Warehouse Slotting: How Storage Location Decisions Cut Pick Travel Time
Half a picker's shift is spent walking, and no productivity target fixes that. How velocity, cube and order affinity should drive slot assignment — and how often to re-slot before demand moves on you.
Walk almost any fulfillment floor at 10 a.m. and you will see the same thing: people walking. Studies of manual picking operations have consistently put travel at roughly half of a picker's shift, and in a poorly organized building it goes higher. That time is not fixable with a faster scanner or a stricter productivity target. It is fixed by changing where things sit.
Slotting is the practice of deciding which item goes in which storage location, and revisiting that decision as demand moves. It is unglamorous, it rarely gets a line in the budget, and it is usually the cheapest meaningful throughput gain available to a warehouse that has already hired enough people.
Slotting is a labor decision wearing a storage costume
The instinct is to treat storage locations as a space problem — fit the most units into the fewest square feet. That instinct produces a warehouse that stores efficiently and picks terribly.
The better frame: every slot assignment is a standing instruction to your pickers about how far to walk, how many times to bend, and how likely they are to grab the wrong thing. A fast-moving SKU parked on a top shelf at the far end of an aisle costs you that walk and that reach on every single order, every day, until someone moves it.
The three inputs that actually drive a slot assignment
Velocity. How often is this item picked — not how many units sold, but how many picks it generates. A case-quantity item that sells 500 units in five orders is a low-velocity item. Rank your catalog by pick frequency and you will usually find the familiar shape: a small share of SKUs driving most of the picking work.
Cube and weight. Physical size caps what a location can hold and dictates replenishment frequency. A high-velocity item in a slot that only holds two days of stock creates a second labor problem — constant replenishment trips — that can erase the travel savings.
Affinity. Which items are ordered together. If two SKUs appear on the same order half the time, putting them in adjacent zones removes a cross-aisle trip on every one of those orders. Affinity is the input most operations never look at, largely because it requires pulling order-line history rather than item master data.
The golden zone, and the vertical dimension people forget
Most slotting conversations are about which aisle. The bigger win is often which shelf.
The band roughly between knee and shoulder height — the golden zone — is where a picker can reach without bending, climbing, or using equipment. It is a small fraction of your available cubic space and should be reserved almost entirely for your highest-pick-frequency items. Everything else earns its way up or down from there.
A practical rule that survives contact with reality:
| Pick frequency | Recommended placement |
|---|---|
| Top decile of picks | Golden zone, nearest the pack-out path |
| Middle band | Above or below golden zone, same general zone |
| Slow movers and seasonal | Upper levels, back aisles, or bulk reserve |
| Dead stock | Off the pick face entirely |
How often should you re-slot?
Continuously is the honest answer, but that is not operationally realistic for most teams. Two rhythms work well together:
- A quarterly full review, where you re-rank the catalog and move the items whose velocity has shifted most.
- A weekly exception pass, where you move only the handful of SKUs that have spiked or collapsed — new launches, items that just got promoted, items that just went out of season.
Ahead of a demand peak, do the full review early rather than late. Moving inventory is itself labor, and the worst time to spend it is the week volume arrives. The outbound readiness checklist for peak season covers where slotting fits in that broader sequence.
What to measure so you know it worked
Slotting projects fail quietly when nobody defines the before-and-after. Three metrics are enough:
- Picks per labor hour — the headline number, but noisy on its own.
- Average travel distance per order line — if your system logs location sequences, this is the cleanest signal that slotting specifically is what changed.
- Replenishment trips per day — the counterweight metric. If this climbs sharply, you slotted for travel and forgot about capacity.
Track them for two weeks before you move anything. Without a baseline you will end up arguing about whether it helped.
Slotting depends on data you may not have yet
None of this works if your location data is unreliable. If items are not actually where the system says they are, a slotting plan just relocates the confusion. The disciplines that hold location accuracy together — clean putaway at the door, and counting that catches drift before it compounds — come first. We covered both in detail in where inventory accuracy is actually won during receiving and putaway and in how cycle counting programs hold accuracy without a full physical count.
Get those right, then slot. Brands that would rather not run this exercise in-house can look at how our warehousing and storage operations handle location assignment as a standing process rather than an annual project.
The velocity ranking that drives all of this has to come from somewhere, and rebuilding it by hand every quarter is how slotting programs die. Our AIDWMS platform keeps that ranking current from live pick history — worth a look on the technology page if quarterly is already feeling too slow for your catalog.
