A walkthrough on a 1,500-SKU synthetic portfolio of the inventory-policy call I run with retail clients. Code, data, and charts are all in the project repo. Methodology is real. Numbers come straight out of the simulation.
The Challenge
Every retail DC eventually outgrows continuous review. The math that worked at 200 SKUs breaks at 2,000. By the time the portfolio is that wide, "reorder when stock dips below s" stops being a policy and becomes a daily firefight — every working day has hundreds of replenishment lines firing, every truck is half-full because nothing coordinates, every store gets its own bespoke wave.
The team I usually walk through this with has the same instinct: switch to periodic review and expect a 30% drop in order volume. That's the wrong number to chase. The simulation below is how I show them why.
The Approach
I needed clean answers to three questions: does periodic review actually cut order volume, what does it really do, and what does it cost.
1. Build a realistic SKU landscape. 1,500 SKUs with lognormal mean daily demand — median ≈ 1.8 units/day, long tail running into the 40s. Per-SKU sigma 25–45% of mean. Lead times uniform 1–3 days. Standard fixture every retail planner recognizes.
2. Simulate both policies on the same demand panel. Continuous (s, Q) with reorder point sized for 95% cycle service and Q at one week of mean demand. Periodic (R, S) with R = 7 days and S sized to cover R + lead time plus matched safety stock. Both policies target the same service level by construction — that's the only way the comparison is honest. Then 100 days simulated SKU by SKU.
3. Track what actually matters. Total orders is the obvious metric. The interesting ones are: how many days have at least one replenishment landing in the warehouse, how much inventory is on hand on average, how many SKU-days end in a stockout.
The Outcome
Periodic review did not cut order volume. 20,730 orders vs 21,603 — essentially flat. The team that walks in expecting fewer lines walks out reading the wrong chart.
The cadence number is where the real story lives. Continuous review fires orders on every single one of the 100 simulation days. Periodic review concentrates everything onto 15 days — the 14 weekly review cycles plus day zero.
That's an 85% reduction in days where the DC has to coordinate inbound. It's the difference between a warehouse that's reactive every morning and a warehouse that knows Mondays are review day, schedules carriers around it, and runs a clean dock the other six days. The cumulative-orders chart shows the same thing from a different angle — periodic review marches up in clean weekly steps while continuous review crawls along noisily every day.
The honest trade-off: periodic review carries about 33% more inventory per SKU (12.5 → 16.7 average units) and runs slightly higher stockout-days (317 → 451 across 150,000 SKU-days — both under 0.3%). You're trading a small amount of working capital and a sliver of service risk for a dramatically simpler operation. The sample-SKU profile makes the shape of that trade-off readable: same demand, same service target, very different cadence and inventory signature.
The lesson for any operations director making this call: pick periodic review when your bottleneck is coordination, not when your bottleneck is inventory. The win isn't in the order count — it's in the cadence. And the cost lands on the balance sheet, which is the one place you can size precisely before you commit.