When I time pickers in a small warehouse, the same thing shows up almost every time. More than half the picking time is walking and searching. Reaching for the product is a small slice.

Slotting is how you shrink the walk: deciding where each item lives based on how often it's picked. For a building with a few thousand SKUs or fewer, a spreadsheet is enough.

Rank by pick lines

The most common mistake I see is slotting by sales dollars or units sold. Neither tells you how often a picker walks to the location.

You want pick frequency, meaning how many order lines included the item. An item that ships 1,000 units on one order a month gets visited once. An item that ships one unit on 300 orders gets visited 300 times. The second one belongs up front.

From your ERP, export order line history for the last 6 to 12 months with item number, order number, ship date and quantity. If you're seasonal, use the full 12 months or slot peak separately. If your product mix changed a lot recently, use the last 3 to 6 months instead.

Build a pivot with items as rows and a count of order lines as the value, sorted descending. That's your velocity ranking. Add a running percentage of total pick lines. In almost every warehouse I've done this for, something like 15 to 20 percent of SKUs account for 80 percent of the picks. That top group is what you're moving.

Add size

Frequency alone isn't enough. A fast mover the size of a dishwasher can't go in a waist-high shelf bin.

Add the item's dimensions, or at least a rough size class (small, medium, large, pallet). Then add one more column: average daily units picked times unit cube. That tells you how big a location the item needs so you aren't replenishing it four times a shift.

Score your locations

The golden zone runs from about knee to shoulder height, closest to where picking starts and ends, which is usually the pack stations. Picking there is faster and easier on backs.

Walk the building and mark it. For each aisle, note how far it is from pack, which levels sit between knee and shoulder, and which locations are on the main travel path. Then score each location. I usually use 1 for golden zone near the front, 2 for golden zone farther back or the levels just above and below it up front, and 3 for everything else. Put the score in a column beside the location list.

Match them up

Now you have two sorted lists, items by pick frequency and locations by score. Match from the top down, checking that each location is big enough for the item. While matching, I keep a few rules:

  • Put items that are often ordered together near each other. A quick look at order history for common pairs is worth the time.
  • Keep heavy items low, whatever their velocity. Nobody should pull a 40-pound case from above shoulder height 200 times a day.
  • Don't split the top movers across opposite ends of the building just because there was room.
  • Leave a few golden zone slots open. New fast movers will come.

One exception to keeping fast movers together: if your top five items share one bay, pickers will queue for each other at peak. Spread the top handful across two or three bays at the front.

Plan the move

A re-slot moves a lot of inventory, and it can wreck accuracy if you rush it. Do it in a slow period or on a shift with no picking, in a planned sequence so you don't need dozens of temporary spots. Record a location transfer for every move as you make it, not at the end of the day. Then count each item at its new home. That count catches mistakes while they're still easy to fix.

Measure before and after

Before you move anything, time a sample of picks, maybe 30 order lines across a couple of pickers, and write down lines per hour. Two weeks after the re-slot, time the same kind of sample.

In small warehouses I typically see pick rates rise 15 to 30 percent from a first velocity slot. Buildings that were already reasonably organized see less. If you see no improvement, travel wasn't your bottleneck, and the next places to look are packing, replenishment and how orders get released.

How often to do it again

This is where people overdo it. A full re-slot is a lot of labor and you don't want it monthly. For most small operations I'd do a full re-slot once a year, ahead of peak. Once a quarter, rerun the pivot, compare the top 50 against where they live now, and move only the handful that drifted. That takes about an hour.

And slot new items properly on day one. "Wherever there's room" is how the building got this way.

When the spreadsheet runs out

A spreadsheet works up to a few thousand active SKUs in one building. Past that, or if your mix changes weekly, WMS slotting tools start paying for themselves. Even then, do it by hand once first. You'll learn what your building actually needs, and you'll be able to tell whether the tool is doing it well.