Home / Insights

Warehouse operations series · Part 1 of 8

We're Out of Space, or Are We? Warehouse Cube Utilization and Slotting

Contents
  1. The example distributor
  2. Step 1: measure three numbers, not one
  3. Step 2: find where the air is
  4. Step 3: fix it in order
  5. The result
  6. Step 4: compare the fixes with a move
  7. Other levers
  8. Keep it from creeping back
  9. When you really are full
  10. Common mistakes
  11. Key points
  12. Sources

Sooner or later every growing distributor hears it from the warehouse: "We're out of space." Receiving has pallets parked in the aisles, the forklift driver spends the morning shuffling loads to make room, and someone starts pricing a bigger building. A move is one of the largest decisions a small business makes. It means a longer lease, racking, a new layout and weeks of disruption.

Very often the building is not full. The positions are full. The space inside them is half air, part of it holds stock that has not moved in a year, and the products picked every day sit at the back on the top level. This first part of the warehouse operations series shows how to tell the two apart, measure it in an afternoon, and win the space back before signing a lease.

The example distributor

The series follows one illustrative business: a distributor in a 40,300 sq ft building with a 9.1 m (30 ft) clear height and about 3,000 SKUs. Most SKUs are small parts on shelving. The 520 bulkiest SKUs are stored in a pallet-rack zone of 10 rows by 20 bays, two pallets a bay, served by 5 aisles 3.6 m wide for counterbalance forklifts. As built the racks have four levels, so the zone holds 1,600 pallets. It is the zone everyone says is full, and it is the one this post works on.

Step 1: measure three numbers, not one

"Full" can mean three different things, and each is measured differently.

MeasureWhat it countsHow to get itWhat it tells you
Floor utilizationShare of the floor covered by racks, shelving and stored goodsThe layout drawingWhether the floor plan is efficient. Says nothing about height
Location utilizationShare of the storage positions that hold somethingA count from your WMS, ERP or a walk-throughHow full the warehouse feels. This is the number behind "we're full"
Cube utilizationShare of the volume of the occupied positions that product actually fillsLoad heights against the clear height of each openingHow much air you are storing. This is where the hidden space is

The two that matter here are calculated as:

\[\text{Location utilization} = \dfrac{\text{positions holding a pallet}}{\text{all pallet positions}}\]

\[\text{Cube utilization} = \dfrac{\sum \text{load volume}}{\sum \text{volume of the openings those loads sit in}}\]

Published benchmarks help put the results in context. Conventional racked warehouses usually run at about 80 to 85% of their locations occupied, and around 85% is where congestion starts: receiving slows, pallets wait in the aisles and loads get handled twice. The same benchmark sources point out that a building can show more than 80% of its positions occupied while product fills well under 60% of the rack volume, because loads are shorter than the beam openings and partial pallets sit in full-size slots. The gap between the two numbers is the space you can recover.

For the example distributor, the count gives 1,469 pallets in 1,600 positions, a location utilization of 92%. That is past the 85% congestion point, and it is why the warehouse feels full. Measuring the load heights tells a different story: the loads fill only 53% of the openings they occupy.

Step 2: find where the air is

Space in a pallet-rack zone is usually lost in the same five places.

  • Beam openings set for a load you no longer receive. The racks were installed for tall pallets years ago, and today's loads are 1.2 to 1.4 m. Every opening carries half a metre or more of air.
  • Partial pallets. Picking leaves broken pallets, and two or three partial pallets of one SKU each take a full position. In the example, 337 of the 1,469 pallets are partial.
  • Dead stock. Product with no sales in twelve months still takes a position. The example has 62 such SKUs on 163 pallets.
  • Put-away to the first free slot. Without slotting rules, fast movers end up wherever there was a hole on the day they arrived, which is often high and at the back.
  • The wrong storage for the item. A pallet position holding three cartons of a small item is a position lost. Small, slow items belong on shelving or in carton flow, not on a pallet.
Highest load top: 8.65 m (18 in below the sprinklers)As-is: 4 levels, 2.0 m beam pitch8 positions · 54% of the opening filledRe-set: 5 levels, 1.7 m pitch10 positions · 76% of the opening filled→Air above the loadPartial pallet
Figure 1. One bay as built and with the beams re-set. The same rack height holds five levels instead of four once the openings match today's loads.

Step 3: fix it in order

The fixes below are applied one after the other, cheapest and most certain first. Each builds on the one before, which is how Figure 2 runs them.

Fix 1: re-set the beam heights. Measure the tallest load each SKU really arrives on, then re-space the beams to suit, with about 100 mm of lift-off clearance. In the example the floor level stays 2.0 m high for the 73 tall loads, and the beams above move to a 1.7 m pitch. The same uprights now carry five levels, and the top load still stays 450 mm (18 in) below the sprinkler deflectors. Positions rise from 1,600 to 2,000. Treat this as an engineering change: the rack manufacturer or a rack engineer confirms the new configuration's capacities, new load signs go up, and the fire protection contractor checks the sprinkler clearance. Re-beaming also means emptying each bay in turn, so plan it bay by bay over quiet weeks.

Fix 2: combine partial pallets. Where a SKU has two partial pallets whose stock fits on one, combine them. In the example that removes 124 pallets and frees the same number of positions. Then add a rule so it stays fixed: a pallet below about a third full is consolidated before a new one is put away.

Fix 3: clear the dead stock. List the SKUs with no picks in twelve months and decide each one: return it to the supplier, sell it to a liquidator, move it to cheaper overflow storage, or scrap it. The money was lost when the stock stopped selling. Keeping it in a prime position only adds the cost of the space. Clearing the example's 149 dead pallets frees 149 positions.

Fix 4: re-slot the fast movers. Rank the pallets by picks a day, as in the ABC inventory analysis, and give the busiest the best positions: the floor and first beam level, nearest the front. Slow movers and reserve pallets go up and to the back. Re-slotting does not add positions, but it decides how far people travel. Reserve pallets follow their pick face, A first, then B, then C, so each class sits above and behind its own picks. In the example that is a full re-slot of 1,195 pallet moves, done in waves over a few weekends, and it cuts the average walk per pick from 95 to 45 m. Published first-time slotting projects often report travel savings of about 20 to 40%. The example saves more because its put-away was close to random.

Figure 2 is the example distributor's pallet-rack zone in 3D, built from the same data as the tables. It steps through the four fixes on its own, about fifteen seconds each. Pick a step to stay on it, colour the pallets by how full their opening is or by how often they are picked, and drag to turn the building.

As-is
Positions occupied92%
Opening filled (cube)53%
Free positions131 of 1,600
Pallets in the racks1,469
Walk per pick95 m
Picks within easy reach16%

As-is: 92% of the positions hold a pallet, so the building feels full, yet the loads fill only 53% of the openings they sit in.

Colour by
Figure 2. The example distributor's pallet-rack zone: 10 rows of 20 bays, two pallets a bay. The fixes apply in order and build on each other. Walk per pick is the round trip from the front cross-aisle, weighted by picks a day. "Easy reach" means the floor or first beam level in the seven bays nearest the front. Loads are drawn at their real heights. Shelving, packing and docks are simplified.

The result

StepPositionsPalletsOccupiedOpening filledWalk per pick
As-is1,6001,46992%53%95 m
1. Re-set beam heights2,0001,46973%66%95 m
2. Combine partial pallets2,0001,34567%72%95 m
3. Clear dead stock2,0001,19660%72%95 m
4. Re-slot fast movers2,0001,19660%73%45 m

After the four fixes the zone holds the same live stock in 60% of its positions, with 804 positions free instead of 131. The loads now fill 73% of their openings. The busiest pallets sit within easy reach: 93% of picks come from the floor or first beam level near the front, against 16% before.

As-is131 free positionsRe-set beam heights+400Combine partial pallets+124Clear dead stock+149Re-slot fast moversNo new positions · walk per pick −52%After the four fixes804 free positions
Figure 3. Where the free positions came from. Re-slotting adds no positions, but it decides how far people walk.

Step 4: compare the fixes with a move

The 673 recovered positions would need about 7,100 sq ft of extra racked floor as the zone was built. Canada's national average net asking rent for industrial space was $14.77 per sq ft a year in the third quarter of 2026. With an assumed $5.00 of additional rent for taxes and common costs, that floor would cost about $140,000 a year, before the cost of moving, new racking and the disruption. The fixes are a one-time cost.

One-time cost of the four fixesBasis (assumed)Estimate
New beam pairs for the fifth level200 bays at $150 a pair$30,000
Rack engineering review, load signsAllowance$6,000
Labour: re-beam, unload and reload each bay496 hours at $40$19,835
Labour: combine, clear and re-slot130 hours at $40$5,207
Total$61,000

About $61,000 once, against about $140,000 every year. There is also time. With 1,196 pallets in 2,000 positions, the zone can grow by 10% a year for about 4 years before it reaches the 85% congestion point again. That is time to plan a move properly, if one is ever needed.

Other levers

  • Narrower aisles. Counterbalance forklifts need aisles of about 3.5 to 4 m. Reach trucks work in about 2.7 to 3 m, which can add a rack row in the same building. It needs new trucks and operator training, so it comes after the cheaper fixes.
  • Right-size the storage. Move small, slow items from pallet positions to shelving, carton flow or bins, and very slow items to a mezzanine or high shelving.
  • Clear the floor. Staging areas full of stock waiting for put-away usually point to a receiving problem, not a space problem. Part 3 of this series deals with the dock.
  • Overflow for the peak. If space runs short for six weeks of the year, short-term overflow storage costs far less than a bigger building for all twelve months.

Keep it from creeping back

Slotting drifts. New products arrive, sales patterns change, and within a year the fast movers are back at the top and the back. A short routine keeps the gains.

MeasureHow oftenSignal to act
Location utilizationWeeklyAbove 85% for more than a few weeks
Cube utilization of occupied positionsQuarterlyMore than about 15 points below location utilization
Partial palletsMonthlyMore than one partial pallet per SKU
SKUs with no picks in 12 monthsQuarterlyAny in the front or lower positions
Picks from easy reachMonthlyFalling as the product range changes

When you really are full

Sometimes the answer is a bigger building. The signs are clear once the fixes are done: location utilization stays above 85% after the beams, partials and dead stock have been dealt with, the cube utilization of occupied positions is already above about 75%, and the sales forecast takes you past the limit within a year. Then the measurements above become the brief for the new building: the pallet count, the load heights and the pick rates it has to serve.

Common mistakes

  • Counting positions and calling it space. Location utilization tells you how full it feels. Cube utilization tells you how full it is.
  • Moving beams without the rack maker. Every configuration has rated capacities. Re-set beams under an engineer's review and post the new load signs.
  • Stacking into the sprinkler clearance. An extra level is no gain if the top load blocks the sprinklers.
  • Re-slotting once. Without a monthly check the fast movers drift back.
  • Keeping dead stock because it was expensive. The cost is already spent. The space it takes is a new cost.

Key points

  • A warehouse can be full by positions and half empty by volume. Measure location and cube utilization.
  • Around 85% of positions occupied is where congestion starts. Above it the building feels full whatever the volume says.
  • Fix in order: beam heights, partial pallets, dead stock, then slotting.
  • In the example the four fixes free 673 positions and cut the walk per pick by 52%, for a one-time cost of about $61,000 against about $140,000 a year for equivalent leased space.
  • Measure again every month, or the gains drift away.

Related reading: ABC inventory analysis ranks SKUs by value and picks, ABC analysis for manufacturers applies it in a plant, and setting safety stock keeps reserve pallets to what is needed. Next in this series: Part 2 follows the pickers through the same warehouse and compares pick methods.

Sources

The distributor, its building, inventory and costs in this series are an illustrative example. The benchmarks are published industry figures. Changes to racking need the rack manufacturer's or a rack engineer's confirmation, and sprinkler clearances need your fire protection contractor and the authority having jurisdiction.

Keep reading

Related articles

All 168 articles

Ready to find your next improvement?

Book a strategy call. We will talk through your goals, pain points and where AI, systems or process work can make the biggest difference.