It is worth opening with a defence of Excel, because it usually comes out of this debate worse than it deserves. A spreadsheet is the most flexible tool a company will ever own: it can be built in an afternoon, needs no budget or approval, anyone can read it, and it adapts to a process that is still changing. A good share of the businesses we know run on spreadsheets for years, and they are right to.
The problem is never that a company uses Excel. The problem is that there comes a point where the spreadsheet stops describing the operation and starts governing it, and that shift happens without anyone announcing it.
How do you know the spreadsheet has run out of road?
There is no row count beyond which Excel breaks. There are behavioural signals, and they are fairly reliable when they show up together.
The first is somebody having to ask which version is the good one. As soon as a “dashboard_final_v3_REVISED.xlsx” is living alongside four other copies in three people's inboxes, the sheet has stopped being a source of data and become a rumour. The second is the operation stalling when one specific person is away, because only they know what formulas sit behind the tab nobody opens.
The third, and most expensive, is the company starting to make decisions on data that cannot be audited. When management asks why March's margin came out differently in two reports and the honest answer is “depends which sheet you look at”, the cost is no longer administrative. It is a management cost.
What actually fails is not the file
A spreadsheet has no concept of “who changed this and why”. All the traceability lives in people's memory.
Excel is not designed for concurrent work or for keeping history. Two people editing at once produce conflicts that get resolved by eye. An accidentally deleted formula cell leaves no visible trace and can take months to surface. And there is no reasonable way to say that Ana can see her branch's data but not everyone else's, short of splitting the file, which multiplies the original problem.
None of this matters when one person uses a sheet for a bounded task. All of it matters a great deal when the sheet is where inventory, cash forecasting or customer tracking actually lives.
Does that mean migrating to a management system?
Not necessarily, and this is the leap that wastes the most money. The usual reaction to an overflowing spreadsheet is to look for an ERP or a CRM to replace it, which means a tool change, a data migration, a training process and several months of reduced productivity. Sometimes it pays off. Often it does not, and there is a middle path nobody proposes.
That path means identifying which part of the sheet is genuinely critical, usually a small fraction, and moving only that part somewhere with access control, history and validation. The rest of the sheet can stay where it is. We have resolved situations that looked like they demanded a full ERP by moving three tabs into a small database and letting the team carry on working in Excel against that data, now reliable.
When is a full replacement the right call?
When the sheet is no longer maintained by whoever built it, when the process it supports has stabilised and is not expected to change much, and when the number of people who need to write to it simultaneously exceeds the number who fit in one conversation. That combination typically shows up somewhere between fifteen and forty employees, although headcount matters less than the volume of concurrent writing.
It should also be replaced, without much further debate, when the sheet holds personal data of customers or patients and circulates by email. At that point it is no longer a question of efficiency.
What it costs to move off Excel
Moving one specific process from a sheet into a system with access control and traceability usually lands between €3,000 and €9,000, depending on how many integrations are needed and what state the source data is in. The clean-up beforehand is almost always the line item the company did not expect: the historical data in a long-lived sheet carries years of mixed formats, duplicates and fields quietly used for two different purposes.
What is worth discarding is the idea that it all has to happen at once. A twenty-tab spreadsheet rarely needs to disappear. It usually needs three of those tabs to stop being a spreadsheet.
When what is missing is visibility rather than storage, the engagement is usually a dashboard. If the sheet coexists with other systems that do not talk to each other, the entry point is systems integration.