Running inventory across multiple warehouses

A second building changes the math. With one warehouse, a single stock number per product is the truth: what the system says is what is on the shelf. The day you open a cross-dock in the next city, that number becomes a story. Forty cases of tomatoes might mean thirty in the main warehouse and ten across town, or the reverse, and the software cannot tell you which. Meanwhile the morning pick list still prints as one pile, so the crew in each building picks from a sheet that mixes both buildings' orders and spends the first hour sorting out whose work is whose.

If your team keeps a side spreadsheet titled something like "what is actually where," you have hit the wall this article is about.

When a second building starts to hurt

The symptoms show up in a predictable order:

• Stock is counted in one pile. The system says 40, building A says 30, building B says 10, and nobody trusts any of the three numbers by Thursday.
• Crews pick from mixed lists. Every order for the whole operation lands on every pick list, so each warehouse crosses out half the sheet before pulling a single case.
• Orders do not say which building owns them. Staff eyeball the delivery address, guess, and occasionally load a pallet onto the wrong truck.
• Transfers are a text message. Ten cases move between buildings in a van, and the record of it is a phone photo, if anything.

None of this is a people problem. It is what happens when produce distribution software built for one location keeps running an operation that now has two.

What per-location operation looks like

The fix is to make the buildings real inside the system. Each warehouse becomes a named location, and the pieces fall out from there:

• Each warehouse holds its own counts. A tracked product carries a stock number per building, not one global figure. Availability for an order is judged against the building that will fulfill it, and a sale decrements that building only.
• Delivery zones are assigned to the building that serves them. You already know which warehouse covers which ZIP codes; you encode that once, and routing follows it from then on.
• Every order is stamped with its fulfilling location and the reason. Staff open an order and see not just "Building B" but why: the address matched a zone assigned to B, the buyer chose B for pickup, or nothing matched and it fell back to the default. No guessing, no tribal knowledge.
• Pick lists and manifests filter per building. Each crew prints a sheet containing exactly the orders routed to their building. Totals across both filtered sheets still equal the full day.
• Transfers between buildings are recorded and audited. Moving ten cases from A to B decreases A by ten, increases B by ten, and logs who moved what and when. The product's total never changes on a transfer, which is the property that keeps counts honest.
• Low-stock warnings fire per building. Twenty cases in building A is no comfort when building B's zones are ordering against zero. Warnings watch each building's count separately.

Three boundaries are worth stating plainly, because they are decisions and not accidents. One location fulfills each order; there are no split shipments, so a single order never ships half from each warehouse. Routing follows your zone assignments, not distance; the system never overrides your judgment with a mileage calculation, because you know that the highway makes the "farther" warehouse the faster one. And the catalog and prices stay the same at every location; buildings differ in stock and coverage, never in what you sell or what it costs.

Pickup at more than one counter

If both buildings offer customer pickup, buyers choose which one at checkout, and the confirmation and order emails carry that location's address and any note like "loading dock, ring bell." If only one location offers pickup, no extra step appears and checkout stays exactly as short as it is today. The choice exists only when there is genuinely a choice to make.

What buyers see

Almost nothing, which is the point. A delivery buyer sees one availability number for each product and places orders exactly as before; the right warehouse is chosen for them from their address. They are never shown a per-building breakdown, never asked to pick a warehouse, and never told that their carrots are "in stock at location 2." Your operational complexity stays yours.

Adopting it without a cliff

The transition matters as much as the destination. In Minori Midori, multi-location is opt-in: until you add a second location, everything runs exactly as it does today, and when you do, all existing stock is attributed to your default location so no count is lost or zeroed. Orders for unassigned zones and parcel shipments route to the default location rather than failing, so nothing is ever blocked while you finish mapping zones. Multi-location ships on the Enterprise plan, with per-building stock, audited transfers, and the per-location pick list and manifest covered on the multi-location page. If your "what is actually where" spreadsheet is due for retirement, book a demo and route a two-warehouse day through it.