The order management process, and the step nobody maps

What the order management process is

The order management process is the sequence an e-commerce retailer runs from the moment a customer completes checkout to the moment they have what they paid for: capture and validate the order, allocate stock, pick, pack, ship, track, and close out through delivery or a return. Every version of it you will find drawn is a straight line that always finishes. The real process is defined by its branches, and the branch nobody draws is the one where the order breaks, so a process map without an exception branch is a picture of a good day. On one brand's launch day every step passed its own test: checkout held, payments settled, order records were created cleanly, and inside ninety minutes the storefront had taken 200 orders for stock that was already gone. Nothing raised its hand, because nothing had been asked to. A person noticed, then spent the week on the phone.

APQC's Process Classification Framework carries order management as a named cross-industry process, so there is a canonical shape to check yours against, and it runs more of the business every year: e-commerce reached 16.9% of US retail sales in the first quarter of 2026.

The seven steps, end to end

Step counts differ from one diagram to the next. The underlying chain does not. Here it is, with the way each step fails, because the failure is the part your process needs an answer for.

1. Order placed. Checkout completes and payment is authorized. Fails when the storefront accepts an order the catalog should never have offered.
2. Order captured and validated. The record is created, the address checked, fraud screened, payment settled. Fails quietly, usually after the money has moved.
3. Inventory allocated. A specific unit in a specific location is committed to the order. Fails when the committed unit exists in software and not on the shelf.
4. Picked and packed. The unit is found, checked and boxed. Fails as a short pick, a substitution, or a warehouse hold nobody writes back to the order.
5. Dispatched. The label prints, the carrier collects, a tracking number is issued. Fails when the label exists and the collection scan never happens.
6. In transit and tracked. Status flows back to the customer. Fails as silence: the parcel sits at a depot and the tracking page keeps saying the same thing.
7. Closed out. Delivered, or returned and refunded. The reverse leg is not a footnote: NRF puts 19.3% of online sales on it in 2025, a projected $849.9 billion of merchandise.

Steps four and five are where the process leaves your software and enters a building, which is why your fulfillment model sets the ceiling on everything downstream.

The step every process map leaves out

Between allocation and delivery the chain fails in a dozen ordinary ways: the allocated unit is not on the shelf, the address fails validation after settlement, the collection scan never happens and the carrier eventually posts a delivery exception nobody is watching for. Every process map on this topic treats those as things that happen, then carries on down the line. The operator's question is the one none of them answers: what does my process do next, and who owns the order while it does it?

At most brands the answer is that nothing fires until the customer emails, and then the order becomes a ticket. A few years ago I was running operations for a brand whose sales had doubled in months. Behind the storefront: an overflowing inbox, missing carrier scans, refunds piling up. None of it was a support problem. It was a process with no branch, arriving one email at a time.

Name that default precisely, because a helpdesk is a system for replying about problems, not for resolving them. Answering the customer and fixing the order are different jobs, and only one keeps the promise. The second belongs to proactive e-commerce operations, which is not a better helpdesk but a different function, and it has a unit cost you can look up: BLS puts the annual mean wage of a customer service representative at $46,590, the real price of routing an operational break to a person to hand-fix.

The clock you are already running against

In the US this process has a statutory deadline attached to it. Under the FTC's Mail Order Rule a seller must have a reasonable basis to expect it can ship within the time it advertised, or within 30 days of a properly completed order where it advertised none. Where the buyer applied for credit to pay, the seller gets 50 days. 16 CFR 435.2 goes further: a seller that cannot ship inside that window must offer the buyer a choice between consenting to a delay and canceling for a prompt refund, no later than the original deadline.

Read that as an operator rather than as a lawyer. The law already treats the shipping date as a promise, and it already says the seller has to act before the promise breaks. Most brands find out it broke when the customer tells them, days past the point at which they were supposed to have done something.

How to tell whether the process is working

Three measures are enough, and the definitions are worth taking from the SCOR Digital Standard rather than from anybody's blog.

• Perfect order rate: the share of orders delivered complete, on time, undamaged and with correct documentation. SCOR counts an order line as perfect only when product and quantity, date and time, documentation and condition are all perfect. It is a conjunction, which is why the number comes in lower than people expect.
• Order fulfillment cycle time: elapsed time from order placed to order delivered. Useful, and honest about its limits: it measures the happy path, which is most of what your on-time delivery rate measures too.
• Exception rate and time to resolution: the share of orders that leave the happy path, and how long they sit off it before anybody acts. It is the one that predicts whether the promise gets kept.

The three handoffs where orders go missing

Breaks cluster at the seams, because no single system sees the whole process. An order management system holds the record, the warehouse holds the stock, the carrier holds the parcel, and they speak in messages: the order, the despatch advice and the invoice are among the most used trade messages in retail. A despatch advice pre-announces what is actually in a shipment, so when one is late, wrong or absent, the receiving system carries on as though the order were fine.

• Storefront to order system: payment settles and the order lands, then the address or the fraud check fails behind it. Nobody notices until fulfillment tries to allocate.
• Order system to warehouse or 3PL: the unit is committed in software and absent from the shelf, or the order never reaches the warehouse at all. At one brand we found 679 orders sitting in the storefront that had never reached the ERP across a rolling 45 days.
• Warehouse to carrier: the label prints and the collection scan never happens, so the parcel is shipped in the record and stationary in reality. Nobody notices, because every system believes its own leg completed.

Each system is telling the truth about its own leg. The order is broken between them.

Designing the exception branch

A process with a branch does three things the straight line cannot. It detects the break from signals the systems already emit, the order that has not moved in days, the tracking number with no scan, the allocation against stock that is gone. It decides what should happen by measuring that order against the promise made at checkout, not against a ticket queue. Then it acts: reship, refund, re-route, cancel cleanly, or tell the customer before they ask.

The edges matter here. Keeyu does not run your order management process. We are not an OMS, not a WMS, not a carrier, not a returns portal and not a helpdesk, and we replace none of them. We sit on the branch: we watch every order against what it was promised, catch the ones that break, and act on them.

If your process map ends at delivered and has nothing drawn for the order that stalls at step three, that is not a documentation gap. It is where your tickets, your refunds and your repeat customers go. Every order is a promise. Keeyu keeps the promise: we detect the break, decide what should happen and act on it, usually before the customer knows there was anything to feel. See what the Keeyu platform does across your post-purchase stack.

• Federal Trade Commission. Mail, Internet, or Telephone Order Merchandise Rule. Business guidance on the shipping window and the seller's obligation when it cannot be met.
• Electronic Code of Federal Regulations. 16 CFR 435.2. The statutory text: the advertised window, 30 days where none is stated, 50 days where the buyer applies for credit, and consent to a delay or a prompt refund.
• ASCM. SCOR Perfect Order Fulfillment (RL.1.1). The four conditions an order line must meet to count as perfect.
• ASCM. SCOR Digital Standard 2025, front matter. SCOR as the cross-industry reference model for supply chain process and performance.
• APQC. Introduction to the Process Classification Framework. Order management as a defined cross-industry process category.
• US Census Bureau. Quarterly Retail E-Commerce Sales, Q1 2026. E-commerce at $326.7 billion, 16.9% of total US retail sales, up 9.8% year over year.
• National Retail Federation and Happy Returns. 2025 Retail Returns Landscape. An estimated 19.3% of online sales returned in 2025.
• National Retail Federation. Returns projected at $849.9 billion in 2025. The headline dollar figure behind the returns leg of the process.
• GS1. EDI trade messages. The order, despatch advice and invoice as the most used order-to-cash messages.
• GS1. Despatch Advice. What a despatch advice asserts about a shipment, and therefore what a missing one hides.
• US Bureau of Labor Statistics. Occupational Employment and Wage Statistics, May 2025. Annual mean wage of $46,590 for customer service representatives.