Amazon Inventory Control, Explained With the Actual Math

What Amazon inventory control actually is

Amazon inventory control is the set of rules that decide how much stock you hold and when you reorder it — reorder point, safety stock, lead time, and sales velocity. It's the operational layer underneath inventory management, which is the broader job of tracking, storing, and moving product.

People search "inventory planning and control" as one phrase because the two halves depend on each other. Planning is the forecast — how many units you'll need over the next quarter, built from seasonality and trend. Control is the ongoing mechanism that turns that forecast into actual purchase orders and triggers, adjusted as real sales run faster or slower than the plan assumed. A good plan with no control mechanism behind it just sits in a spreadsheet until someone remembers to check it — usually after the stockout.

Amazon's environment adds three variables most inventory theory doesn't account for: a lead time that includes manufacturing plus FBA check-in, an Inventory Performance Index that penalizes both too little and too much stock, and storage fees that jump specifically in Q4. Control on Amazon means running the general formula through those three filters.

The reorder point math, worked through with real numbers

The formula behind reorder point is: Reorder Point = (average daily sales × lead time in days) + safety stock. ABC classification, EOQ, and JIT are all variations on getting these three inputs right.

Take a SKU selling 40 units a day, with a supplier-to-FBA lead time of 35 days, and a 15-day safety stock buffer to cover demand spikes and shipping delays. That's ROP = (40 × 35) + (40 × 15) = 1,400 + 600 = 2,000 units. When on-hand plus in-transit inventory drops to 2,000 units, you place the next order — not before, and not thirty days later because someone was busy.

Days of cover is the sense-check on top: on-hand units ÷ average daily sales. At 2,000 units and 40 units a day, that's 50 days of cover at the reorder trigger — enough to clear the 35-day lead time with 15 days of slack. If your days of cover at reorder time is consistently lower than your lead time, you will stock out before the next shipment lands, full stop. That gap is checkable in about five minutes with your own sales report and your supplier's actual lead time — not the one quoted at signing.

When the reorder point is wrong, or the fix didn't work

Three failure modes cover almost every case.

• You stocked out anyway. Usually the lead time used in the formula was the supplier's quoted figure, not the one that includes customs delays, FBA check-in queues, or a rejected shipment. Pull the actual receive dates from your last five reorders and use that average instead. If it's materially longer than quoted, your safety stock needs to grow, not your reorder frequency.
• You're paying storage fees despite a "correct" reorder point. A reorder point controls when you buy, not what happens to slow-moving stock already on the shelf. Aged inventory surcharges are a separate problem — the decision at that point is remove or liquidate, not reorder less, and it has to be made SKU by SKU against actual sell-through, not category averages.
• The fix worked for one SKU and broke another. Raising safety stock across the board to solve one stockout usually creates excess cash sitting in SKUs that didn't need it. Reorder points are a per-SKU calculation. Applying one buffer percentage to a whole catalog is a shortcut, not control.

Across the accounts inside the Full Circle group — more than $500M in managed revenue across 100+ brands — the gap between quoted and actual lead time is the single most repeated correction we make. It's rarely dramatic. It's usually four to nine days, and four to nine days is exactly the margin that turns "fine" into "out of stock in week three of a launch."

The sales-velocity mistake almost everyone makes

Average daily sales is the input everyone gets wrong, including us. The easy version is a trailing 30-day average, which is fine most of the time and badly wrong exactly when it matters most: right after a promotion, a Prime Day spike, or a deal that pulled forward three weeks of demand into three days. Feed that inflated average into the reorder point formula and you'll order too much, right when demand is about to fall back to normal — the path to an aged-inventory surcharge six months later.

We've made this exact call wrong on an account before: read velocity right off a post-promotion week, sized the next purchase order against it, and ended up with a six-week excess position that took a price drop to clear. The fix isn't a smarter formula — it's excluding promotional and stockout days from the velocity calculation before averaging, and rebuilding the average weekly instead of trusting a rolling window to self-correct.

The mirror-image mistake is a velocity average that includes stockout days, which understates demand — you weren't selling because you had nothing to sell, not because customers stopped wanting it. That one quietly sets up the next stockout too.

What handles inventory control, and what handles something else entirely

Worth being precise here, because the tools get lumped together and they're not doing the same job.

• A 3PL physically stores and ships your inventory. It doesn't decide when to reorder — that's still on you or whoever's watching the numbers.
• An ERP or multichannel inventory system runs the whole business — purchasing, accounting, inventory, often across several sales channels at once. If you sell on Amazon plus your own site plus wholesale, this is usually the right backbone, and nothing built for Amazon alone should try to replace it.
• A product research suite is built to find what to sell, not to manage what you already have. Different job, different buyer.
• Amazon's own tools — FBA, Fulfilled by Merchant, Amazon Supply Chain Services — handle fulfillment and give you the raw data on sell-through, aged inventory, and IPI. The interpretation and the reorder decision are still yours to make.

Brands running a handful of hero SKUs against long overseas lead times — cookware, for instance, where a brand like HexClad carries seasonal gift-set demand against a months-long manufacturing lead time — feel every one of these gaps harder than a seller with fifty low-cost SKUs and same-week reorders. The math is identical. The margin for error isn't.