Warehouse Wave Picking Release Timing: When to Release the Next Wave

Here is the answer up front: wave picking release timing is not about the clock. A wave should go out only when three lines meet at the same time — the order pool is big enough for one clean wave, the floor has free people and equipment to absorb it, and the carrier cutoff still leaves a real buffer. If any one of those three is missing, releasing the wave creates the mess you will spend the rest of the shift cleaning up. I have seen this pitfall three times, in three different warehouses, and each time the team blamed picker speed when the release decision was the actual problem.

The hard part is that most warehouses do not measure those three lines before they press release. They release on a schedule, or because the floor looks quiet, or because a big batch just landed and everyone is nervous about the cutoff. Feeling is not a release criterion. Numbers are.

Wave Release Timing Starts With the Order Pool

Start with the order pool, because a wave is only as good as the orders inside it. Before release, I want a snapshot: how many orders, how many lines, how many units, and how those lines spread across zones. That snapshot tells me whether the wave is worth running at all, and whether it fits the floor.

Every warehouse needs a minimum wave size and a maximum wave size, and both numbers should come from downstream capacity, not from habit. If your checking stations can clear 120 orders an hour, do not release 200 orders in one wave. You have not made picking faster; you have moved the bottleneck from the aisles to the checking benches, where it is more visible and more expensive. If the pool is below the minimum, keep accumulating. A wave of 30 scattered orders sends pickers walking the whole building for very little shipped volume.

I learned the cost of ignoring this in Dongguan in November 2023, in a small-appliance warehouse coming off the Singles' Day peak. Orders were still high, the supervisor was worried about the carrier cutoff, and at 9 a.m. he released 340 orders in a single wave. Fourteen pickers were on the floor, but only three checking stations were staffed. Picking was mostly done by 11 a.m. Checking had six pallets of unchecked goods stacked beside it. The carrier cutoff was 5 p.m., and the last load did not leave until 5:40 p.m. The customer complaint was not about slow picking. It was about late dispatch. After that week we capped a single wave at 150 orders and staffed five checking stations. The same order volume cleared before 5 p.m. I have seen this pitfall three times, and every time the symptom looked different — late trucks, overtime, checking queues — while the cause was the same: the released wave was bigger than what the next step could handle.

Zone spread matters just as much as size. A wave with lines in three distant zones forces every picker into long travel. When the snapshot shows heavy cross-zone spread, split the wave by zone. Two smaller, zone-clean waves will beat one large, zone-messy wave on almost every shift I have measured.

How to Do the Capacity Math Before a Wave Release

Once the pool is ready, check whether the floor can take it. "Can take it" has a concrete meaning in my go-lives: pickers actually on shift, pick carts and totes actually available, checking and packing stations actually free, and dock doors not already queued with waiting vehicles. Any one of those in the red means the wave waits, or shrinks.

The math I use on site is deliberately plain. Divide total wave lines by lines picked per picker per hour, and you get picking time. Divide total wave orders by orders checked per station per hour, and you get checking time. Take the larger of the two, add 30 minutes for packing and loading, and you have the shortest realistic finish time. If that finish time lands less than 45 minutes before the carrier cutoff, do not release this wave as it stands. Split it, or wait for the next shift's staff to arrive.

The value of this table is not sophistication. Every number can be read from the WMS, and a supervisor can defend the decision from it. If a number cannot be found, that is itself a finding: fix the data first, then argue about release rhythm. For any policy claim, refer to the official website for the latest policy rather than relying on memory.

A second case, from a beauty warehouse in Suzhou in May 2024, shows the opposite mistake. The morning pool was small, about 80 orders by 10 a.m., and the supervisor liked to wait until 2 p.m. to bundle 180 orders into one big wave. On paper that looked efficient. In practice, morning pickers spent nearly half their shift waiting for work, and the afternoon wave then overwhelmed checking. We switched to a small wave at 10:30 a.m. for those 80 orders, followed by a second wave at 2 p.m. Daily lines picked per person rose by roughly 20%. Here is the counterintuitive judgement: smaller, more frequent waves are often more stable than big, infrequent ones. What you save is waiting time, not picking motion. A warehouse that waits for the perfect big wave pays for that perfection in idle mornings and queued afternoons.

The Wave Release Checklist: Tick Every Box Before You Press Release

Release itself must leave no room for interpretation. I make the floor run this checklist, and a missing tick means no release.

• Open the WMS wave pre-check screen and confirm orders, lines, and zone spread
• Confirm pickers on shift and checking stations staffed, then calculate the estimated finish time
• Confirm the carrier cutoff and verify at least 45 minutes of buffer remains
• Confirm carts, totes, and pallets on hand, and reallocate before release if any item is short
• Remove exception orders from the wave: stock-outs awaiting confirmation, address problems, and orders needing special checks stay out
• Check progress 15 minutes after release: first-order start time, aisle congestion points, and checking queue length
• After the wave, review three numbers: estimated versus actual finish time, longest checking queue, and orders still unshipped at cutoff

The step teams skip most often is removing exception orders. One bad order in a wave stops a picker at the shelf, triggers a phone call to a supervisor, and slows the whole route behind that picker. Let the exception wait one wave rather than letting it drag a full wave down. The same discipline applies to short picks: if the system cannot confirm stock, do not let that order hold up twenty good orders.

Common Wave Release Failures

The first failure is watching only order volume and ignoring checking capacity. Fast picking with slow checking still ships late; the goods just wait in a different place. The second failure is mixing too many zones in one wave, which turns picking routes into long walks and quietly cuts lines per hour. The third failure is forcing a wave out just before cutoff when the math already says it will finish late. Overtime can recover 15 minutes. It cannot recover a 60-minute gap. The fourth failure is releasing and then walking away. If the first order starts 20 minutes late and nobody notices, the whole wave is half an hour late before anyone reacts.

Wave release is, at its core, a reservation of resources. Orders, people, equipment, and cutoff time have to line up inside the same window. A supervisor's job is not to push waves out the door. It is to confirm that alignment, then release. Get that judgement right and the rhythm takes care of itself: waves go out when the floor is ready for them, not when the clock says it is time. That is the difference between a floor that looks busy and a floor that actually ships on time.

By Lin Xiao, WMS Implementation Consultant