Amazon Restock Planning

Amazon FBA restock planning is the discipline of answering two questions — how much and when — before Amazon’s fee structure answers them for you. Order too late and you stock out, losing the organic rank you paid months of PPC to build. Order too much and you eat storage fees, aged inventory surcharges, and capacity you needed for faster movers. The sellers who get this right do not have better intuition; they have reorder point math, a forecast they actually maintain, and buffers sized to the season. This guide covers the core formulas, sell-through forecasting that survives contact with reality, seasonal and Q4 planning, and the AWD and 3PL workarounds for capacity limits.

Reorder Point Math: The Formula That Runs Everything

The reorder point is the inventory level that triggers a new purchase order:

Reorder point = (average daily sales × total lead time in days) + safety stock

The first term is lead time demand — what you will sell while the replenishment is in transit. Both inputs get fudged constantly, so be strict about them:

Average daily sales should come from a trailing 30–60 day window, adjusted for anomalies. Strip out the spike from last month’s Lightning Deal and the trough from the week you were stocked out — a stockout period in your velocity average is the most common way sellers systematically under-order.

Total lead time is door-to-sellable, not the supplier’s quote. Count all of it: manufacturing (often 30–45 days), quality inspection, freight (14–21 days air, 30–50 days ocean west coast, longer east coast), customs, prep or 3PL cross-dock, and Amazon receiving — which can quietly add 5–15 days in normal months and worse in Q4. A “30-day lead time” product is usually a 75–90 day product measured honestly.

When sellable inventory plus inbound (units already on POs and shipments) drops to the reorder point, you order. The order quantity itself balances your MOQ, cash cycle, and capacity limit — but the trigger is non-negotiable math.

How Much to Order: The Quantity Question

The reorder point says when; quantity is a separate decision with three competing constraints:

Days of cover. The clean starting point is ordering enough to last one full reorder cycle plus lead time — for a 60-day lead time product reviewed monthly, roughly 90 days of forecasted demand per PO. Shorter cycles mean less capital tied up but more freight events and more chances for a timing miss.

MOQ and price breaks. Suppliers will happily sell you 12 months of inventory for a 6% unit discount. Price the discount against carrying cost honestly: FBA storage, aged surcharge risk past 181 days, capacity consumption, and the cash locked up. For most products, a price break that requires holding more than 5–6 months of supply is a loss dressed as a deal.

Cash conversion cycle. You pay the supplier at order or shipment; Amazon pays you two weeks after the sale. A 90-day cover order on a 60-day lead time means fronting roughly five months of product cost before the last unit converts back to cash. Growing brands stock out most often not from bad forecasting but from POs shrunk to fit the bank balance — if that is the real constraint, fix it with payment terms, inventory financing, or a smaller catalog, not with silent under-ordering on your best sellers.

Safety Stock: Sizing the Buffer Instead of Guessing It

Safety stock exists because both demand and lead time vary. The formula that requires no statistics degree:

Safety stock = (max daily sales × max lead time) − (average daily sales × average lead time)

Example: a product averages 20 units/day with a 60-day honest lead time, but has hit 30 units/day in good weeks and lead time has stretched to 75 days before. Safety stock = (30 × 75) − (20 × 60) = 2,250 − 1,200 = 1,050 units, and the reorder point becomes 1,200 + 1,050 = 2,250.

If that number makes you flinch, tune it by product tier rather than cutting it blindly. A-items with page-one rank get the full buffer — their stockouts cost 2–3x the lost sales in rank recovery, a math we work through in our Amazon inventory management guide. C-items can run lean; a two-week gap on a long-tail SKU costs little. The statistical version (Z-score × demand standard deviation × √lead time) is worth adopting once a tool is doing the arithmetic, because it lets you set an explicit service level — 95% for A-items, 85% for the tail — instead of one-size buffers.

Also size buffers in weeks of cover, not just units, when reviewing: safety stock of 1,050 units at 20/day is about 7.5 weeks — sensible for an ocean-freight product, absurd for one you can air-ship in 10 days.

Sell-Through Forecasting That Survives Reality

Every formula above leans on a demand forecast, and flat trailing averages are where restock plans die. Build the forecast in three layers:

1. Base velocity: trailing 30/60/90-day averages, cleaned of stockouts and deal spikes. If the 30-day is meaningfully above the 90-day, the product is trending — weight recent data, don’t average the trend away.
2. Known events: Prime Day, planned Lightning Deals, price changes, a PPC push, new competitor on your main keyword. Each gets an explicit multiplier on the affected weeks rather than a vague mental note.
3. Seasonality: last year’s monthly index for the ASIN (or its category, for younger products). A product that does 2.4x December vs. its annual average needs that index applied to lead-time demand, not just to the December order.

Then close the loop: once a month, compare forecast to actual and adjust the multipliers. A forecast nobody reconciles is a spreadsheet, not a plan. This layered review is exactly what tools like SoStocked operationalize with per-SKU seasonality curves and exception alerts, but the tool is only as good as the event calendar you feed it.

Seasonal Buffers and the Q4 Plan

Q4 breaks naive restock math in three places at once: demand multiplies, Amazon’s receiving times stretch, and capacity limits tighten. Plan it as its own project, working backward from the demand peaks:

• Early September: Q4 POs leave suppliers. Ocean freight booked. This means Q4 quantities were decided in July–August using last year’s seasonal indices — if you are starting your Q4 plan in October, you are choosing between air freight margins and stockouts.
• Early-to-mid November: Black Friday/Cyber Monday inventory checked in at Amazon. Amazon publishes hard inbound cutoff dates each year (typically late October to mid November for BFCM); treat the published date minus 14 days as your real deadline, because Q4 check-in queues are the worst of the year.
• Late November: December inventory already in the network — there is no restocking between BFCM and Christmas, only what is already in the building.

Size Q4 buffers asymmetrically. Under-stocking BFCM on an A-item costs rank during the highest-velocity weeks of the year; over-stocking costs January storage fees and maybe a removal order. For proven A-items, err heavy. For unproven products, err light — January is a brutal month to discover you own 4,000 units of a C-item at $2.40 per cubic foot storage. And if a deal or seasonal spike is part of the plan, coordinate ad budgets with stock depth; a Prime Day or holiday PPC push into thin inventory just buys your stockout sooner.

Capacity Limit Workarounds: AWD and the 3PL Buffer

Your reorder math can be perfect and still collide with an FBA capacity limit that will not hold the plan — especially in Q4, when limits tighten as everyone inbounds at once. Two structural fixes:

Amazon Warehousing and Distribution (AWD). Amazon’s upstream bulk storage: send pallets to AWD, and it auto-replenishes your FBA stock as it sells down. The decisive features are that AWD inventory does not count against your FBA capacity limit, storage runs far cheaper than FBA rates, and there are no aged inventory surcharges upstream. For import-heavy sellers, shipping ocean containers straight into AWD and letting it drip-feed FBA solves capacity, storage cost, and restock cadence in one move. Watch replenishment lag — AWD-to-FBA transfers take days, so hold roughly 3–4 weeks of cover in FBA proper.

The 3PL buffer. The same architecture with more control: bulk inventory at a 3PL, with weekly or biweekly forward shipments into FBA sized to your sell-through. Costs more labor than AWD but gives you multichannel stock, FBM backup capability during FBA stockouts, and freedom from Amazon’s transfer timelines. Many of the brands we manage run 30–45 days of cover in FBA, the balance at a 3PL, and an FBM offer on every A-item as the last line of defense.

Either way, the principle is the same: FBA is your shelf, not your warehouse. Keep the shelf stocked and the depth upstream.

Amazon FBA Restock Planning Is a Weekly Habit, Not a Spreadsheet

Everything above decays without a cadence: weekly restock-alert reviews, monthly reorder-point and forecast recalculation, and immediate updates when velocity, lead times, or capacity limits move. It connects to everything else in Amazon account management — PPC pacing against stock depth, deal calendars, IPI protection, cash planning — which is why restock planning fails most often not from bad math but from nobody owning it week over week. That ownership is a core deliverable of our Amazon account management service: we run the forecasts, place the restock triggers, manage AWD and 3PL buffer layers, and coordinate it all with your ad calendar so stockouts and surcharges stop deciding your P&L. If your current restock process is a spreadsheet someone updates when they remember, that is the gap to close first.