Multi-Store Retail Inventory: Size Runs, Transfers, Dead Stock

A single store can run its inventory on the owner’s eyes. Walk the floor, look at the back room, reorder what is getting thin. It works surprisingly well, right up until there are two stores, and then three, and the owner’s eyes are in one building while the stock is in several.

At that point the problems stop being about how much inventory you have and start being about where it is and what shape it is in. The numbers most retailers watch are blind to both.

Total units is the wrong number

A style showing seventeen pairs in stock looks healthy. It is not healthy if none of those seventeen is a size 7, and size 7 is what walks in the door.

That is a broken size run, and in apparel and footwear it is the most common way inventory lies. The count says in stock; the customer hears we do not have your size. The sale is lost, and worse, the style’s sell-through data now says it is not selling, so it is less likely to be reordered — which makes the problem permanent.

The fix is to measure the run rather than the total. Identify the core sizes for each category — the sizes that account for most of the demand — and flag any style that is missing one of them in any store, regardless of how many units remain elsewhere in the run. In the franchise portal we built for Milly’s, broken runs are surfaced explicitly, with core sizes weighted more heavily than the edges.

The stock is in the wrong building

The second failure is that the size a customer needs frequently exists — in another store.

Without network visibility, the associate does not know that, so the customer leaves. With it, the associate can see the pair twelve miles away and raise a transfer before the customer reaches the parking lot. The difference is not a technology question so much as a habit: stores need to see each other’s stock and have a routine, trusted way to move it.

Two details make transfers work in practice. First, record the request even when the transfer does not happen. A customer asking for a size you do not have is demand, and it is invisible in sales data because no sale occurred. Second, make the transfer a real movement of stock — out of one store, in transit, into another — so the counts stay honest while the pair is on the road.

On-hand should be a sum, not a typed number

Most small retailers keep a count that people overwrite. Somebody does a stock take, types in the new figure, and the history of why it changed is gone.

The more durable approach is a ledger: every receipt, sale, transfer, return and adjustment is a movement, and on-hand is the sum of those movements. When a count looks wrong, you can trace it to what caused it — a transfer that left but never arrived, a receipt entered twice, a shrink adjustment nobody explained. An overwritten number cannot tell you any of that. It just tells you the new number.

Dead stock needs a plan before the season

Every retailer buys some styles that do not sell. The expensive part is not the mistake; it is how long the stock sits before anyone acts on it, tying up cash and floor space in the wrong building.

Decide the markdown ladder before the season starts rather than in a panic at the end of it. For example: a style that has not reached a given sell-through by a set number of weeks takes a first markdown; if it still has not moved a few weeks later, a deeper one; after that it is consolidated to one location or cleared. The specific thresholds depend on your margins and your categories. The point is that the rule exists in advance, so markdowns become a routine decision rather than an emotional one.

Transfers help here too. A style that is dead in one store is sometimes selling in another, and moving it is cheaper than marking it down.

Compare stores honestly

Once there are several locations, head office wants to know which ones are struggling. The trap is a scorecard that treats missing data as a bad score.

If a new store has no sell-through history yet, or a location has not completed an audit, a scorecard that counts that as zero will rank the wrong store last and send attention to the wrong place. The better rule is to drop a component a store has no data for and reweight the rest, and to show clearly which measures are missing. A dashboard should say when it does not know something. Silence and zero look identical on a screen, and only one of them is honest.

Where to start without buying anything

Before any system, run one weekly exercise: for your top styles, record whether each store holds every core size. That single sheet will show broken runs, surface the transfers you should be making, and tell you whether your inventory problem is quantity or distribution.

If it is distribution — and in multi-store retail it usually is — the operating fix comes before the software. That is the sequence behind our retail and franchising work, and it connects to the broader point in what has to exist before you franchise: a system has to be documented before it can be run across locations. If you would like a second set of eyes on which problem you have, book a conversation.