Bottom line first: when stock runs short, order allocation comes down to one sentence — rush paid orders first, key accounts second, regular orders third, deferrable orders last. Write those four tiers into your OMS allocation rules. Let the system sort it out. Never let a human do it by gut, and never use first-come-first-served.
I've seen this exact pit three times. Every time it was the same setup: a peak-season order spike plus a delayed inbound shipment, not enough stock to go around, someone asking in the group chat "which orders do we cut," the boss making a gut call, and everyone regretting it. Here's the full reasoning and the fix, once and for all.
Why First-Come-First-Served Is a Fairness Trap
First-come-first-served looks the fairest. In practice it's the most expensive way to handle order allocation in a stockout. It hands scarce inventory to whoever ordered earliest — not to whoever matters most.
Here's the counterintuitive call: the best stockout allocation strategy doesn't protect big customers or early buyers. It protects the orders whose cancellation costs you the most. Buried in those first-come orders might be a rush order with a promised delivery time — cancel it and you pay penalties. Or a key distributor's replenishment order — cancel it and you lose a quarter of partnership. Fairness delivered, and the boss pays the bill.
And then there's the blanket "pause all order intake" move. Some sellers shut down intake when stock gets tight, thinking "at least our fulfillment rate stays clean." Fulfillment rate is clean, sure — but marketplace category rankings run on sales and conversion. Pause for three days, your ranking slips, and climbing back takes two months. That's trading short-term dignity for long-term bleeding.
So the core principle of allocation rules is one line: rank orders by the cost of cutting them, and let the system execute automatically. Leave no room for gut calls.
Four Priority Tiers: The Standard Stockout Allocation Framework
| Tier | Order type | Allocation policy |
|---|---|---|
| P0 | Paid rush orders, promised-delivery orders | Fully protected, even if it means splitting wholesale orders |
| P1 | Key accounts, distributor replenishment | High fulfillment rate guaranteed, priority reallocation |
| P2 | Regular retail orders | Proportional allocation, auto-split or delay on shortfall |
| P3 | Pre-orders, low-margin, deferrable orders | Allocated last, proactively communicate delays |
P0 and P1 are red lines — untouchable no matter how tight stock gets. P2 is the cushion, absorbed through split shipments and delays. P3 is the pressure valve: most buyers accept a few days' delay if you tell them upfront.
Three real disasters, all from skipping this framework:
Client one: November 2024 peak season, a 3C accessories seller. Order spike plus inbound delayed by 5 days, zero priority rules in the OMS, default first-come-first-served allocation. Among the cancelled orders was a key distributor's bulk order — roughly 20% of that month's volume. The distributor went to a competitor for replenishment. The seller didn't just lose an order; they lost a quarter of channel partnership. Afterward the boss said "we should have protected the key account" — but the system had nowhere to write "protect key accounts."
Client two: June 2025 peak, a home goods seller. Stock ran critical, operations paused all order intake for three days. Fulfillment rate stayed clean. Category ranking slid from the 40s to outside the top 50. When rankings drop, traffic follows — it took two full months to climb back. The customer-service cost saved in those three days didn't come close to covering the traffic lost from the ranking drop.
Client three: an apparel seller, the night before a big promotion. I've seen this pit three times: channel-reserved inventory with no hard isolation. 2,000 units reserved for a key account got eaten alive by the retail order spike. When the key account came to collect, the warehouse had nothing. Contract breach, channel penalties, emergency air freight — a six-figure loss all told. Since then, rule number one in every allocation setup I configure is hard isolation of reserved inventory.
OMS Allocation Rule Checklist: Copy This and Use It
Take this checklist to your OMS implementer or IT team. Whatever's missing, add it:
- Priority field on every order: four tiers P0–P3, derived from customer level × order type × promised delivery time, auto-tagged at order creation. No manual edits.
- Hard isolation of reserved inventory: reserved stock is visible to retail orders but untouchable by them — physical separation or logical lock, pick one.
- Oversell circuit breaker: auto-pause sales or cap quantities when available stock drops below a safety threshold. Don't set the threshold to zero — zero means no breaker at all. I've seen three companies do exactly that, because "it might hurt sales."
- Hardcoded allocation sequence: reserved → P0 → P1 → P2 → P3. The OMS runs allocation in this order, and nobody gets to jump the queue.
- Auto-split on shortfall + notification: P2/P3 orders short on stock split automatically — ship what's available, generate a follow-up for the rest, and notify the buyer about the delay. Don't make customer service explain it.
- Re-run allocation twice daily: morning and evening. Inbound receipts and cancelled orders free up stock — skip the re-run and you're leaving recovered inventory sitting idle.
Common Crash Points: I've Seen This Pit Three Times
Crash point one: the priority field exists but nobody maintains it. The customer tier table hasn't been updated in three years, a dead distributor is still flagged as a key account, and every P1 tag is wrong. Fields are static; maintenance is a living job. Set a quarterly review rule.
Crash point two: the circuit breaker threshold set to zero. The boss fears hurting sales, so the breaker is decorative. When the spike hits, overselling is instant — cancelled orders tank your ratings, emergency air-freight replenishment bleeds shipping cost. Both ends cost money; better to trip the breaker early.
Crash point three: reserved inventory eaten by overselling. One more time — I've seen this pit three times. Reservation must be hard isolation. Soft isolation melts like paper in front of peak-season traffic.
One honest question to close: in your OMS, when stock runs short, who decides which orders get cut by gut feel? If the answer is "we discuss it in the ops group chat," this checklist was written for you.
By Lin Xiao, WMS Implementation Consultant
