Ecommerce order fulfillment: from accepted order to parcel on the van

Ecommerce order fulfillment is everything that happens between the moment you accept an order and the moment the parcel leaves with a carrier. In the United Kingdom the word is also spelled fulfilment; this guide uses the one spelling throughout. Getting the stock figure right beforehand is a separate job, covered in stock management for webshops.

This page is for the person running a webshop warehouse of their own, from one picker up to a team of a dozen. It assumes you hold the goods and despatch them yourself. It is not written for sellers whose orders are fulfilled by a third party, and not for operations already running conveyors or robots; the sequence is the same but the decisions are not.

First: when is paper simply fine?

A printed pick list, one careful person and a box of parcel tape will run a webshop properly up to a surprising volume. If one person picks, that person knows where everything lives, and customers rarely receive the wrong thing, you do not have a fulfillment problem and nothing in this guide needs buying.

The question that separates the two situations is not volume. It is whether a second person can pick an order correctly without asking the first person anything. When the answer is no, the operation is running on one person's memory, and that is the constraint worth fixing before any device is bought.

The ecommerce order fulfillment process, in six steps

Stated plainly, so that each step can be examined on its own.

Release. The order arrives from your webshop or marketplace and becomes work for the warehouse. Whether it should be released at all depends on your stock figure being honest.

Pick. The ordered items come off the shelf.

Check. Somebody confirms that what was picked is what was ordered.

Pack. The items go into a suitable box with the paperwork and any returns documentation.

Label. The carrier label is produced for the right service and the right address.

Despatch. The parcel is handed over or collected.

The useful thing about writing the six steps down is that it stops the conversation about errors being a conversation about carelessness. Each step has its own failure, and most of them have their own fix.

What belongs on a pick list

A pick list is the instruction to the person at the shelf. It is a cheap thing to get right and expensive to get wrong.

It needs the location first, because the picker's first question is always where, not what. It needs the quantity written so that one and seven cannot be confused at arm's length. It needs enough product identity to distinguish two similar items, which usually means the SKU and a short description rather than a long marketing name. And it needs to be sequenced in the order the picker will physically walk, not in the order the customer added items to the basket.

That last point is the one most often missed, and it is the difference between a route and a wander.

Pick and pack: two halves that fail differently

Pick and pack is usually treated as a single task, because in a small operation one person does both. They are worth separating, because they go wrong for unrelated reasons.

Picking errors are errors of identity and quantity: the wrong item, or the right item in the wrong number. They come from similar-looking products stored next to each other, from a list read one line out, and from items with no fixed location. Packing errors are errors of fit and completeness: a box too large for the goods and too small for the padding, a missing component, a despatch note for a different order.

Separating them lets you see which one you actually have. An operation convinced it has a picking problem quite often has a packing problem, and the fix is a bench layout rather than a scanner.

What an error costs

The reason to take a picking error seriously is arithmetic rather than principle. A wrong item despatched costs the outbound carriage you have already paid, the return carriage, the labour to receive and inspect the return, the labour to pick and pack the replacement, the second outbound carriage, and a customer who now tells other people about it. None of those are large on their own.

As an illustration, consider an operation despatching 200 orders a day with an error rate of two in a hundred. That is four wrong parcels a day, around a thousand a year. Whatever you believe one of those costs you end to end, multiply by a thousand before deciding that a scanner is an extravagance. The cost per order line calculator will do that sum with your own figures, which is the only version of it worth acting on.

The ecommerce warehouse: what to change first

Most advice about an ecommerce warehouse starts with the racking, and that is the wrong end. Before anything is bought or moved, two things are worth knowing: how far your pickers actually walk for an average order, and whether the items that sell most often are stored closest to the packing bench.

Those two questions, and what to do about the answers, are the subject of our warehouse picking guide, which covers layout, where to place which products and the choice between picking methods in proper depth. This page deliberately stops short of that ground, because the answer depends on the shape of your order profile rather than on anything specific to selling online.

What is specific to selling online is the mix: webshop orders are typically many orders of few lines, which is the profile that rewards picking several orders in one walk. That approach is called batch picking, and alongside zone picking and wave picking it is the first thing to look at once a single picker is no longer keeping up. Which of them suits your building is a question the picking guide answers properly.

Carriers and cut-off times

Your despatch deadline is not set by you. It is set by whichever carrier collects, and your own cut-off has to sit far enough in front of it to survive a bad afternoon.

Work the sum backwards. Take the collection time for Royal Mail, Evri, DPD or whoever you use. Subtract how long picking, packing and labelling actually take when you are busy, not when you are calm. Subtract a margin for the order that needs a conversation. What remains is the latest time you can safely promise same-day despatch, and it is almost always earlier than the time a webshop advertises.

Publishing an honest cut-off is worth more than an optimistic one. A customer who orders at four and is told tomorrow is content. A customer promised today and given tomorrow is a complaint, and on a marketplace a late despatch is also a metric.

A shipping platform such as Sendcloud will handle label production and carrier selection; what it cannot do is shorten the part of the process that happens at your bench.

When a scanner pays for itself

A barcode scan at the point of pick removes the most common error in the whole operation: a person reading one line of a list and taking the item from the line below. It replaces reading with confirmation.

The threshold is not a volume. It is the moment your error cost over a year exceeds the device plus the work of getting barcodes onto your stock. In practice that moment tends to arrive when a second picker joins, because two people cannot share one person's memory of where things are, and when a line starts to be picked frequently enough that a small error rate becomes a weekly occurrence rather than a monthly one. If your supplier does not supply barcodes, our barcode generator will produce them.

Measure before and after with the same number: picks per hour and errors per thousand lines. The picks per hour calculator gives you the first of those from a single timed observation.

Frequently asked questions

What is ecommerce order fulfillment?

Everything that happens between accepting an order and handing the parcel to a carrier: releasing the order, picking the items, checking them, packing them, producing the label and despatching. It does not include getting the stock figure right beforehand, which is a separate job.

What does pick and pack mean?

Pick is taking the ordered items off the shelf. Pack is putting them in the right box with the right paperwork and label. They are usually treated as one task because in a small webshop operation the same person does both, but they fail for different reasons and are worth measuring separately.

When do I need a barcode scanner?

When the cost of your picking errors over a year exceeds the cost of the device and the work of labelling your stock. A scanner removes the error of reading one line and picking another, which is the most common error there is. Below roughly a hundred order lines a day with one picker, a printed list and a careful check usually hold.

How late can I accept orders for same-day despatch?

Work backwards from your carrier's collection time, subtract how long your own picking and packing actually takes at your busiest, then subtract a margin. Publish that time and keep it. A cut-off you miss costs more goodwill than a cut-off set an hour earlier.

Continue with

• Stock management for webshops — getting the figure right before the order is released
• Returns — the reverse of this process
• Peak season — the same six steps under three times the volume
• E-commerce — the wider picture for online retail

Two questions about your own operation

Could a new person pick an order correctly tomorrow morning without asking anyone where anything is?

What is your error rate per thousand order lines, and when did you last count it rather than estimate it?