· 4 min read
How to Shorten a Warehouse Pick Walk
Heshan Fernando
Co-founder & COO
An eighteen-line pick list in the order the customer typed it sends a picker up aisle A, across to C, back to A, out to E, and back to B.
The same eighteen items in aisle order is one pass. Nothing about the warehouse changed, and the walk halved.
Travel is most of the picking time
In manual picking operations, walking typically accounts for a large majority of the time — commonly cited as around half or more, depending on the layout and order profile.
The remainder is the actual picking, scanning, and handling. Those are hard to speed up without changing equipment. Travel is not, and it responds to ordering the list.
The distance itself is not the main lever, either. Most wasted travel is aisle revisits — entering an aisle, leaving, and coming back. Each revisit costs the full length of the aisle twice over, and a list in entry order produces several of them.
Eliminating revisits is where the saving is, and it is achieved by sorting rather than by any change to the warehouse.
Serpentine routing
The standard approach: work up one aisle, across, down the next, and so on, visiting only the aisles containing picks and each of those exactly once.
It is not the mathematically shortest route. Finding that is a version of the travelling salesman problem, and the optimal path is sometimes meaningfully shorter.
Serpentine wins anyway, for a practical reason: pickers can follow it. It requires no memory, no decisions, and no reference to a route diagram. A picker who understands “up this aisle, down the next” executes it correctly at speed.
An optimal route that requires checking a sequence at each junction is slower in practice than a slightly longer one that runs on autopilot, and it produces errors when someone loses their place.
| Approach | Distance | Executable? |
|---|---|---|
| Entry order | Worst | Trivially |
| Serpentine | Near-optimal | Easily |
| True shortest path | Best | Hard to follow |
Slotting matters more than routing
Routing optimises the walk for a given item placement. Slotting changes the placement, and it has a larger effect.
The principle: fast-moving items go closest to the dispatch point, because they are picked most often and every pick saves the walk. Slow movers go furthest.
A common analysis finds a small proportion of items account for a large majority of picks, so moving that small proportion forward shortens the average pick substantially and permanently.
Slotting is more work to implement — physically relocating stock, updating locations — and it compounds with routing rather than competing with it. Routing is the cheap win to do first; slotting is the larger win to do next.
Batch picking, where it fits
Picking several orders in one pass is the other large lever. One walk serves five orders rather than one, and the travel per order drops sharply.
The cost is sortation — the items have to be separated into orders afterwards, either at a station or into compartments on the trolley. That adds handling and introduces a new error mode.
It works well for many small orders with few lines each, which is why e-commerce operations use it heavily, and poorly for large multi-line orders where a single order already fills a trolley.
Common mistakes to avoid
- Picking in order-entry sequence.
- Optimising the route while ignoring where stock is slotted.
- Choosing a mathematically optimal route pickers cannot follow.
- Batching orders without a reliable sortation step.
- Measuring picks per hour without measuring distance walked, which is the thing actually changing.
How to do it with Warehouse Pick Path Optimizer
The Warehouse Pick Path Optimizer reorders the list by location.
- Paste the pick list with a location code per line.
- Choose the layout and routing strategy — serpentine for most parallel-aisle warehouses.
- Give pickers the reordered list rather than the original.
- Look at slotting next; it is the larger and more durable saving.
Other operations tools are in the tools directory.
Frequently asked questions
What is serpentine routing?
Working up one aisle and down the next, visiting each aisle containing a pick exactly once. It is simple to follow and close to optimal in a standard parallel-aisle warehouse.
Why not use the mathematically shortest route?
Because it is hard to follow. A route requiring a decision at each junction is slower in practice and produces errors, while a slightly longer route runs on autopilot.
Does slotting matter more than routing?
Usually yes. Placing fast-moving items near dispatch reduces travel more than any routing change, and the two compound. Routing is the cheaper one to implement first.
Final thought
Sort the list before touching the warehouse. Eliminating aisle revisits costs nothing and captures most of the available saving.