FBA Replenishment: Work Backwards From "How Much Ships Today"

In September 2021, a seller friend in pet supplies called me, voice shaking: Black Friday was 60 days out, his bestseller had 12 days of cover left at FBA, and ocean freight was already off the table. He air-freighted three emergency batches — $40,000-plus in freight, the year's profit gone. What had he been doing in August? The line I've heard a hundred times in twenty years: "Sales felt fine, so I never ran the numbers."

Nine out of ten inventory disasters start with "sales felt fine."

Bottom line up front: replenishment is arithmetic in reverse. How much you ship today = how many days of sales you want to cover − what's in the warehouse − what's on the water. Get the "how many days" right and the rest mostly takes care of itself.

Step one: know your real daily sales

Daily sales are the foundation. Many sellers use yesterday's units as the baseline — a dangerous shortcut, since yesterday might have been a heavy ad day or just a weekend.

Look at three layers instead:

Flat baseline: trailing 30-day daily average, promo spikes stripped out. A storage bin does 4,500 units in 30 days including one 600-unit lightning deal day — the baseline is (4,500−600)÷29 ≈ 134 units a day, not 150.

Seasonal multiplier: the line item rookies forget. The same product can do 2–3x its flat rate during Black Friday week, 1.5–2x through December, 3–5x on Prime Day. Plan on flat sales and you will stock out in Q4. Split the next 90 days into "flat stretches" and "peak stretches" on a sales calendar, with a different daily rate for each.

New products with no history? Find a comparable older product — same category, same price band — and take 70% of its rate as the conservative estimate. Under-buy at launch and re-fire small air batches. Dead launch inventory hurts worse than a launch stockout.

The core formula: days of cover × daily sales, minus what you have

The whole thing fits in one line:

This shipment = daily sales × target days of cover − on-hand inventory − in-transit inventory

Target days of cover = inbound transit days + FBA receiving days + safety buffer days

A worked example. A home product doing 200 units a day, ocean to the West Coast: 35 days inbound, 10 days FBA appointment plus receiving, 10 days safety buffer — 55 days of cover, 11,000 units of demand. FBA holds 1,500, with 3,000 on the water: ship 11,000 − 1,500 − 3,000 = 6,500 units this round.

The discipline hidden in this formula: it forces on-hand and in-transit onto the same ledger. Most stockouts I see aren't "we didn't ship" — they're "we shipped three batches and they're all floating in the Pacific while the FBA shelf sits empty."

Inbound lead time: don't trust the sailing schedule alone

The carrier quotes "14 days to LA." That's port to port. From order placement to sellable units, the real chain is: factory production 7–15 days, drayage and export clearance 3–5, ocean 25–35 days (West Coast) or 35–45 (East Coast), destination customs and pickup 3–7, domestic trucking 3–5. End to end, 40–60 days West Coast and 50–70 East Coast is normal.

Add 7–10 days of congestion buffer in peak season. August through October, the LA/Long Beach queue is a standing tradition — in 2021 some vessels sat at anchor 20-plus days. Plan on "normal plus peak buffer," never your fastest run ever.

Air is 7–10 days door to door; express 3–5. Air is a fire extinguisher, not a strategy: per unit it typically runs 5–8x ocean. Sellers who skip the math when things are calm burn their margin on emergency air when the shelf goes empty. Keep that ratio in mind next time you're tempted to "figure it out later."

The FBA side: appointments and receiving eat days too

Arriving in the US is not the same as being sellable. FBA runs on appointments, and in peak season getting a delivery slot — then waiting in the truck queue — is routine. From "freight at the local warehouse" to "available for sale," plan on 7–10 days off-peak and 12–20 in peak (September–December): appointment wait, truck queue, and Amazon receiving each eat several days. Ahead of Black Friday 2021, one seller's early-November delivery didn't finish receiving until mid-December — units sat in the building while the listing showed out of stock.

So for the "FBA receiving days" in your cover calculation: 10 days off-peak, 15–20 in peak. You won't be far off.

Safety stock: leave room for the spikes, don't run naked

Demand isn't a straight line, and stockouts happen on the spikiest days. A simplified version does the job:

Safety stock (in days) = inbound transit days × demand volatility

Volatility is how far peak daily sales overshoot the average over the last 90 days. Average 200 a day, promo days hit 300 — 50% volatility. With 35 inbound days: 35 × 50% ≈ 17 days of cover, or 3,400 units.

How conservative to be depends on margin. Above 40% margin, pad an extra 5 days without guilt — one stockout day costs more in lost profit than two weeks of extra carrying cost. Under 15% margin, flip it: warehousing fees on idle units will eat you alive.

Splitting across multiple FBA warehouses

Amazon will scatter your inventory across West Coast, East Coast, sometimes central warehouses. The split logic is simple: mirror your regional sales mix.

Pull the geographic sales report. If California, Texas, and Florida drove 60% of sales over the last 90 days, split West/East shipments roughly 60/40. No data on the first shipment? Start with 40% West, 40% East, 20% central, and adjust after a month of real data.

One more practical note: smaller, more frequent batches beat one giant push. The same 10,000 units split into two shipments two weeks apart lets you correct quantities and splits on round two based on round-one sales. Split shipments cost a bit more per unit on ocean — worth it for new and seasonal products.

IPI and storage caps: having the goods doesn't mean you can send them

The newer headache: your inventory is ready, and Amazon won't let it in — you've hit your storage cap. Amazon rations capacity through the Inventory Performance Index (IPI), and low scores get limits cut ahead of peak. The index weighs excess inventory share, sell-through rate, stranded inventory, and in-stock rate, but the exact thresholds change constantly — always check the current policy in Seller Central rather than trusting an old screenshot.

Three moves: clear dead weight — anything untouched for 90 days gets liquidated or removed instead of squatting on capacity. Nurse sell-through 2–3 months before peak with promos on slow movers. When capacity gets cut, prioritize bestsellers and seasonal SKUs, shifting long-tail products to a third-party warehouse for merchant-fulfilled orders. Capacity is scarce; spend it where it pays.

Stockout vs. overstock: which one actually hurts more

I get asked this constantly. My answer: it depends on the product's stage.

For new and growing products, stockouts hurt more. One stockout tanks listing rank, ad weight, keyword positions — and competitors move in. One seller measured it: two weeks out of stock on a hero SKU, and the recovery ad spend equaled three months of extra safety-stock carrying cost. In this phase, an extra 10 days of cover is cheap insurance.

For mature and declining products, overstock hurts more. Every extra day on the shelf is warehousing fees, and long-term storage surcharges get ugly past the one-year mark. Compress cover under 30 days, top up with small air batches, sell through to zero.

Short version: when the product is climbing, extra inventory buys insurance; when it's sliding, lean inventory stops the bleeding.

The pre-shipment checklist

Before every shipment, run through this. Ten minutes, and it dodges most disasters:

• Daily baseline refreshed? Trailing 30 days, promo outliers stripped, seasonal multiplier applied
• Target cover days = inbound + FBA receiving + safety buffer — what are the three numbers
• On-hand and in-transit reconciled? Where is each in-transit batch right now
• Warehouse splits updated against the latest regional sales data
• Carton and FBA shipment labels correct? Mislabeled cartons get refused at the dock
• IPI score and remaining capacity checked? Will this shipment fit
• Does arrival land ahead of the promo calendar (Prime Day, Black Friday, Christmas)

Done right, replenishment comes down to discipline, not math. The hard part is updating the data and reconciling the ledger every month without fail. Every seller I've seen do this well shares one trait: a spreadsheet, updated monthly, no skipped months.

When did you last run your numbers properly? If it's been more than a month, open the spreadsheet today and start with the daily baseline.