Five signs a company has outgrown spreadsheets
Spreadsheets work well up to a point. These are the signals that the point has already passed, and why replacing them does not have to happen all at once.
Written by Ramzan Karya Persada
Almost every system we build replaces something that previously ran in spreadsheets, and that is entirely normal. Spreadsheets are cheap, immediate, and everybody already knows how to use them. For a small team with a small amount of data they are genuinely the right tool.
The problem is not that spreadsheets are bad. It is that they have a limit, and that limit gets crossed quietly, without anything announcing it.
Here are the five signals we see most often.
The five signs
- There is a file named final that is not the latest version. Usually accompanied by one named final v2 and another named final revised.
- One person has become the custodian. Work stops when they take leave, because only they know which file is authoritative and how the formulas fit together.
- The same number appears differently in two reports. Nobody can say which is correct without rebuilding both from source.
- Changes leave no trace. A figure is different from last week and there is no way to see who changed it, when, or why.
- Monthly reporting takes more than a full working day. Not analysing the numbers. Just assembling them.
One of these on its own is not necessarily a reason to replace anything. Three or more usually means the cost of leaving it alone has already passed the cost of fixing it.
The reason that goes unnoticed is that the cost of doing nothing is spread thinly across many people as wasted hours. It never appears as a single line in the accounts, so nobody ever sees the total.
It does not have to change all at once
The biggest worry is almost always operational downtime. That worry is reasonable, and it is also why replacements get postponed for years until something breaks badly enough to force the issue.
But a replacement does not have to be a single switchover. What works most reliably is this:
- Move the one process that causes the most trouble
- Leave everything else running exactly as it is
- Let the new piece prove itself for a month
- Move the next one
Each step is small enough to reverse, which means no single step carries the risk that makes the whole project frightening.
What actually changes
The visible gain is time back at month end. The larger gain is quieter: the work stops depending on one person remembering how it fits together.
A system nobody can explain is a system with a single point of failure, and that point of failure is a human being who is entitled to take a holiday.
That is the part worth paying for.
Read next
- When automation is worth the cost, and when it is notThree tests that decide whether a repetitive task deserves to be automated, applied before any budget is committed.
- A slow site is rarely a hosting problemUpgrading the hosting plan almost never fixes a slow website. Three causes that actually explain it, and how to check each one yourself.
Need this built rather than explained?
Send a short brief and RKP comes back with a written scope, a cost estimate and a schedule within two working days.