Your pick rates looked fine six months ago. Same headcount, same WMS, same shifts. Now they're down 12 to 18 percent and nobody can tell you why. The supervisors have already run through the usual suspects: coaching gaps, a few newer associates, one zone that always runs behind. The labor looks fine and the WMS hasn't changed. So it must be the people, right?
Probably not. In a lot of warehouses, the pick rate problem isn't a labor problem. It's a slotting problem that's been quietly compounding since the last time anyone looked at velocity classifications. The WMS is still routing putaway and directing picks exactly as it was told. What it was told is now wrong.
What Slotting Decay Actually Looks Like
Slotting decay doesn't announce itself. There's no alert in the WMS that says "velocity tiers are 14 months stale." What you get instead is a slow erosion of pick path efficiency that shows up in metrics most operations don't track at the slot level.
A SKU that was a B-mover when the facility opened might be an A-mover now. Demand shifted. The customer mix changed. A product line that moved in pallet quantities now ships in eaches. But the WMS still has it classified as medium-velocity, parked in a mid-aisle location far from packout. Every time a picker walks to that slot, they're covering ground the system never intended to cover for a SKU of that pick frequency.
The math compounds fast. If 15 percent of your A-velocity SKUs are sitting in B or C locations, your pickers are walking an extra 40 to 80 feet per pick for those items. Across a shift with hundreds of picks, that's hours of additional travel time that no coaching program can fix. Those A items should be in the golden zone, waist-to-shoulder height, nearest to packout. Instead they're in the back aisle because nobody told the WMS they became important.
Most warehouses treat slotting like infrastructure. Build it once, maintain the building. That's the mistake. Slotting behaves more like inventory. If you don't manage it actively, it degrades.
How Velocity Classification Goes Stale Inside the WMS
The WMS makes putaway decisions by cross-referencing product attributes against available location attributes: velocity tier, unit dimensions, weight, hazmat flags. The logic is solid. The problem is that velocity tier is a snapshot, not a live calculation. Most systems won't dynamically reclassify SKUs based on recent pick history unless someone configured that recalculation to run. If your WMS was set up three years ago and no one has touched the velocity rules since, the system is routing product based on demand patterns from 2023.
A real example. A 3PL operation added a new ecommerce customer in mid-2024. By late 2025, that customer's top 30 SKUs were generating more picks per day than half the legacy A-movers combined. But the WMS still had them classified as C items because the velocity recalculation job had never been configured for the new customer's product group. Pickers walked to the back of the warehouse 40 times a shift to retrieve what should have been golden-zone product. The WMS was not wrong. Its inputs were.
This is one of the most common sources of invisible travel time waste: a WMS executing perfectly according to outdated rules.
The Symptoms Your Operation Is Already Showing
Slotting decay leaves fingerprints. You just have to know which ones to look for.
If your pick rate variance between identical shifts is running above 15 percent without a clear staffing or volume explanation, slotting is usually involved. When product sits in the wrong location tier, some shifts get lucky because wave composition happens to favor properly-slotted SKUs. Other shifts don't. Same headcount, same WMS, very different output.
If your replenishment team is running more emergency fills per shift than they were six months ago, that's another signal. Forward pick locations sized for lower-velocity product run dry when demand increases and the slot dimensions were never adjusted. Pickers get redirected to reserve storage mid-wave. Travel time spikes. The replenishment crew looks understaffed, but the real issue is that the pick face was dimensioned for demand that stopped being accurate months ago.
If supervisors are manually overriding directed putaway, telling associates to put product somewhere other than where the WMS sends it, that's the loudest signal. It means the people on the floor already know the system is wrong. They've been working around it instead of fixing the root cause.
None of these point to a labor problem. They point to location logic that's out of date.
Building a Slotting Review Cadence That Works
The fix isn't a one-time re-slotting project. That's what got you here. What works is treating slotting as a recurring discipline with a defined cadence and clear ownership.
Start by pulling 60 to 90 days of pick data and running an ABC analysis: rank SKUs by pick frequency, not units moved. A SKU that ships in large quantities once a week puts very different demand on a forward pick location than one picked in small quantities 40 times a week. Both might move the same total units, but the slot-level operational demand is completely different. Frequency is what drives travel. Units moved hides it.
Run a velocity reclassification quarterly at minimum. For operations with seasonal product mixes or rapidly changing customer profiles, monthly is better. The WMS can do this automatically if the recalculation rules are configured properly, but someone has to own the configuration and verify the output.
Focus physical re-slotting on the highest-impact mismatches. You don't need to move every SKU. Identify the top 50 to 100 items where the velocity tier in the WMS doesn't match actual pick frequency from the last 90 days. Move those. You'll get most of the benefit with a fraction of the disruption.
And audit the WMS velocity rules themselves. Check whether the recalculation job is actually running, whether it covers all product groups and zones, and whether the tier thresholds still make sense for your current order profile. A re-slotting effort that uses stale velocity data is just moving product from one wrong location to another.
The operations that stay ahead of this don't have better WMS software or a bigger CI team. They have someone who owns slotting accuracy the way ICQA owns inventory accuracy, as an ongoing discipline, not a project. The WMS won't tell you the classifications are wrong. It'll just keep executing them at scale until someone updates what the system knows.