Bottom line first: a warehouse replenishment trigger point is one formula. Reorder point equals average daily demand multiplied by replenishment lead time, plus safety stock. Take daily demand from the last 28 days, measure lead time in hours from task creation to stock on the shelf, and give safety stock only to items whose demand actually swings. All three numbers come from your own system. Never copy a ratio off the internet.

I have seen this exact failure three times, and the third time still stung. In September 2023, I was on a go-live at a 3PL warehouse in Ontario. Every pick-face trigger had been set by hand. A fast mover sat at 12 units, and the replenishment crew ran four trips a day and still stocked out. A slow mover sat at 40 units and had not moved in three weeks. By week two of go-live, pickers were losing more than 90 minutes a day waiting on stockouts. The crew was not the problem. The trigger points had no math behind them. They were guesses.

Why Your Warehouse Replenishment Trigger Point Keeps Going Wrong

The math is rarely the hard part. The inputs are wrong. Daily demand gets averaged over last month, and then peak season doubles it overnight. Lead time gets written down as supplier lead time, when the real clock runs from the system raising the alarm to the goods sitting on the pick face, with task dispatch, travel, and putaway in between. Safety stock gets added as a flat percentage, so stable items waste pick-face space while volatile items still run dry.

A quieter mistake is merging the reorder point and the minimum stock into one number. They are two different signals. The reorder point says "time to replenish." Minimum stock says "you are about to stock out." Collapse them into one and the system either screams all day or nobody listens when it does. Keep them apart: crossing the reorder point creates a normal task, dropping below minimum stock escalates it to urgent.

Here is the counterintuitive part: setting the trigger too high costs more than setting it too low. Set it too low and the damage is visible — a picker standing idle, a supervisor shouting. Set it too high and the pick face fills with slow stock, fast movers have nowhere to go, and the extra replenishment trips nobody counts quietly pile up. In the projects I have worked on, the hidden waste from triggers set too high has run an order of magnitude above the cost of stockouts, because nobody gets blamed for it.

Setting the Warehouse Replenishment Trigger Point in Five Steps

Step one, pull the data. Export daily pick-face issues per item for the last 28 days — pick-face issues only, not full-pallet shipments. I settled on 28 days after trying shorter and longer windows: 7 days is too short and one promotion wrecks it; 90 days is too long and drags stale numbers through a season change. From that export you need two numbers per item: average daily demand and the standard deviation of daily demand. Most WMS reports carry both. If yours does not, export the raw issue records and calculate them.

Step two, measure replenishment lead time yourself. Do not ask the supplier. Do not ask the supervisor. Walk one trip. Start the clock when the system creates the replenishment task and stop it when the goods are on the pick face and confirmed in the system. At that Ontario site the measured time was 4.5 hours, and 3 of those hours were travel and putaway — dispatch and confirmation took only 1.5 hours. Measure in hours, not days. Measuring in days assumes the replenishment lands the same day, and most of the time it does not.

Step three, calculate safety stock, and only for two kinds of items: those whose daily demand standard deviation exceeds 30% of average daily demand, and those whose suppliers habitually deliver late. The formula is simple: safety stock equals 1.65 times the standard deviation of daily demand times the square root of lead time in days. The 1.65 factor corresponds to a 95% service level. That is enough. Do not chase 99% — those last four points cost double the inventory.

Step four, calculate the reorder point and the minimum stock. Reorder point equals average daily demand times lead time, plus safety stock. Minimum stock equals half the reorder point, rounded down to the smallest replenishment unit. Take one item: average demand 20 units a day, lead time 0.5 days, safety stock 15 units. The reorder point is 25 units and the minimum stock is 12. Write both numbers into the item master. The system raises a task automatically when stock falls below the reorder point and flags it urgent below minimum stock.

Step five, run it for two weeks, then tune. After the first version goes live, watch two reports every day: the stockout count and the on-time completion rate of replenishment tasks. Any item that stocked out gets its reorder point raised 10%. Any item that never stocked out and whose pick face keeps running full gets its reorder point cut 10%. Two weeks of that and the numbers settle. Nobody gets the warehouse replenishment trigger point right on the first calculation. It is tuned, not computed.

Approach When it fits Cost
Fixed reorder point Stable-demand consumables Stockouts in peak; needs a manual seasonal review
Rolling 28-day average Most items; the recommended default Recalculated weekly; needs system support
Forecast plus safety stock Promotional and seasonal items Complex math; skip it in a small warehouse

The Go-Live Night Checklist

I have seen this failure three times, always on go-live night. First, the replenishment unit does not match the sales unit. The system replenishes by the case and picks by the piece. The trigger is 25 pieces, the system tops up one case of 24, and the task sits open forever. The check: on go-live night, print the replenishment unit, sales unit, and minimum replenishment quantity side by side for every item and read them row by row.

Second, one item lives in two pick faces. The system watches total stock across both faces. Each face holds 5 units, the total is 10, nothing triggers, and both faces are nearly empty. The fix is to set trigger points per location, not per item. Two pick faces means two sets of numbers.

Third, safety stock quietly becomes permanent stock. In March 2024, at an e-commerce warehouse in Los Angeles, nobody had reviewed safety stock for six months. At the count, safety stock occupied 22% of pick-face locations, and half of those items had shipped only 3 times in the previous 90 days. My rule now is a monthly clean-up: any safety stock untouched for 90 days goes to zero, and it comes back only if demand turns volatile again.

Fourth, the trigger changes but task priority does not. Raising reorder points doubles the task volume, and the crew keeps working in the old order, with urgent tasks queued behind routine ones. On the day you change a trigger, change the sorting rule in the same breath: anything below minimum stock goes to the top of the queue.

One more habit: triggers are not set-and-forget numbers. Spend half an hour at the start of each month recalculating the 20 items with the most stockouts last month, and recalculate everything one month before peak season. Once the demand mix shifts, old triggers are instantly stale. A regular review matters more than one accurate calculation.

So here is my question for you: how many of the trigger points in your warehouse were actually calculated, and how many were guessed? List the guessed ones, run them through these five steps, and check the stockout count two weeks later. The numbers will give you the answer.

By Lin Xiao, WMS Implementation Consultant