The four numbers you need
Outbound pallet loads per week. Average days before an empty pallet returns to you, if it returns at all. Peak-week multiplier. And your tolerance for running out, expressed as days of buffer.
Everything else is detail. If you can estimate those four numbers you can size a pool properly, and most operations have never written them down.
The one-way case
If your pallets leave with product and never return, your requirement is simply consumption: weekly loads multiplied by the peak multiplier, plus buffer. A plant shipping 240 loads a week with a 1.4 peak factor and one week of buffer needs roughly 670 units on hand at peak.
The mistake here is sizing on average rather than peak. Averages are a reporting concept; your dock lives in peaks.
The closed-loop case
If pallets come back, your pool size is governed by cycle time rather than consumption. Pool size equals daily throughput multiplied by cycle time in days, plus buffer.
The leverage here is enormous. Cutting cycle time from fourteen days to ten reduces required pool size by nearly 30%, which is capital released without buying anything. Chasing cycle time is almost always cheaper than chasing unit price.
Where pallets actually disappear
Three places, in our experience. Customers who keep them. Internal departments using them as shelving, workbenches and ramps. And attrition — genuine breakage, which in a well-run operation runs somewhere around 5 to 12% per cycle depending on handling.
Measure attrition before buying a bigger pool. We have seen sites add 400 units to cover a shortfall that turned out to be two departments quietly hoarding.
What to do with the answer
Put it in a supply program. Reserved stock at a fixed price with a ±20% flex clause removes the two bad outcomes: emergency spot buying at peak, and a yard full of capital you do not need.
We will do this arithmetic with you for free, including the uncomfortable part where the answer is that you are already over-stocked.