Why your pickers walk more than they pick — and how to fix it without changing your WMS

Ask a warehouse manager what their pickers do all day, and they’ll answer “they pick orders.” Measure it, and the answer changes: in a typical manual picking process, 50 to 60% of a

picker’s time is spent walking between locations, not picking. Scanning, grabbing, dropping into the bin — the part that actually creates value — only accounts for a minority of the job.

This gap is the main productivity lever in most warehouses, and it usually goes untapped because it looks like a fixed cost tied to the building.

It’s not.

Where does all this walking come from?

Distance walked isn’t one problem, but three problems that add up:

- Order-by-order picking. Each order becomes a separate trip, even when three consecutive orders go through the same aisles.

- Naive routing. The pick list is sorted by location code, which follows rack numbering rather than the shortest physical path. Add one-way aisles, blocked crossings, or zones

prioritized for forklifts, and the “sorted” route involves multiple U-turns.

- Slotting that has drifted. High-turnover SKUs were placed smartly three years ago. The assortment has changed, seasonality too, and now A-items are at the back of a B-aisle.

None of these three issues stands out as a mistake. The order ships, the dashboard shows “on time,” and the extra kilometers are quietly absorbed in staffing and overtime.

Why doesn’t the WMS solve the problem?

A WMS is excellent at knowing what is where, and who should go get it. It’s generally not designed to answer in what order, by which path, grouped with which other orders. In most WMS,

routing is just sorting, not optimization, and it has no model of your actual floor — distances, obstacles, traffic rules, or the fact that aisle 14 is closed on Tuesday mornings for restocking.

Switching WMS to get better routing is a two-year project with lots of risk, for a feature that’s at the edge of what a WMS is meant to do. The pragmatic path is an overlay: keep the WMS

as the system of record, and let a dedicated engine take the pick list, apply a real model of the site, and return an optimized sequence of tasks.

What actually reduces distances

Three levers, ranked by speed of ROI:

1. Path optimization for picking

Give the engine a digital twin of the site — aisles, distances, one-way routes, priority zones — and let it calculate the shortest legal path for each task, instead of sorting by location code. On its own, this lever

typically delivers a 5 to 20% reduction in distances walked, depending on traffic constraints. It requires no changes to layout or to how orders arrive.

2. Order batching

Instead of one order per trip, group orders that share aisles into a single multi-order mission, with a picking scenario adapted to your containers and sorting constraints. The number of trips drops, picks per hour go up, and e-commerce growth is absorbed without adding pickers proportionally.

3. Smart slotting

Once flows are modeled, the same digital twin lets you put items back in the right place: fast movers near dispatch, SKUs often ordered together in the same aisles, sales forecasts integrated before the season, not after. With this lever, you can see up to 30% less distance in picking and up to 15% fewer forklift moves for replenishment.

What does 15% look like on the ground?

Take a site with 40 pickers, 7 effective hours per shift, 55% of which is spent walking. That’s about 154 hours of walking per day. Cutting distances by 15% frees up 23 hours per day — the equivalent of three full-time shifts — with no layoffs, no robots, and not a single extra square meter.

Those hours show up in later cut-offs, peaks absorbed without temps, or simply less fatigue at the end of a shift.

Where to start:

1. Measure before optimizing. One week of pick data is enough to calculate actual distances walked and spot back-and-forth trips.

2. Model the site as it is, with its real traffic constraints — not the original blueprint.

3. Start with routing, which doesn’t touch layout or processes, then add batching and slotting once you see the first gains.

The question isn’t whether your pickers walk too much. They do. The question is how many of those kilometers you can recover without touching your WMS.

Want a quantified estimate for your site? Our ROI calculator gives you a ballpark figure in just a few minutes.