Here is the short answer. When stock runs short, OMS out-of-stock order allocation priority decides which order gets cut first — not the loudest voice in the room or the biggest order value. Score every order on four things — customer tier, delivery promise, margin contribution, and fulfilment cost — protect the high scorers, cut the low scorers, and let the system run the ranking. People only handle the exceptions. Without this rule, every stockout turns into an argument between customer service, sales, and the warehouse floor, and the loudest person wins while the most important customer loses.

I have seen this mistake three times, and the plot is the same every time. The system says stock is available. Picking finds it is not. Dozens of orders chase the same units. Customer service phones customers one by one to ask if they can wait. Sales drops voice messages in the group chat saying this account must not be upset. The warehouse supervisor stands in front of the rack not knowing which order to pick first. A stockout itself is not the disaster. The disaster is having no priority, allocating by favour and volume of voice, and finishing with no record of who got cut.

The Short Answer — Three Rules for Out-of-Stock Allocation

Rule one: set the rule before the stockout, never after it. OMS out-of-stock order allocation priority has to be written into the system while shelves are still full. The moment you change the rule after a stockout starts, you have opened a back door for favoured accounts. Change it once today, and tomorrow everyone will come asking. Once the rule is set, the answer to anyone who asks is the same: the system ran the rule. Exceptions go through approval and leave a record.

Rule two: protect the promise before the amount. An order with a same-day or next-day delivery promise outranks a large order with no time commitment. The logic is simple. Break a delivery promise once and the customer may never come back. Ship a large order two days late and the customer will very probably accept it. Big amount does not mean high priority — the most common misjudgement in shortage allocation.

Rule three: every cut leaves a trace, and every customer gets a call with a date. Each cancelled or partially shipped order needs a record stating why it was cut, how many units were cut, and when the rest will ship. When customer service calls, they must bring a replenishment date. A call without a date just pushes the customer toward a competitor.

OMS Allocation Priority in Five Steps

Step one, tier your customers. A tier is not a one-off ranking by annual spend. Combine recent order frequency, payment history, and strategic value. The result needs a signature from the sales lead, stored in the system, and called directly when a stockout hits. Nobody sets a tier on the spot.

Step two, score every order. Here is the table I use most often on projects: four dimensions, weighted, higher score ships first. The weights are illustrative, so tune them to your own business.

Scoring dimension Weight What a high score looks like
Customer tier High Strategic account, long-term partner
Delivery promise High Same-day or next-day promised
Margin contribution Medium High margin, not a clearance sale
Fulfilment cost Low Stock in the same warehouse, no cross-warehouse transfer

Step three, let the system run the allocation. The OMS allocates stock from the highest score down. When stock runs out, the remaining orders move automatically into a backorder state. The allocation must produce a list stating how much each order received, how much it is short, and when replenishment is expected. That list is the script customer service uses with customers. Without it, they can only guess.

Step four, people handle exceptions only. There are two kinds. A strategic account needs an urgent push, or a customer asks to split an order or delay it. Every exception needs approval, and the approver, the time, and the reason all stay on record. On every go-live night checklist I write, this item is always there: is the exception approval flow configured? If not, do not go live. On stockout day, someone will definitely come asking for an exception.

Step five, review the stockout list every morning — a stockout is not finished when the allocation is done. Spend ten minutes each morning on three questions: how many orders were cut yesterday, has the replenishment date moved, and has any cut customer complained? The point is to find repeat stockouts and push them to purchasing, instead of firefighting the same shortage every day.

In early 2025, a beauty e-commerce project in eastern China hit a typical stockout. A bestselling face cream arrived four days late, and more than 200 orders were waiting for the same batch. The OMS already had the allocation rules configured, so the system automatically protected next-day member orders and two long-term distributors, while the rest moved to backorder with a replenishment list. Customer service called down the list with expected dates attached. Complaints that day were lower than usual, and the sales lead told me it was the first stockout day without an angry customer call. The dates and figures here are an illustrative scenario. The rules themselves are the real practice from the project.

Three Mistakes I Have Seen More Than Three Times

The first mistake is first-come, first-served. Many warehouses sort by order creation time, and the earliest order ships first. It sounds fair, and it is the worst option. The early order may be a stockpiling one-off, while the later order may be a strategic account with a delivery promise. First-come, first-served hands priority to the clock and leaves your most important customer to luck. Here is the counterintuitive judgement: fair allocation is not first-come, first-served. Ranking by promise and value is fair. First-come, first-served is the least fair method of all.

The second mistake is manual changes with no trace. Someone phones the floor and says this customer is a friend of the boss and must ship first. The operator changes the allocation by hand, and there is no record. The next day a customer complains about a wrong shipment, and nobody can reconstruct what changed. My rule is simple: any manual change to an allocation needs a reason and an approver. Missing either one, and it does not happen.

The third mistake is allocating the stock and forgetting the replenishment. The stock is divided, the customers are called, and everyone thinks the job is done. Then replenishment arrives and nobody owns it. Cut orders sit in the system for a week with no follow-up. Allocation and replenishment have to be one line. The day replenishment arrives, the system must prompt customer service, and cut orders ship again in their original priority order. I have seen too many projects drop this loop, and the cost is a customer who thinks you forgot them.

One more detail: split shipments. When stock cannot cover a whole order, the rule must state whether to ship part of it now. My approach is that high-priority orders can split, while low-priority orders wait whole for replenishment. Splitting creates extra shipping cost, and the rule must also state who pays it. If it does not, finance will come looking for you at month end.

The replenishment order needs setting in advance too, not improvised on arrival day. My approach is that replenishment ships against the original cut list, in the original priority order, rather than letting new orders jump the queue. The system should hold a backorder list, ranked by the original allocation priority, and tick orders off as stock arrives. In late 2024, a home goods e-commerce project in southern China received replenishment and left the backorder list unmanaged for three days. Customer service kept chasing the warehouse for new orders, while more than a dozen cut customers waited a week for their goods, and two high-repeat customers were lost. The dates and figures here are an illustrative scenario. The lesson is real: a backorder list nobody watches means the allocation was wasted effort.

So back to the opening line: OMS out-of-stock order allocation priority is not a multiple-choice question. It is a rule written in advance. Before the rule, a stockout is an incident. After it, a stockout is routine. Three questions will tell you where you stand: does your OMS hold this rule today, can the system produce an allocation list automatically on stockout day, and does customer service hold a notification script with a replenishment date? Answer those three, and the next time stock runs short, nobody has to stand in front of a rack and decide by shouting.

By Lin Xiao, WMS Implementation Consultant