Warehouse picking methods, and how to choose one

Why the picking method matters

In most warehouses, walking between locations takes far more of a picker's time than taking items off the shelf. A picking method is a way of arranging work so that each walk collects as much as possible, without creating so much sorting and checking afterwards that the savings disappear.

No method is best in general. The right one depends on how many orders you ship, how many lines each order has, how much the orders overlap in SKUs, how big the building is and when the carriers collect. Most operations end up combining two methods, for example batch picking for small parcel orders and discrete picking for large B2B orders.

The method also shapes everything around picking: how orders are released, how the pack station is laid out, how many scanners and carts you need, and how you measure the team. Changing it later is possible, but it is a process change, not a setting, so it is worth choosing deliberately.

Discrete picking

In discrete picking, one picker picks one order from start to finish, then starts the next. The pick list holds that order's lines, sorted in walk order. There is no sorting step, and the picker can take the order straight to packing.

Discrete picking fits large orders with many lines, such as a distributor's 40-line wholesale order or a pallet build for a retailer. It also fits low volumes, where there are not enough orders to batch, and orders that need special handling. Its weakness is travel: for small orders, the picker walks the whole route to collect one or two items.

Discrete picking is also the easiest method to train and audit. Every item in the tote belongs to one order, so a mistake is caught at packing against a single order, and a new picker can be productive on the first day. That makes it a sensible starting point for any new operation, even one that will move to batches later.

Batch picking

Batch picking groups several orders into one walk. The picker collects the total quantity of each SKU for the whole batch, then the items are sorted into individual orders, usually at the pack station. This is called pick-then-sort.

For example, a batch of 30 parcel orders might include 12 orders that each contain one phone case CASE-114. Discrete picking would visit that bin 12 times; a batch visits it once and picks 12. The saving grows with the overlap between orders.

Batch picking fits high volumes of small orders with shared SKUs, which is typical of e-commerce. The cost is the sort step. It needs a clear sorting convention at the pack station and scan-verified packing, or the time saved on the floor turns into mis-shipped orders.

Keep batches to a size the pack station can sort without confusion. A batch of 10 to 30 small orders is easy to sort into labeled slots or packages; a batch of 200 needs a put wall and a clear process. Build batches from orders with the same cut-off, so an urgent order is never held back by a slow one.

Cluster picking

Cluster picking is batch picking that sorts at the shelf. The picker pushes a cart with several totes, one per order, and puts each item straight into its order's tote. There is no separate sort step, but every pick needs a second confirmation: which tote.

Cluster picking fits small orders when a sort station would be a bottleneck, and when orders overlap less, so a combined total would not save many visits. It needs carts, totes and discipline about tote positions, because a unit dropped into the wrong tote ships to the wrong customer. Scanning the tote as well as the item closes most of that gap, at the cost of one extra scan per pick.

Zone picking and pick-and-pass

In zone picking, each picker owns an area of the warehouse and picks only from it. Pickers learn their zone well and never walk the whole building. Orders that span zones have to be brought together, and there are two ways to do it.

In pick-and-pass, the order's tote travels from zone to zone, often on a conveyor or a cart relay, and each picker adds their lines before passing it on. In pick-and-consolidate, zones pick in parallel and the partial orders are merged at a consolidation area before packing. Pick-and-pass is simpler to control; pick-and-consolidate is faster when orders touch many zones.

Zone picking fits large buildings, many pickers, and areas with different rules, such as a chilled room, a secure cage for high-value items or a mezzanine. Small warehouses rarely need it. The overhead is real: someone has to balance work between zones so one picker is not idle while another is buried, and orders that touch five zones wait on the slowest one.

Wave picking

Wave picking is about timing rather than routing. Orders are released to the floor in waves, each planned around an event: a carrier pickup, a truck departure or a shift break. Within each wave, the work can be picked discretely, in batches or by zone.

Waves fit operations with fixed carrier cut-offs and enough volume to plan labor in blocks. For example, a wave released at 9:00 might hold all orders for the 12:00 courier collection, and a 13:00 wave the orders for the evening line haul. The risk is idle time between waves and urgent orders stuck waiting for the next one. Many smaller operations get most of the benefit by simply sorting released work by cut-off time and urgency.

If you do run waves, size each one to the labor you actually have in that window, and keep a way to release a single urgent order outside the wave. A wave plan that cannot absorb a late same-day order will be bypassed within a week.

Walk sequence and pick routing

Whatever method you choose, the order of stops on a pick list decides how far people walk. A walk sequence gives every location a number in the order a picker should visit it, and pick lists are sorted by that number.

Number your locations along the route people actually take. In a warehouse with narrow aisles, that is usually an S-shaped path: up one aisle, down the next. For example, A-01-01, A-01-02 and onwards up aisle A, then back down aisle B from its far end. Reflect one-way aisles, doors and anything else that changes the real path, and renumber when the layout changes.

Leave gaps in the numbering, such as steps of 10, so a new bin can be inserted without renumbering the whole aisle. Check the sequence by walking a real pick list: if the picker has to double back, the numbers are wrong, not the picker.

Routing and slotting work together. Put the fastest movers on accessible pick faces near the pack area and slow movers deeper in. The warehouse slotting guide covers how to decide what goes where.

How to choose a picking method

Pull a month of order history and answer the questions below. The answers usually point to a primary method and a secondary one for the exceptions.

For example, a brand shipping 400 parcel orders a day, averaging 1.3 lines per order across 600 SKUs, should batch pick. A distributor shipping 60 orders a day at 35 lines each should pick discretely in walk order, and consider zones only when the building or the team gets large. To size the team for either, try the picker staffing calculator.

• Lines per order: one to three lines favors batch or cluster; dozens of lines favors discrete or zone.
• Orders per day: low volume favors discrete; high volume favors batch, cluster or zone.
• SKU overlap between orders: high overlap favors batch pick-then-sort; low overlap favors cluster.
• Building size and zones: a large site or distinct temperature and security areas favors zone picking.
• Carrier cut-offs: fixed pickups favor waves or at least cut-off ordering.
• Traceability: lot, expiry or serial capture at pick favors methods where each pick is scanned against a single order or a verified sort.
• Team size: two pickers cannot run zones; twenty pickers probably should.

Measuring picking performance and accuracy

Whichever method you choose, measure it before and after a change so you know whether it worked. Four numbers cover most of what matters.

Measure accuracy from what customers and packers find, not from what pickers confirm. Scan-verified packing catches a wrong item before it ships, and each catch tells you whether the problem is the picking method, the slotting or the labels. If errors cluster on look-alike SKUs placed side by side, separate them rather than retraining the team.

• Lines picked per hour: the core productivity measure. Compare like with like, because a line of one small item and a line of a full case are different work.
• Pick accuracy: the share of lines picked with the right item and quantity, measured at pack verification or from customer complaints.
• Order cycle time: time from order release to packed or shipped. Batching improves travel but can lengthen cycle time for individual orders.
• On-time shipment against cut-off: the measure customers actually feel.

Picking methods in NextStock

NextStock supports discrete picking and batch pick-then-sort today. In a batch, one walk collects the totals for several orders into one container, and items are sorted to their orders at the pack station with scan-verified packing. Pick lists follow each location's walk sequence, allocation prefers pick faces over storage, and released work can be ordered by cut-off time and urgency. See order picking for the full flow.

Wave planning and zone pick-and-consolidate are on the roadmap and not available yet.