Français
Español
Order Fulfillment Process
Handoff one: the order lands in a system, not on a bench
Fulfillment starts the moment a paid order reaches something that can act on it. If that something is a spreadsheet a person reads twice a day, the process is fine until the day volume arrives faster than the reading. The functional requirement is a queue that assigns a pick instruction automatically and records the time it did so.
Two data points must be captured at this handoff and are routinely missing: whether the item sold is the item available, and whether the destination can be served by the lane you intend to use. An order that passes both checks is a fulfillable order; the rest are a work queue for someone.
Handoff two: locating the unit, not counting it
Picking is a search problem. A worker who knows where things live is fast until they are absent, and the knowledge leaves with them. A location system that names the bin, validates the scan and updates the count as work happens is what turns picking into a repeatable operation. Ask any provider how many SKUs one person can pick per hour — the honest answer is that the number is irrelevant to you; the miss rate is not.
Batching changes economics: picking fifty orders in one pass beats fifty walks. It also changes risk, because a mis-sorted batch produces a cluster of wrong shipments instead of one. A control that catches this is a per-order weight or scan check before the parcel closes.
Handoff three: packing is a destination decision
The pack step decides whether the parcel survives the journey and whether the recipient accepts it. Materials, void fill, carton dimensions, labelling placement and hazard documentation are all destination-specific. Three consequences of getting it wrong recur:
- Dimensional repricing — a carrier re-measures an oversized box, and a modest margin disappears in one shipment.
- Refusal at delivery — a damaged or unlabelled parcel bounces back at your cost, twice.
- Platform rejection — marketplace inbound rules reject a case that was packed to a generic standard.
Photographing the closed parcel before handover costs seconds and resolves a large share of later disputes about what actually left the building.
Handoff four: handover and the first scan
A tracking number is not a shipment. The moment that matters is the carrier's acceptance scan, because until then the parcel exists only in your system. The gap between "label created" and "carrier received" is where delivery promises quietly fail, and it is the single most useful operational number a seller can watch.
Cut-off times define this step more than any other. An order that misses the handover window is one day older than the dashboard says. Our own rule is a 16:00 China-time order cut-off with same-working-day handover and same-day tracking visibility; transit remains an estimate, because the carrier, the lane, peak season and customs queues decide that part.
Handoff five: the export and import documents
Cross-border orders add a documentary layer that domestic ones do not have: export declaration, transport documents, a destination entry, and someone accountable for duty. Entry models are not interchangeable — a low-value parcel regime that a destination withdraws or re-prices changes your landed cost without any action on your side. That happened through 2025 and 2026 as the long-standing low-value parcel exemptions were dismantled for goods shipping into the United States, while parcels entering the European Union began attracting a fixed low-value charge that several platforms pass straight through as a line item.
The practical question for a fulfillment partner is who is importer of record and what happens to your goods when an entry is queried. A partner who cannot answer in one sentence is outsourcing that answer.
Handoff six: the customer's view of the same event
Buyers do not experience six handoffs; they experience three: confirmation, a tracking number, and delivery. Everything they complain about clusters into one category — where is my order — and the fastest fix is not speed but visibility. Carrier consumer studies consistently find that most recipients open a tracking link on the day a parcel is due and begin checking well before it, so the scan history you publish matters more than the promise you made.
Proactive notification converts a support ticket into a non-event. Late-cycle deliveries are the ones that generate the most contacts, so the message that says "your parcel is now expected on Thursday" is worth more than a faster one that never arrives.
Handoff seven: when the order does not go out
The unglamorous half of the process is what happens to the orders that fail: out of stock at the pick face, address rejected by the carrier, damaged in storage, held at a border. A process that only measures successes will look excellent and disappoint customers anyway.
Define four outcomes per exception class — who is told, how fast, what the fallback is, who pays for the retry — and you have an operating standard most providers cannot produce on request. The exception taxonomy is worth more than the happy-path flowchart in a selection decision.
How to measure the whole chain
| Metric | What it actually tells you | Trap in the definition |
|---|---|---|
| Order accuracy | Wrong-item rate at delivery | Per order hides multi-line errors |
| Same-day handover | Whether the cut-off is real | Measured from label creation, not acceptance scan |
| On-time in full | Promise kept, complete | Partial shipments counted as success |
| Perfect order rate | Composite of on-time, complete, damage-free, correct docs | Each sub-partner may own one component |
| Cost per order | True unit economics | Exceptions and storage often excluded |
| Support contacts per 100 orders | Customer-visible quality | Only counts customers who complained |
Frequently asked questions
What are the steps of order fulfillment in the right order?
Receiving supply into a counted location, storing it, then per order: pick, pack, hand over to a carrier, get the first acceptance scan, deliver, and handle anything that failed along the way. Most descriptions stop at four of those.
What is the difference between order fulfillment and delivery?
Fulfillment ends at carrier handover; delivery is the carrier's performance. Mixing them is the main reason sellers argue with the wrong party.
How fast should order fulfillment be?
Same-day handover before a stated cut-off is a genuine standard; anything slower is a queue, not a promise. Delivery speed is a separate decision about which lane you pay for.
Does dropshipping change the process?
It removes your stock but not the handoffs — someone still picks, packs and scans, and the visibility problem gets worse because you no longer own the location data.
What is WISMO and why does it matter?
"Where is my order" — the dominant post-purchase support contact. It is a visibility failure rather than a speed failure, which is why tracking that updates at real events fixes more tickets than a faster carrier.
Which part of fulfillment breaks first as you grow?
Receiving and location accuracy, well before picking speed. Growth amplifies an inventory ledger that was already wrong.
Operations note from the FulfillDock by SBT desk, Shenzhen, 2 October 2026. Cut-off and handover figures are our own process standards; transit times are estimates: the carrier, the lane, seasonal peaks and border processing decide them.
