How far can a spreadsheet take you?

A spreadsheet is not a tool to look down on. Plenty of businesses run on one for years, and they should. The problem is not that spreadsheets are bad; it is that past a certain point they become the wrong tool.

That point usually arrives quietly. Nobody wakes up one morning and declares the spreadsheet insufficient. Instead small failures accumulate: a file gets corrupted, a report comes out wrong, two people fill in the same row differently. The following are the signs you have arrived.

1. Several versions of the same file are circulating

"report_final.xlsx", "report_final_v2.xlsx", "report_final_FINAL.xlsx". If which one is current is a matter for debate, you have lost your single source of truth.

This is not something a spreadsheet can solve. It is inherent to file-based work: a file is a copyable object, and the moment it is copied there are two truths.

The symptom hides in these sentences: "Can you send me the latest version?", "Which one do you have?", "It does not look like that in mine." If those come up several times a week, this is not a file naming problem.

2. Two people cannot work in it at once

Cloud versions partly solve this. But in files with complex formulas and many sheets, simultaneous editing is still problematic: calculations lag, one person overwrites another's row, linked cells break.

If someone says "can you close the file, I need to get in," you have hit the limit. That sentence also carries a hidden cost: the person waiting is not working.

3. You cannot control who sees what

The sales team should only see their own customers, but everyone in the file sees everything. Hiding sheets or setting a password is not real access control: passwords can be cracked, hidden sheets can be unhidden, and the whole file can be copied in one click.

For any process handling personal data, this alone is reason enough to move. Under data protection rules you have to be able to answer "who accessed which data," and a spreadsheet cannot tell you. A related framework is covered in the KVKK compliance checklist.

4. There is no history

A cell changed. Who changed it, when, and what was the previous value?

Spreadsheets have change tracking, but it does not get used in practice. In complex files the output becomes unreadable, and the history disappears the moment the file is copied.

For processes subject to audit this is mandatory. Price changes, stock adjustments, and discount approvals all need an answer to "who did this?"

5. Nobody understands the formulas any more

The person who built the file has left, and nobody who remains dares touch the formulas. When a new column is needed, everyone looks at everyone else.

This is the spreadsheet version of the legacy system problem, and it carries the same risk. The one difference: in software, at least the code can be read and tests can be written. Nobody can test a nested lookup chain in a spreadsheet.

6. The same data gets typed into other systems by hand

If you are copying from the spreadsheet into accounting and from accounting into e-commerce, you are simulating an integration layer with human labour.

That has two costs. The first is time. The second, more expensive one, is error: manual transfer never has a zero error rate, and where the mistake happened cannot be found afterwards. API integration exists precisely to close that gap.

When staying put is the right call

When we hear these, we say do not build a system:

• One or two people use the file.
• The process changes often and flexibility gains you speed.
• What you are doing is analysis, testing scenarios and building models. Spreadsheets are still excellent at that.
• The data volume is small and no growth is expected.
• The process is temporary. Building a system for three months of work is waste.

If most of those apply, you do not need software. In those cases tidying the file, keeping it in a single cloud location, and reducing the number of columns is usually enough.

How to measure the cost of staying

Before deciding, record the following for two weeks:

1. How many times an argument about which version is current came up.
2. How many minutes a day went to manual copying.
3. How many errors were found and how long each took to fix.
4. How many people waited for the file to be free.

Those four numbers make the whole decision. When the discussion is based on guesswork, everyone defends their own impression; a measured table ends the argument. The same measurement method is covered in detail in where to start with automation.

How to make the transition

Do not switch the spreadsheet off all at once. Run the new system in parallel for a while. When both sides agree, make the switch. The parallel period usually runs two to four weeks.

Do not migrate everything. Most of the columns accumulated in a spreadsheet over the years are no longer used. Migration is a good opportunity to clean up. Looking at which columns were filled in over the last six months is enough of a filter.

Start with one process. Pick the file that hurts most and move only that one. Projects that try to move every file at once drag on, and the team gets tired and reverts.

Leave analysis in the spreadsheet. Moving data entry and process tracking into a system while leaving an export for analysis is usually the most practical answer. Trying to prise everyone off spreadsheets creates unnecessary resistance.

Migrate history, but only some of it. The last year or two of data is usually enough. Keeping older data as an archive file substantially lowers the migration cost.

The point

Leaving a spreadsheet is not a technical decision, it is a threshold decision. If three or more of the six signs above are true for you, the file is no longer making the work easier; it has become the work.

---

We can work out which of your files is worth moving into a system. Our approach is on the business process automation page.