Wave, batch and zone picking: choosing a picking method for apparel.

Why the picking method matters

Picking is usually the largest labor line in a warehouse, and within picking, the largest share of time is not picking at all. It is walking. Industry rules of thumb put travel at half or more of a picker’s day, which means the fastest way to pick more with the same people is to walk less per unit picked. That is what a picking method is: a strategy for organizing who walks where, for which orders, and when.

The four standard methods are discrete, batch, wave and zone. They are routinely presented as a menu to choose one item from, which is misleading in two ways. First, wave picking is not the same kind of thing as the other three, it governs when work releases, not how a route is walked. Second, real floors mix methods by order profile. Getting the vocabulary straight is worth two minutes before the comparison.

Discrete picking: one order at a time

Discrete picking, also called order picking or single-order picking, is the intuitive method: a picker takes one order, walks the route it needs, picks every line, and delivers it to packing. It is the simplest method to run and to audit, one picker owns one order end to end, and it requires no sorting afterward.

Its cost is that every order pays the full travel price of its route. That is fine for a large wholesale order where the picker fills a pallet across forty lines, and ruinous for two hundred D2C orders of one tee each, where the same aisle gets walked two hundred times. Discrete picking remains the right call for large multi-line orders, full-case picks, fragile or high-value work that wants single ownership, and floors with volumes low enough that optimization is not worth its overhead.

Batch picking: many orders, one route

Batch picking (also called multi-order picking) combines a set of orders into one pick run. The picker works a consolidated list, collects the total quantity each location owes across the whole batch, and the units are sorted back to their orders, either on a multi-tote cart as they are picked, one tote per order, or downstream at a put wall.

The math is what makes it the default for D2C apparel. If forty single-line orders share a route, the travel that discrete picking would spend forty times is spent once. The trade is a sorting step, and the sorting step is where errors move: a mis-sort becomes a mis-ship unless each unit is scanned to its order. Batch size is a tuning problem, large enough to save real travel, small enough that carts stay manageable and cutoffs are still met.

Wave picking: release timing, not routing

Wave picking is the most misunderstood of the four, usually described as a variant of batch picking. It is better understood as a release strategy. Orders are grouped into waves and released to the floor together, timed around the things a warehouse actually lives by: carrier pickup cutoffs, shipping priorities, retailer routing windows, replenishment cycles and shift changes.

Within a wave, the actual picking can be discrete, batched, zoned or any mix. That is precisely the point: because a wave is a planned unit of work, it is the natural place for optimization to happen. Before anyone starts walking, the system decides which orders belong in the wave, how they batch, how routes sequence, and which zones carry which work. A well-built wave turns a pile of orders into a plan. A badly built one just batches the chaos and releases it on schedule.

Zone picking: pickers own areas

Zone picking divides the floor into areas with pickers assigned to each. An order touching three zones is picked by three people, either passed zone to zone in sequence (pick and pass) or picked in parallel and consolidated at the end. Pickers work short, familiar routes in territory they know, and aisles stop jamming with people crossing the whole building.

Zones earn their overhead as scale arrives: growing headcount, larger footprints, or a floor physically split into reserve, active pick faces and a returns area. The costs are a consolidation step for parallel zones and the slowest-zone problem for sequential ones, an order moves at the pace of its most backed-up zone. On a small floor, zones add process without adding speed.

The four methods compared

The comparison below keeps the distinction from the sections above: discrete, batch and zone answer how orders are walked; wave answers when they release. That is why wave appears alongside the others in every guide but combines with all of them in every real warehouse.

Read as a set, the table points at one conclusion: these are not four competing answers to the same question. Discrete, batch and zone are routing choices that fit different order shapes, and wave is the planning layer that decides which orders hit the floor when. The practical question is not which method to adopt but which mix each day’s order profile deserves, which is exactly where apparel gets interesting.

What makes apparel different

The order profile is bimodal. An apparel brand selling wholesale and D2C off the same floor ships two completely different shapes of work every day: a stream of one-line and two-line consumer orders, and a smaller number of wholesale orders measured in cartons, size runs and prepacks, picked against retailer routing windows and ship-by dates. The consumer stream wants batching. The wholesale stream wants discrete and full-case picks, built into waves against each retailer’s window. One fixed method serves neither well.

SKU density raises the mispick stakes. One style in eight colors and thirteen sizes is 104 SKUs of nearly identical garments in adjacent locations. Eyeball picking mispicks a medium for a large eventually and always. The answer is directed picking with scanning: the device sends the picker to a location, the barcode confirms the exact SKU at the shelf, and the error is caught at the pick face instead of in a customer’s parcel or a retailer’s chargeback.

Wholesale picking feeds compliance. Retailer orders are not done when picked. Cartons need to be packed to the retailer’s pack rules, labeled with SSCC/UCC-128 carton labels, and announced on an ASN that describes exactly what each carton contains. Scan-based packing carries the pick’s accuracy through to the label and the ASN, which is where apparel chargebacks are won or lost.

Choosing the mix, not the method

The practical selection logic for an apparel floor is short:

• Batch the D2C stream. Single-line and two-line orders ride combined routes, sorted to orders by scan on the cart or at the wall.
• Pick wholesale discretely, by the carton. Full-case and prepack picks, packed to retailer rules, labeled and ASN’d.
• Release everything in waves. Timed to carrier cutoffs and retailer routing windows, with the system building each wave from the day’s order profile.
• Add zones when scale demands them. Growing headcount, aisle congestion, or a physically split floor, not before.

The deciding factor is less the method chosen than where the choice is made. On the best-run floors, the method is a per-wave decision made by the system from the actual orders in front of it, not a fixed policy for the building. That requires a WMS that can build and optimize waves, direct each pick to a scanned location, keep batched units attributed to their orders, and hand accurate cartons to packing and the ASN. With that in place, the picking method stops being a debate and becomes an output.

Frequently asked questions