Warehouse management systems: holding stock truth by location

Stock records with enough dimensions to direct work

The stock record carries far more than quantity: a location, a handling unit or licence plate identifier, a batch or lot, sometimes a serial, a condition status such as available, quarantined or damaged, an expiry date where relevant, and an owner where the site holds goods for several parties. Each dimension exists because some instruction depends on it. Without status you cannot block stock; without batch you cannot recall it; without owner a contract warehouse cannot bill or report. Adding a dimension after go-live means re-labelling physical stock, so the model is chosen before racking is populated.

Directed work is the operating mode

Rather than telling staff what needs doing and hoping, the system issues discrete tasks tied to identifiers that get scanned back: receive against this expected consignment, put away to this location, replenish this pick face, pick these lines in this sequence, pack to this container, load to this door. The rules behind those tasks, put-away strategies, picking methods, allocation logic and replenishment triggers, are configuration rather than code in mature products. Their quality determines walking distance, which is the largest controllable cost in a manual operation.

The location model is the decision you live with

Location naming, zoning and capacity definition look trivial during design and become structural afterwards. Names encode aisle, bay, level and position in a sequence the pick path follows, so a naming scheme that does not sort into a sensible walk order produces permanently inefficient routes. Capacity by volume, weight and unit count determines whether put-away suggestions are usable. Sites that guessed here tend to disable directed put-away entirely, at which point the system is recording locations rather than managing them.

Where records meet customs, quality and finance

Goods held under a customs procedure such as customs warehousing or temporary storage carry record-keeping obligations to the supervising authority, and the warehouse record is usually the evidence trail. Consult the relevant customs administration for what a specific procedure demands, since requirements differ by country and by authorisation. The practical implication for system design is stable: the stock record needs an attribute distinguishing goods in free circulation from goods under suspension, and movements between the two must be deliberate transactions rather than a status edit that leaves no history.

Reconciliation with the enterprise stock ledger

The finance or planning system tracks stock by item and site for valuation and availability; the warehouse tracks by location and handling unit for execution. They will diverge, and the interface design decides how quickly that is discovered. Transactional interfaces that post each movement keep the gap small but flood the receiving system; periodic summary interfaces are quieter but hide errors for longer. Either way a reconciliation routine is not optional, and the operational question worth settling before go-live is which system wins a dispute, and who is allowed to force an adjustment.

Frequently asked questions

Is the stock module in an ERP enough for a small warehouse?

It can be, while the operation is small, product variety is low and staff know where things live. The threshold is usually reached when temporary staff must find stock unaided, when batch or serial traceability becomes contractual, or when picking errors start reaching customers.

Why do warehouse and finance stock figures disagree?

They are counted at different moments and at different granularity. Goods picked but not yet despatched, stock in a staging location, and adjustments posted in one system before the other are the usual causes, so a reconciliation routine with a defined tie-breaker matters more than trying to keep both perfectly synchronised.

Can one system run several warehouses with different processes?

Yes, and multi-site configuration is normal, but variation has a governance cost. Every site-specific rule is a permanent branch in testing and upgrades, so it is worth distinguishing genuine physical differences from habits that could be standardised.