How to handle an out-of-stock item at the counter without losing the sale

A foodservice distributor with a walk-in counter attached to its depot had a rule nobody had written down but everyone followed: if it's not on the shelf, it's not for sale. A chef would come in needing forty pounds of a specialty cheese for a weekend menu, the counter rep would check the shelf, see three blocks instead of ten, and say the same thing every time — "we're out, try next week." The chef would walk, usually to a competitor, and the distributor would log nothing. No record of the lost sale, no signal to purchasing that demand had outrun supply, nothing to connect the empty shelf to the order that could have refilled it. The depot just quietly lost the business and moved on, three or four times a week, without ever seeing the pattern add up.

The sale that looks like it never happened

The owner only noticed because a regular customer mentioned, almost as an aside, that he'd stopped bothering to ask for the cheese at all — he'd switched suppliers for that one item and just never switched back for anything else either. That's the real cost of treating "out of stock" as a dead end instead of a decision: it doesn't show up as a cancelled order or a return. It shows up as a customer who quietly starts buying somewhere else, one SKU at a time, until the relationship is gone and nobody can say exactly when it happened.

Digging into it, the pattern was worse than one SKU. The counter had no way to tell the difference between three kinds of "out of stock," and it treated all three the same way — turn the customer away:

• Out of stock, but a shipment is three days out and the quantity is already committed to this depot.
• Out of stock, with no reorder placed yet, but the vendor can fill an order inside a week.
• Out of stock, and the item is being discontinued or the next shipment is weeks away.

Those are three completely different conversations with a customer. The first is "yes, Thursday." The second is "yes, but it'll take a week — want me to hold it?" The third is "no, but here's what else works." A counter system that can't tell them apart collapses all three into "no," and "no" is the one answer that reliably costs a sale.

Why the fix isn't just "track inventory better"

The instinct is to say the depot needed better inventory visibility, and that's part of it — but the harder problem was upstream of the shelf count. Their POS had no concept of a backorder at all. An item's quantity on hand was either enough to sell or it wasn't; there was no third state for "sell it anyway, tag it as owed, and fulfill it when the truck arrives." So even when purchasing already knew stock was three days out, that information lived in a spreadsheet the counter rep never saw. The person facing the customer and the person who knew the real ETA were two different people with two different systems, and the customer paid for the gap between them.

The second problem was that nothing captured demand for an item you can't currently sell. Every "we're out" conversation at the counter was a real, immediate signal that this item needs reordering sooner, and every one of those signals evaporated the moment the rep said it out loud and moved to the next customer. A reorder point calculated only from sales history will always under-order for an item that's selling customers away before it can log a sale.

What changed at the counter

The fix didn't require new hardware, just a different flow at the register, built around three rules:

Make backorder a real state, not a workaround. When a line item exceeds what's on hand, the counter rep sees the committed inbound quantity and date right there in the sale screen — not in a separate spreadsheet — and can accept the order as a backorder with a real ship date instead of defaulting to "no." The customer leaves with a receipt that says "partial now, remainder Thursday," not with nothing.

Default to the honest answer, not the optimistic one. If there's no inbound shipment and no confirmed date, the rep doesn't guess. They either offer a substitute SKU the system already knows is a reasonable swap, or they log the request against that item so it counts as demand — even though no sale happened.

Let every declined or backordered line feed the reorder decision. A request that couldn't be filled gets recorded exactly like a sale would for purposes of demand tracking, so the next purchase order for that vendor reflects what customers actually asked for, not just what the depot managed to sell.

None of this eliminates stockouts — vendors still run late, demand still spikes unpredictably. What it changes is whether an empty shelf costs you the sale, the customer, or just the one item.

Takeaway: what to check at your own counter

• Can your POS accept an order for an item that's below available quantity, and clearly mark it as backordered with a real expected date — or does it just block the sale?
• Does the counter rep see the same inbound shipment data purchasing sees, or are they guessing when they tell a customer "maybe next week"?
• When a customer asks for something you don't have, does anything record that request, or does the demand disappear the moment the rep says no?
• Do your reorder points use sales history alone, or do they also account for requests you couldn't fill?
• Is there a standard substitute suggested automatically for your most commonly out-of-stock SKUs, so the rep isn't improvising under pressure?

iotoms' point-of-sale lets a counter rep see real-time stock and incoming purchase orders in the same screen where they're ringing up the sale, so a line item that's short can be accepted as a backorder with an honest date instead of turned away — and every one of those requests feeds back into when the item gets reordered next.