Warehouse

Because knowing how much stock exists is useless without knowing where it is, and every fulfilment decision is a question about place before it is a question about quantity.

The warehouse structure decides what the business can promise. Availability, picking, replenishment and stock-take all resolve against it, so the granularity chosen at implementation sets the ceiling on how precisely anyone can answer 'can we ship this today?'. Too fine and the data rots because nobody maintains it; too coarse and the system cannot answer the question at all.

• Stock always sits in exactly one warehouse. Quantity without a location is not stock, it is a rumour.
• A warehouse has a physical address, and in India that address determines which GSTIN governs movements in and out of it.
• Moving stock between warehouses is a transaction, not an edit — even when both belong to the same company.

The hierarchy gets designed for a warehouse manager's mental model rather than for the transactions people actually post. It looks right in a workshop and is unusable at the receiving dock, where someone with a trolley has to pick a location in seconds — so they pick the first one, every time, and the structure quietly stops meaning anything.

• Stock in transit between two of your own warehouses — owned by you, present in neither, and invisible unless modelled.
• Consignment stock at a customer site, which you own but do not hold.
• Third-party logistics warehouses, where the location is real but the operator is not you.
• Warehouses in different states, where an internal transfer becomes a taxable supply requiring an invoice and an e-way bill.