You don't need a WMS to map your warehouse flow

One of the first things a warehouse consultant asks for is a WMS report. Pick paths, travel times, throughput by shift — all of it neatly exported, timestamped, ready to analyse. It is a reasonable starting point when the system exists.

But plenty of operations do not have that system. Paper-based picking is more common than the trade press suggests. Many smaller sites run on spreadsheets, manual pick sheets, and tribal knowledge. And even sites with a WMS often find the relevant data is in a spreadsheet export or a hand-annotated PDF rather than a clean database view.

The absence of a WMS is not a barrier to understanding how your warehouse moves. It just means you work from what you have. In most cases, that is enough.

Paper runs more warehouses than software does

Mid-sized warehouses — thirty to two hundred thousand sq ft, a handful of pickers, regional distribution rather than national fulfilment — are underserved by warehouse management software. The enterprise systems are overbuilt and expensive; the cheaper options often lack the implementation support to stick. Many of these sites run on pick sheets printed from a spreadsheet, scanners that feed a basic stock system, and experienced pickers who know the building.

The data exists. Orders get picked and dispatched. SKU movements are recorded somewhere — in a stock control system, in a spreadsheet, in a paper record that someone reconciles at the end of the day. It may not be tidy, and it may require some work to extract, but it is there.

What these sites lack is a way to turn that data into insight about the layout. There is no dashboarding, no travel-time report, no heat map. The manager knows the warehouse is busy and that picking takes too long, but they cannot easily quantify what “too long” costs or where the specific problem lies.

That is the gap. And it does not require a WMS to fill it.

What you actually need: a layout and a pick list

To model warehouse flow, you need two things: a description of where the stock lives (the layout) and a record of what got picked (the pick history).

The layout can be a CAD drawing, a hand-drawn floor plan, or a sketch on graph paper. It needs to show the racking positions with enough accuracy to trace routes — which aisles exist, roughly how long they are, and where dispatch is. It does not need to be an architect’s drawing.

The pick history can be a week’s worth of completed pick sheets, a month of stock movements from the stock control system, or a spreadsheet of order lines with location codes against each. It needs location codes (bin references, aisle labels, shelf numbers — whatever your system uses) and ideally a date or order reference so you can group picks into runs.

That is the minimum. With those two inputs, you can trace routes and estimate travel distance for a representative set of orders. From travel distance, you can calculate a cost.

More data helps — a longer history catches seasonality, a more detailed layout captures aisle widths and cross-links — but you can start with remarkably little. The first analysis is usually about order of magnitude: is the walking bill tens of thousands a year, or hundreds of thousands? That distinction matters even if the exact figure is uncertain.

Sketch the aisles, run the flow

The practical starting point is to draw the layout as accurately as you can, at a scale that lets you measure routes. Metre-square graph paper works well. Mark the aisle positions, the racking blocks, and the dispatch area. Label the aisles and sections to match your location codes.

Then trace a representative set of picks. Take twenty or thirty completed pick runs — one picker, building one outbound load — and mark the locations on the plan in pick-sequence order. Connect them with a line. Measure the line.

Do this for a handful of runs and you will have a reasonable estimate of average travel distance per run. Multiply by the number of runs per day and you have a daily travel distance. Multiply by the loaded labour rate and you have a cost.

The analysis will be imperfect. You will not have captured every shortcut or every inefficiency. But it will be directionally correct, and it will tell you whether the walking bill is a problem worth solving or a rounding error.

The next question is usually: what would change if we moved things around? And that is where a model earns its keep, because rearranging stock by hand and re-tracing routes is tedious. A model does it instantly.

From picture to pounds

The jump from a sketch to a model is smaller than it sounds. The core of any warehouse flow model is the same as the pen-and-paper analysis: locations, routes, and pick sequences. A model just runs that calculation faster and across more scenarios.

The value of the model is in the comparison. Not “how far do pickers walk?” but “how far would they walk if we moved the top fifty SKUs to the front aisles?” That counterfactual is the thing that justifies action — the £/yr saving from a slotting change, expressed as a number you can put in a business case.

You do not need perfect data to get a useful comparison. You need consistent data: the same pick history applied to the current layout and the alternative layout. The uncertainty is in the absolute figures; the relative difference between layouts is more robust.

WalkBill is designed for exactly this starting point: sketch your aisles in the browser, add a pick history export, press Build, and the model runs. It works without a WMS integration, without a CAD file, and without a specialist consultant to interpret the output. If you would like to see it run on a real site — with your drawing and your pick data — a founding pilot does exactly that.

The warehouse that thinks it cannot be modelled usually just needs someone to start with a pencil.