Accuracy is the prerequisite
We are regularly asked for demand forecasting by businesses whose recorded stock differs materially from their actual stock. No model corrects for that; it produces confident recommendations from wrong inputs.
Count twenty high-value lines against the system today. The variance tells you which project you actually need.
Capture at three moments
- Goods in — scanned against the purchase order, discrepancies flagged at receipt rather than discovered later
- Issue or pick — scanned out, so the system reflects reality within minutes
- Dispatch — confirmed against the order, catching picking errors before they become complaints
Barcode or QR capture on a phone is enough for most SMEs. The technology is unremarkable; the discipline it enforces is the point.
Reorder points that move
Static reorder points set two years ago are wrong now. We calculate from recent usage and supplier lead time, review monthly, and alert rather than order automatically until you trust it.
Lead time is the variable people forget to update. A supplier who has quietly slipped from five days to twelve makes every reorder point in your system wrong.
Multi-location visibility
Several sites, vans or channels each holding stock creates the classic problem: unavailable here, sitting unused there. A single view with transfer suggestions recovers stock you already own.
For field service and multi-site businesses this is frequently the highest-return element, because the stock is already paid for.
What it costs
Capture with scanning typically £10,000–£25,000 including devices. Reorder automation is usually a smaller addition once capture is reliable. Forecasting is a separate project and a later one.