Separate stock on hand from stock available

Goods physically present may be reserved, damaged, in quarantine or owned by a supplier. Agree which quantities are available to sell and which are excluded. A transfer should not appear at both locations while it is in transit. Define these states with the staff receiving and dispatching stock, then use the same definitions in every report.

Use one example through the full cycle

Use a real product with a size, colour or batch if that matters to your operation. Receive it, reserve some for an order, transfer some to a branch and return part of a sale. Ask the vendor to explain the resulting balance without editing the database manually.

MovementQuestion to test
Purchase orderDoes this record expected stock without adding it prematurely?
Goods receivedCan staff receive less than the order and keep the balance open?
Branch transferWho confirms dispatch and receipt?
Customer returnDoes an inspection decide whether it can be sold again?
Stock adjustmentIs the reason, approver and original value retained?

Make branch access deliberate

Decide whether a branch can see other branches' quantities, costs or customer details. Headquarters may need a consolidated view while local staff need only their own work. Test this with separate accounts. Barcode support also needs testing with your actual labels, scanners, printers and units of measure.

Compare total operating effort

A subscription may fit standard stock flows well. Custom software becomes worth considering when important allocation, consignment, approval or integration requirements remain awkward. Compare migration, hardware, staff training, hosting, maintenance and export capability alongside the purchase price. Avoid selecting an ERP merely because you have many SKUs.

Run a small, reconciled pilot

Choose a location and a representative product set. Agree the opening count and record every movement during the pilot. Investigate differences before expanding. Track time spent reconciling, unexplained adjustments and stock availability errors. These measures tell you more than a polished demo using perfectly clean data.