Does Your Small Business Really Need an ERP?

Headcount is a poor signal. The signal is a failure between departments that handle one order: sales promises a delivery date the warehouse cannot meet, and production hears about it from the customer. Finance then spends days each month reconciling figures that three spreadsheets all claim are correct, while nobody can say which one is wrong. A purchase order raised on a Tuesday should not surprise the stores team on a Friday, yet that is what happens when each department keeps its own copy.

Take a garment maker as an example. Sales keeps orders in a spreadsheet, production works from a whiteboard, and accounts sit in a standalone package, so the three disagree at month end. No single tool is at fault. Nothing connects the three. Meanwhile the cutting room commits fabric rolls to one buyer’s order while merchandising promises the same rolls, trims and zips to another.

Count who touches one transaction. Count the times someone re-types data that already exists elsewhere, such as an order keyed into accounts again or a stock count repeated because two systems disagree. A few handoffs a week are a nuisance. Dozens of handoffs, each carrying a customer promise or real money, form a coordination problem that no amount of diligence will remove. Dyeing and washing jobwork leaves the building and returns with different quantities, which a chalkboard cannot reconcile and a month-end spreadsheet reconciles only after the shipment has left.

Next, list the tools doing overlapping jobs. One tool for billing, another for inventory, a spreadsheet for orders and WhatsApp for internal updates add up to an informal, fragile ERP that nobody designed or maintains. That pattern says more than revenue or headcount.

Dispatch errors show it clearly. Track how many invoices were corrected after dispatch, because every correction marks a point where two systems held different quantities, rates or tax codes for one sale. Frequent corrections are the evidence.

ERP is not a universal cure. If the accounting package cannot produce an aged-receivables report, or purchase approvals run on email chains, one better package or automation fixes that pain sooner than a rollout.

Ask where the problem lives. A fault inside one department’s tool rarely justifies replacing everything, whereas a gap between departments usually does, because the fix needs several teams looking at the same record. Everyone must read one stock figure.

Reporting gaps are the clearest example. A missing margin report usually needs a reporting add-on, whereas a missing link between quotation, production order and dispatch note needs shared data.

Check the data first. Master data, meaning the item list, customer list and price list, needs a named owner and a cleanup before migration, since duplicates simply move into the new system. A reorder level, a unit of measure and a tax code must mean the same thing in sales, stores and accounts, or the system exposes the disagreement faster.

Readiness also needs a sponsor senior enough to settle arguments between department heads who run the same process three different ways without noticing. Block weeks for testing. A project without a sponsor stalls at the first hard trade-off, usually when two departments want opposite settings for the same field.

Put the findings together. Real cross-department failures, overlapping tools that disagree, and a clean-enough dataset with a sponsor who has authority all point one way: ERP is worth starting now. Contained pain means a smaller fix.

“Not yet” is a legitimate answer. Firms that spend a few months tightening one process and cleaning master data usually see a calmer rollout than firms that start while everything is in flux.

Run the scorecard as a group. Ask each department head to score independently before comparing, since the gaps between their answers often reveal the coordination failures faster than any workshop.