I have never met anyone who started their career saying, “One day, I’m going to build the entire operational backbone of a company in a spreadsheet.” It happens anyway. Almost always by accident. Someone — usually you — needs to track something. Excel is sitting there. You open it. You add a column. Then another. Then you colour-code things. Then you add some formulas. Then three other people need access, so you email it around with increasingly creative filenames. Before you know it, you’ve got “Final_FINAL_v3_ACTUAL_USE_THIS_ONE.xlsx” circulating through your company, and seventeen people are sort of editing it, and nobody is quite sure what the source of truth is anymore.
Spreadsheets are extraordinary tools. I will defend this to the end. They’re flexible, they’re intuitive, they’re forgiving. You don’t need to be a developer to use them. They’re one of the reasons ordinary people without technical training can do genuinely sophisticated data work. The problem isn’t spreadsheets themselves. The problem is when we confuse a spreadsheet with a system.
A spreadsheet is a document. A system is infrastructure. These are not the same thing.
Here’s how you know you’ve outgrown your spreadsheet. The first sign is multiple people editing it at once. Excel is not a multi-user editing tool. Google Sheets is slightly better but still not built for it. When you’ve got three people trying to update the same spreadsheet, you’re operating in a permanent state of version conflict and lost data. Someone saves over someone else’s changes. Someone makes an edit they think is in the master file but they’ve actually got an old version open. The source of truth becomes unclear. It’s just a question of time before you lose something important.
The second sign is version control via filename. If your file naming convention has become genuinely creative — if you’re using timestamps, or full words like “ACTUAL,” or multiple “final"s — you’ve lost the thread. You can’t remember which version is current. Nobody can. You’ve reinvented version control but worse, because humans are terrible at it and spreadsheets have no mechanism to enforce it.
The third sign is that no one person fully understands the spreadsheet anymore. You’ve got interdependent sheets. There are formulas that reference other formulas. Someone knows why that column is there, but that person doesn’t work here anymore. New people inherit it and add things without really understanding how it all connects. You’ve got what amounts to technical debt in cell form.
The fourth sign — and this one kills me because I’ve seen it so many times — is data entered twice in two different sheets. Or data that lives in the spreadsheet and also lives in the system you’re paying for. You’re maintaining two sources of truth because the spreadsheet does one thing and the system does another, and neither does everything, so nobody trusts a single version. This is when you know the spreadsheet has become a liability rather than a tool.
When you hit this point, something inside you usually breaks a little. You realise you’re spending operational energy maintaining a spreadsheet instead of doing actual operations. You’re chasing data entry errors. You’re asking people to remember to update a document. You’re managing a tool instead of managing a process.
Here’s the hard bit, though. Recognising that you need something else doesn’t make it easy to actually move on. Because the spreadsheet has one critical advantage: everyone knows how to use it. Excel or Google Sheets is already on everyone’s computer. There’s no learning curve. There’s no onboarding. There’s no setup time. The pitch for moving to a “real system” requires you to ask people to learn something new, to change their workflow, to go through yet another implementation project that will probably feel like it slows things down before it speeds things up.
And everyone will say, “But can’t we just keep using the spreadsheet?”
“We’re already using Excel. Can’t we just add another sheet?”
“Do we really need to spend money on a tool when we have this?”
These are not stupid questions. They’re just missing the point that a spreadsheet is not a system. A system has rules. It enforces data structure. It prevents conflicting edits. It creates an audit trail. It scales. A spreadsheet is a frontier town — you can do almost anything, but there’s no law enforcement, and you’re one person editing one document away from chaos.
The real barrier to moving away from the spreadsheet isn’t the tool. It’s that organisations are uncomfortable paying for things they can’t see an immediate dollar value in. And “not losing data” or “not spending three hours a week hunting down the right version of a document” doesn’t show up on a budget line in the same way that “tool subscription” does.
If you’re reading this and you’re in operations, or you inherited this mess, or you’re the person who started it and now you’re genuinely panicked — here’s what helps. First, document what the spreadsheet is actually doing right now. What workflows depend on it? What happens if it breaks? What decisions get made from the data in it? This is not bureaucratic. This is clarity.
Second, get honest about the cost of keeping it. The person-hours spent maintaining it, chasing version issues, re-entering data, hunting for mistakes. That’s real money. Compare it to the cost of an actual system. Usually you’ll find they’re not as far apart as you thought.
Third, pick something that does one thing well. You don’t need a massive ERP system. You need something that solves the specific problem the spreadsheet is creating. Usually it’s a tool designed for the exact workflow you’re actually doing.
And finally — because this matters — have a real migration plan. Don’t just stand up a new system and hope people switch. That’s how you end up with the new system AND the old spreadsheet, which is somehow worse.
Spreadsheets are not a systems strategy. They’re a tool for humans who don’t have access to a system. When you do, it’s time to move on. Not because spreadsheets are bad, but because you’ve outgrown them, and pretending you haven’t is going to cost you far more than the alternative.