In this guide 6 sections
Staff keep an Excel sheet after go-live because it still does something the ERP does not: a missing report, an awkward workflow, a locked permission or lost trust. Banning the sheet rarely works. The fix is to find out which of those reasons applies to each sheet, then close that specific gap before anyone retires the file.
What the side spreadsheet is doing
Find out what it is for. Pull a copy, sit with the person who maintains it, and ask which numbers come from the ERP, which are typed by hand, and why. Most shadow spreadsheets do not duplicate the whole system, because each one solves a single narrow problem that the ERP handles badly. Treat a clever hack as evidence. Spreadsheet sprawl is a symptom, not the disease.
The purposes repeat from one business to the next, and these three turn up wherever a shadow sheet survives the go-live month after month.
- A forecast or pipeline view that the ERP’s standard reporting cannot produce, usually maintained by sales or finance to answer questions about next month.
- A working draft, kept until the numbers are final enough to enter formally and the person is willing to put their name on them.
- A personal cross-check, kept because an ERP number was once wrong in front of the team and nobody has fully trusted that report since.
A missing feature versus a clunky one
One sentence hides two problems. Sometimes the ERP truly cannot produce what is needed because no report, field or workflow exists for it, and only a build will fix that. Other times it can, but the task needs a long run of clicks while the spreadsheet needs one paste, so people default to the easier tool.
The second case is a usability problem. Do not dismiss the sheet as a kludge. It usually has a cheaper fix than replacing the ERP or adding bloat through customisation: a saved report template, a dashboard view or a small configuration change.
Permissions, trust and training
Sometimes the ERP can do the job, but the person was never given the access or training for that module, so they built a workaround. Compare each role’s permission level with what the role has to do each week, because a missing working login for one module can explain a whole spreadsheet.
Trust is harder to see. Ask whether an ERP figure was ever visibly wrong in front of this team and never explained, because one unresolved error can outlast the fix that corrected it. The sheet becomes a safety net. Each spreadsheet is an island: a figure changed there never reaches the ERP, so two versions of the truth drift apart week after week.
Fixing the gaps worth fixing
Suppose a manufacturer went live two years ago, yet its sales team still runs the pipeline in a personal spreadsheet and updates the ERP only at month-end. Investigation finds that the ERP has no view of deals by expected close date across a rolling six weeks, and sales cannot plan without it.
That is a validated gap worth fixing, and it needs a small report addition rather than a re-implementation, so the cost is one report, not a second project. Not every sheet earns that investment. Some prove redundant, a personal preference, or covered by an existing ERP report nobody was shown, and those need a training or communication fix instead.
Retiring the spreadsheet in a controlled transition
Never retire it by decree. Once the real gap is fixed, run both in parallel for a defined period, so the owner can confirm for themselves that the ERP produces numbers they trust. A few weeks is usually enough.
Set the date in advance. Then check that the sheet really stopped being updated, because a file reappearing six months later usually signals that the underlying gap was never closed. It is rarely defiance. Treat a trusted file held together with duct tape and one person’s memory as a risk to the business, not as a nuisance.
Root-cause tree for shadow spreadsheets
A decision tree that starts from the symptom ‘staff still use a spreadsheet’ and runs four checks: capability gap, usability friction, permissions or training, and trust in numbers. Each branch ends in a different fix. No branch gets a generic reply.
Open a blank worksheet to printQuestions owners ask
Should we ban spreadsheets?
No, because banning a sheet without fixing what it compensated for only pushes the problem into a personal file nobody can see. Close the real gap first. Once no reason remains to keep the file, it usually stops being needed without any order from management and without anyone losing face.
Can reporting gaps be fixed without replacing the ERP?
Often, yes, because many reporting complaints trace back to a missing view, dashboard or saved report rather than a structural limit of the system. Those are typically a configuration or small development task, well short of a re-implementation, and a sketch of the missing view makes the scope easy to price.
How do we know the ERP numbers can be trusted?
Run the spreadsheet and the ERP in parallel across a full month-end close and have the sheet’s owner reconcile every difference line by line. Explained differences rebuild trust.