The ecommerce warehouse process runs in seven stages: order capture from your webstores and marketplaces, receiving, putaway and storage, picking, packing, dispatch and returns. Each stage hands work to the next, so a delay at goods-in shows up hours later as a missed courier collection. Online sales made up 28.8% of all retail sales in Great Britain in September 2026, according to the ONS, which is why more warehouses now run a flow built for thousands of small, single-customer orders rather than a few large ones.
This guide walks through the whole flow at overview level. Each stage links to a deeper guide where we cover it properly, so you can use this page as the map and the linked pages as the detail.
The ecommerce warehouse process at a glance
Most guides agree on the core. Outvio’s breakdown of ecommerce warehouse operations lists receiving and inspection, storage and slotting, picking and packing, carrier handoff and returns. FreightAmigo names six fundamental warehouse processes: receiving, putaway, picking, packing, shipping and returns. Both lists start at the dock door. We add one stage in front of them, order capture, because in ecommerce the order usually exists before anyone in the warehouse touches the stock, and most problems on the floor trace back to how orders arrive.
| Stage | What happens | Where it usually breaks | What a WMS records |
|---|---|---|---|
| 1. Order capture | Orders flow in from webstores and marketplaces | Re-keying, oversells, orders arriving after cut-off | Order, channel, carrier service, expected dispatch date |
| 2. Receiving | Stock is checked in against what was expected | Unannounced deliveries, shorts and damages | Quantity, batch, dates, discrepancies |
| 3. Putaway and storage | Stock moves to a bulk or pick location | Stock left on the floor, wrong location | Location, rotation date, hold status |
| 4. Picking | Items are collected for each order | Wrong item, long walks, empty pick faces | Pick task per order line, operator, time |
| 5. Packing | Items are checked, packed and labelled | Wrong box, missing item, damage in transit | Pack confirmation, parcel count, label |
| 6. Dispatch | Parcels are sorted and handed to carriers | Missed cut-offs, parcels on the wrong trailer | Carrier, service, dispatch time |
| 7. Returns | Goods come back, are inspected and restocked or written off | Returns piling up uninspected | Return receipt, condition, hold reason |
Read the table from left to right as a chain. The third column is the useful one: each failure in it creates work for the stage below. An oversell at order capture becomes a short at picking. A pallet booked into the wrong location at putaway becomes a picker standing at an empty shelf.
Stage 1: order capture from marketplaces and webstores
An ecommerce warehouse rarely has one order source. A typical seller runs a webstore on Tienda virtual alongside Amazonas, eBay and, increasingly, TikTok Shop. Each channel has its own order format, its own delivery promise and its own penalty for getting it wrong.
The first job is getting every order into one queue without anyone typing it. Copying orders from a marketplace screen into a spreadsheet works on a quiet day and falls apart the first time volume doubles. The second job is keeping stock levels honest across every channel, so you don’t sell the last unit twice. That’s the core of multichannel inventory management, and our guide to software de gestión de existencias para el comercio electrónico covers how stock syncs back to each channel.
Order capture also sets the shape of the warehouse day. Every order should carry its carrier service and an expected dispatch date when it lands, because those two fields decide what gets picked first. In Clarus, the default pick list is ordered by expected date of dispatch, and orders arrive through native connectors handled in gestión de pedidos.
Stage 2: receiving stock into the ecommerce warehouse
Receiving is where stock becomes real in the system, and it’s the first physical step in the ecommerce warehouse process. If the count at goods-in is wrong, every later stage inherits the error, and the first person to find out is usually a customer.
The biggest choice here is whether you know what’s coming. A pre-advised receipt, where the supplier sends an advance shipping notice (ASN), lets the team check deliveries against an expected list. A blind receipt means counting what turns up with nothing to compare it against. Mathew Buttar, Head of Solutions at Clarus, argues for pre-advice wherever you can get it: you should know “what’s coming, how much of it, what it is, when it’s arriving, so you know how many people to spot on it.”
Barcodes matter as much as paperwork. Pallet and carton labels that follow GS1 standards let a scanner read the product code, batch and quantity from one label, so receiving becomes a scan rather than a typing exercise. Clarus handles this in its goods-in workflows, and a short receipt won’t complete against a pre-advice until the shortage is dealt with. Our warehouse goods-in procedure guide covers inspection, discrepancies and damage step by step.
Stage 3: putaway and storage
Once stock is received, the ecommerce warehouse process needs it somewhere a picker can find it. Ecommerce storage usually splits into two zones: bulk reserve, where full pallets and cartons sit, and pick faces, where single units are picked. The link between them is replenishment, which tops up pick faces from reserve before they run empty.
Good putaway is directed, not left to whoever is driving the forklift. Suggested putaway tells the operator which location to use and confirms it with a scan. Stock with a date on it also needs a rotation rule. Clarus records first in, first out as the creation date of the stock record, down to the millisecond, and supports FIFO and FEFO rotation for products with best-before dates.
Storage layout is a bigger subject than this overview can cover. Our storage process in a warehouse guide goes through location types, slotting and replenishment in detail.
Stage 4: picking in the ecommerce warehouse process
Picking is where most warehouse labour goes, and ecommerce makes it harder. Orders are small, often a single item, and there are lots of them. A picker might visit dozens of locations to fill dozens of orders that a wholesale operation would have filled from a handful of pallets.
The fix is to cut walking and wrong picks at the same time. In Clarus, pick routes follow a walk sequence set against each location, so you can design the route to finish at goods-out. Every pick is a task for one order line, and if the operator scans the wrong location the handheld stops them and shows where they should be standing. That’s the mechanism behind fewer errors, though Mathew is honest about its limit: a scanner can’t stop someone who scans the right product and then picks up the box next to it. “There is not a warehouse in the world that does not have errors,” as he puts it. The goal is to catch them at the shelf, not at the customer’s door.
For methods such as batch, zone and wave picking, read our guía de selección de pedidos de almacén. For how picking and packing work together as one flow, see our pick and pack guide.
Stage 5: packing
Packing is the last chance in the ecommerce warehouse process to catch a mistake before it costs you a return. The packer confirms the items against the order, chooses the right packaging, adds any inserts and prints the carrier label.
Three things go wrong here more than anywhere else. The wrong box size wastes postage and filling. Missing items slip through when packers trust the pick rather than checking it. And fragile goods get damaged in transit because the packaging was never designed for a parcel network. Packing benches are also where repetitive lifting and twisting happen all shift, so it’s worth reading the HSE guidance on manual handling at work when you lay one out.
A scan at the empacar escritorio confirms each item against the order line before the label prints. That second check is cheaper than the alternative most paper-based sites use, which is a supervisor walking round every finished order to recheck it.
Stage 6: dispatch and carrier cut-offs
Dispatch is where the ecommerce warehouse process meets the courier, and the courier doesn’t wait. Each carrier has a collection time, and anything not packed, labelled and sorted by then misses the van.
That makes pick order a dispatch decision. Mathew describes the common case: if half your orders go on one carrier collecting at 2pm and half on another collecting at 4pm, you want to pick the 2pm carrier’s orders first. Clarus can pick by carrier and service for exactly this reason, so the team works through, say, every Royal Mail order before moving on to Correos del país o Evri. An automation can also send a file to the next party when a pick list completes, so nobody has to email a dispatch notice by hand.
Nuestro warehouse dispatch process guide covers sortation, manifesting and proof of dispatch. If you also ship B2B or retail orders on pallets, mixed palletising explains how to build stable pallets from many SKUs. And our page on software de gestión de envíos looks at label printing and carrier selection.
Stage 7: returns
Returns are the stage of the ecommerce warehouse process that most sites underplan. In the UK, customers who buy online can usually cancel within 14 days of receiving an item, then have another 14 days to send it back, and GOV.UK guidance says sellers must refund within 14 days of getting the item back. So a pile of uninspected returns isn’t just untidy. It holds up refunds you’re legally obliged to make.
A workable returns process has four steps: book the parcel in, inspect it, decide whether it goes back to stock, to refurbishment or to write-off, and update the channel stock level. Give returns their own space at goods-in so they don’t mix with new stock.
To be straight about our own product: Clarus doesn’t have a dedicated returns module today. Returns are handled as a manual goods-in process, and stock that can’t go straight back on sale can be put on hold with a reason code, which records who held it and when. If returns are a large share of your volume, ask any vendor to show you their returns flow live rather than describing it.
Where the ecommerce warehouse process breaks down
The stages fail together, not one at a time. Mathew, who managed warehouses for the best part of 25 years before joining Clarus, describes the cascade: a busy day at goods-in means picks run late, so the next morning you have to pick first to make room, while today’s deliveries are already queuing outside.
Ecommerce tightens the loop because the customer’s clock is short. In Mathew’s words, the challenge is “the amount of people you need to do it in the time that you need to do it, i.e. within 24 hours, because that’s what’s now expected.” A wholesale order can slip a day. A parcel promised for tomorrow can’t.
Three patterns cause most of the trouble:
- Stock records drift from the shelf. Pickers find empty locations, orders go short and customer service apologises. Regular cycle counts keep records honest. JODA Freight reached 99% stock accuracy by its second year on Clarus, and stock takes that took weeks in Excel now take a day.
- Peaks arrive faster than headcount. A system that guides new starters on a handheld shortens the time before they’re useful. KATEM Logística scaled monthly picking volumes 10x after switching.
- Channels and the warehouse disagree. When stock levels sync late, you oversell. The fix sits back at stage one, with integrations rather than exports.
Uniserve’s guide to using warehousing in an ecommerce business makes a similar point from the logistics side: warehousing only works when it’s joined up with shipping and returns, not run as a separate function.
Choosing a system to run the ecommerce warehouse process
You can run this flow on spreadsheets and paper at low volume. The trouble starts when one person holds the logic in their head, or when a busy week means the paper pick lists and the stock records stop matching. A warehouse management system (WMS) turns each stage into scanned tasks with a full audit trail, so you can see where any order is and who touched it.
When you compare systems, test them against the seven stages above. Ask to see your own channels connected, a receipt with a short delivery, a pick routed by carrier cut-off and a return booked back to stock. Our best WMS for ecommerce guide compares the main options. Clarus is a cloud-native WMS with no servers to maintain, and a standard implementation with the right support on both sides aims to go live in around 12 weeks.
Data is also becoming more useful outside the WMS itself. The Clarus Asistente de IA can find and pull operational data on request, such as every transaction for one customer on a given day, and an MCP server lets AI tools query live warehouse data directly.
Habla con un experto en almacenes
If you’re evaluating your options and want to see how a purpose-built WMS handles the ecommerce warehouse process in practice, Clarus is worth a conversation. We work with ecommerce operators, 3PLs and distributors across the UK to implement warehouse management software that fits the way you operate, not the other way around.
Ponte en contacto con nuestro equipo para discutir tus requisitos.