Outgrown your spreadsheets? When to replace them, and when not to

"We need to get off spreadsheets" is one of the most common lines in an operational review, and it is often wrong, or at least wrong as a first move. Spreadsheets are cheap, flexible and understood by almost everyone in the business. In a lot of cases, they are exactly the right tool for the job they are doing. The failure is not the spreadsheet. It is what has been quietly asked of it over time.

What a spreadsheet is good at, and what it is not

A spreadsheet is well suited to connective tissue: pulling together information from a few places, doing a calculation, giving one person or a small team a working view of something. It is poorly suited to being the single source of truth for a growing team, providing real-time alerts, enforcing an approval chain, or surviving being edited by ten different people with ten different habits.

Research backs this up. In spreadsheet development experiments, between roughly a fifth and most of the spreadsheets participants built contained errors, and even people working carefully got around 1 to 6 percent of their formula cells wrong. In day-to-day use, the same kinds of mistake creep in: a formula range not updated, a figure overwritten, a manual step skipped under time pressure. That is not an argument against spreadsheets. It is an argument against asking a spreadsheet to do a job it was never designed to do without any check on the way.

The question is never "should we still be using a spreadsheet." It is "what is this spreadsheet actually being asked to do, and does it still do that safely."

Signs a spreadsheet has outgrown its job

It is the only place critical information lives. If the file disappeared tomorrow, would anyone notice before it mattered, or would the business lose knowledge nobody else has.

It requires the same information to be typed in more than once. Repeated rekeying between a spreadsheet and another system is a strong sign the real problem is a missing connection, not the spreadsheet itself.

It is expected to provide something it cannot. Alerts when something is late, an audit trail of who changed what, a live view for people in another department. These are legitimate needs. They are not things a spreadsheet does reliably on its own.

Nobody remembers why a formula does what it does. A spreadsheet built up in layers over several years, by several people, with no documentation, becomes a risk in its own right, regardless of what it is used for.

An illustrative example

Consider a business (illustrative, not a specific client) using a macro-driven spreadsheet to combine several supplier reports into one inventory-planning view. On the surface it looked like exactly the kind of fragile, ageing tool a review should recommend replacing. In practice, it was doing a genuinely difficult job well, combining data that no single system in the business held in one place, and doing it in a way the team understood and trusted. The right call was to leave it alone and document how it worked, not to commission a new system to replicate it.

What this means in practice

Before recommending a spreadsheet be replaced, ask what it is actually being asked to do, whether that is a reasonable job for a spreadsheet, and whether the risk sits in the tool or in how it is being used. Sometimes the honest answer is a better spreadsheet, with clearer ownership and a documented structure. Sometimes it genuinely is a system. The two get confused far more often than they should.

If your spreadsheets have become the system nobody dares touch, see how I help with outgrown spreadsheets and systems.