How to Run a Warehouse Slotting Analysis Without Software

Every warehouse has a SKU that moves forty times a week and lives in the back corner of the building on the bottom shelf. Usually nobody put it there on purpose. It went into an empty location during a busy receiving week two years ago, and it has been there ever since, quietly costing about ninety extra seconds a pick.

Slotting analysis is the exercise of finding those and fixing them. The software that does it automatically starts in the five figures and assumes you have clean cube data for every item, which most operations don't. The manual version takes a spreadsheet and a couple of days, and for a facility under a few thousand SKU locations it gets you most of the way there.

---

TL;DR: Slotting is the difference between a picker walking to the item and the item being where the picker already is. This guide runs the analysis in six steps: pull twelve months of pick line data, rank SKUs by pick frequency rather than revenue, map the golden zone in your actual building, check cube and weight before committing to any move, execute the moves in waves so picking never stops, and measure travel time before and after so you know whether it worked. No software required. Start by exporting one column you already have — pick lines by SKU — and sorting it descending.

---

Step 1: Pull the pick data you already have

You need one thing: pick lines per SKU over the last twelve months. Not units shipped, not revenue — the number of times a picker went to that location and took something. That count is what generates travel.

Every WMS exports this. If you are running on paper or a spreadsheet, your order history has it: count the number of order lines each SKU appeared on.

Twelve months matters because it catches seasonality. A SKU that does nothing for ten months and then carries your Q4 is not a slow mover, and slotting it as one is how you build a beautiful layout that falls apart in November. If you only have six months, use it, but note which months you are missing and check your peak SKUs by hand.

Two columns is enough to start: SKU, pick lines. Add current location if you can get it in the same export — you will need it in step three, and joining it later by hand is miserable.

Step 2: Rank by picks, not by revenue

Sort descending by pick lines and look at where the curve bends. In most operations it bends hard. A fifth of the SKUs generate somewhere around half to three-quarters of the picks, and the exact split matters less than the shape: there is a small group doing most of the walking.

Band them. A common cut is the top 20% by pick lines as A items, the next 30% as B, the remainder as C — but the bands are a tool, not a rule, and if your curve bends at 12% then band it at 12%.

The word "revenue" should not appear anywhere in this exercise. Your highest- margin SKU might get picked twice a month, and slotting it in prime real estate costs you every day for the benefit of two picks. Slotting optimizes labor, and labor responds to trips, not to dollars.

One check worth doing before you go further: pull the bottom of the list and look at what has been picked zero times in twelve months. In most facilities that is a real number of locations. Those are not a slotting problem — they are an obsolescence conversation with your customer, and the fastest square footage you will recover all year. If that inventory belongs to a 3PL client, it is also a conversation about their storage habits rather than one about your layout.

Step 3: Map your golden zone

The golden zone is the band of your racking between roughly mid-thigh and shoulder height where a picker can take a case without bending or reaching. Call it 30 to 60 inches. Picks from it are faster and they are the ones that don't generate back injuries.

Draw your actual building. Not the CAD file — a whiteboard sketch with the pick face, the aisles, and the P&D station. Then mark two things:

Vertical: which levels in each bay are golden zone. In most selective racking that is the floor level and the first beam level, and everything above is a reserve or a step-ladder pick.

Horizontal: distance from the packing station along the actual pick path. The closest bay in the aisle nearest to pack is your best location; the far end of the last aisle is your worst. If your pickers walk a fixed route, rank locations by their position along that route rather than by straight-line distance, because the route is what they actually walk.

Now you have a ranked list of SKUs and a ranked list of locations. The analysis is joining them.

Step 4: Check cube and weight before you move anything

This is the step people skip, and it is the one that turns a good analysis into a plan that doesn't survive contact with the building.

Does the item physically fit the location you want to put it in? An A item that ships two pallets a week needs a location that holds two pallets a week, or you have traded a long walk for a replenishment every day — and a replenishment is a full trip by a lift operator, which usually costs more than the pick trip you saved.

Is it too heavy for where you are putting it? Heavy items belong at or below waist height regardless of velocity. A 60-pound case at shoulder height is an injury waiting to be recorded, and no travel-time saving is worth it. Heavy and fast means floor level in the closest bay; heavy and slow means floor level somewhere else.

Does it need to be somewhere specific for another reason? Food-grade segregation, temperature zones, hazmat separation, lot control, or a customer who has audited and approved a location — all of these outrank velocity. Note them as fixed before you plan a single move.

The output of this step is your move list with the impossible moves already struck off it.

Step 5: Move in waves, not in a weekend

The instinct is to shut down on a Saturday and re-slot the building. Resist it. A mass move means every location in the WMS is wrong at once, every picker is learning a new building on Monday, and if anything goes wrong you cannot tell which change caused it.

Move in waves instead:

1. Start with your top 20 SKUs. They are the biggest win and the smallest number of moves. If your data is wrong about anything, you find out here while it is cheap to unwind.
2. Move one aisle or one zone at a time, and update the system location as each move completes rather than at the end of the day. A pallet that has physically moved but not systematically moved is a lost pallet.
3. Tell the pickers before you move, not after. They have the building memorized. The people who will notice your data was wrong about a SKU are the people who pick it every day, and they will tell you in the first hour if you have given them a reason to.
4. Cycle count what you moved. Every touch is a chance to lose count accuracy, and a re-slot is a lot of touches. If you don't have a counting program, start one before you start moving inventory around.

A wave a week for a month beats a weekend, and you can stop after any wave.

Step 6: Measure whether it worked

Take the baseline before you move anything, because you cannot reconstruct it afterward. The two numbers worth having:

Lines per labor hour, measured over a full week, for the pick function only. It's crude and it's affected by order mix, but over enough weeks it moves in the right direction if the slotting worked.

Travel distance per order, if your WMS reports it. Most do not. A workable substitute is timing a picker on a representative multi-line order — same picker, same order profile, before and after.

Give it four weeks before you judge. The first week will look worse than the baseline no matter what you did, because everyone is learning where things are. If it still looks worse at week four, something in the analysis was wrong, and it is usually step four.

Then put it on a calendar. Velocity drifts — a SKU that was an A item in March is a C item by September, and the layout you built decays quietly. A quarterly re-run of steps one and two takes an hour and tells you whether the moves are worth doing again.

For the metrics side of this and what else is worth tracking on a pick operation, the field guide on warehouse KPIs that actually matter covers what to measure and what to leave alone.

The minimal version

If two days is more than you have: export pick lines by SKU for the last twelve months, take the top twenty, and find out where they physically are. Move the ones that are in bad locations into good ones. That is a morning's work, it captures a large share of the available saving, and it will tell you whether the full exercise is worth scheduling.