From spreadsheets to an ERP: when it is time to move

Almost every company starts with spreadsheets, and that is the right decision.

They are fast, cheap, familiar to everyone and require no implementation at all. In a small company they work well, and rushing to replace them is money spent for nothing.

The difficulty lies elsewhere: the moment spreadsheets start costing you is hard to spot. They do not break suddenly - they become expensive gradually.

When spreadsheets are enough

Software companies rarely write this section, because it works against sales. But it saves money.

Spreadsheets work well when:

• One person maintains them, or only one edits at a time
• Headcount is under twenty
• The catalogue holds fewer than a hundred lines
• History is not needed - only the current state matters
• The process is still forming and changes often

The last point matters. If the business model changes every month, fixing it into a system is premature. At that stage a spreadsheet is flexible, and it is the right choice.

Seven signs it is time

1. The file exists in several copies. "Stock_new_2.xlsx", "Stock_final_latest.xlsx". Nobody knows for certain which one is correct.

2. Two people edit at once. One person's changes disappear with the other's. Cloud versions solve this partly, but stock figures still come out wrong.

3. Formulas break. Someone deleted a row and three calculations started producing wrong numbers. It can take months to notice.

4. Nobody knows who changed what. A figure changed, but who altered it and why is unknown.

5. Assembling a report takes days. Every month data is pulled together from several files.

6. The same data lives in several places. Customer details sit in the sales file and in accounting, and they do not agree.

7. The file has become slow. Tens of thousands of rows have accumulated and opening it takes minutes.

Three of these mean it is time to think about moving. Five mean you are already losing money.

What spreadsheets cost

They look free, because the cost is not in expenditure but in losses.

A worked example: twenty hours a month assembling reports at $4 an hour is $80 a month, or $960 a year. Errors and lost time come on top.

In a company of fifty, the figure usually lands between $3,000 and $8,000 a year. That is a fraction of an ERP's price - but it repeats every year.

Intermediate options

Moving straight to an ERP is not compulsory. There are two intermediate options and sometimes they are enough.

A cloud spreadsheet. This solves simultaneous editing and adds a change history. But formulas still break and the permission model is limited.

A single module. Inventory only, or CRM only. It is not an ERP, but it closes the biggest pain point and costs less.

When an intermediate option is enough: if the problem sits in one process only. If it spans three or four, separate tools do not talk to each other and you end up with the same problem again.

How the move runs

The second step takes the longest and is skipped most often. Put messy data into a system and you get a digitised version of the mess.

What migrates and what stays

Answering this saves a great deal of time.

The last one causes the most debate. Migrating full history is expensive and rarely needed. The practical answer: old files stay in an archive and are opened if required. Only balances and open transactions go into the system.

Working with staff

Resistance is stronger here than on most projects, for a simple reason: everybody knows spreadsheets.

The new system feels slower at first, and that is true - in the first week every action takes longer. Say so in advance.

What works:

Show a first result quickly. Within a week something should be working - stock visible on screen, for instance. That builds confidence.

Keep the export to spreadsheet. If people can pull data out into a familiar format, they relax. It is a temporary bridge and it lowers resistance noticeably.

Consider who will actually use it. The interface has to suit the storekeeper as well as the accountant.

In summary

Spreadsheets are not a bad tool. They simply stop working once several people and a need for history arrive.

Practical steps:

1. Check the seven signs - three matches mean it is time to think
2. Convert the time spent on reports and corrections into money
3. If the problem is in one process only, consider an intermediate option
4. Clean the data; that is work before the system
5. Start with two processes rather than everything at once
6. Archive the spreadsheet in month three