel operation is double maintenance by design.
- The old sheet becomes an archive. Read-only, in a known place. Don't delete it — but never maintain it again either.
What Excel gets to keep
Even after the switch, Excel stays in the house, and that's a good thing. Ad-hoc analyses, one-off reports, the quick back-of-the-envelope before a decision — it's unbeatable for that. The line is easy to draw: Excel for thinking about data, real software for managing data that processes and people depend on. Whoever draws the line like that rarely has to move it again.
And if someone with exactly such a workbook has just landed on this piece: the list of signs above is an honest self-test. Zero to two hits — carry on, the sheet is your friend. From three hits, one hour of taking stock is worth more than the next macro will ever save.
The spreadsheet was never the problem. The problem is the day it's your only system.Whiteboard note from a kickoff workshop
