If you sell on multiple platforms and you have not locked your peak-season inventory split 90 days out, you are not planning anymore — you are about to improvise. Every October I get the same call: "FBA appointments are booked solid, can you emergency-transfer from your 3PL?" Usually yes. And the expedited transfer will cost enough to hurt for months.
Peak-season stocking is not a courage contest. It is a calendar. Here is how I would run the 90 days.
Work backward from the promo date
Assume the big promos start in late November. Count back 90 days to late August, and the schedule writes itself.
Weeks 1–2 (late August into early September): commit to the plan. Pull last year's same-period sales for every platform — Amazon, Walmart, TikTok Shop, your own site — one by one. Do not guess the growth rate. Base it on three things: this year's month-over-month trend, whether you are launching new SKUs before the holidays, and whether you have off-site traffic plans.
Weeks 3–6 (September): place orders and move freight. The deadline for factory purchase orders is mid-September at the latest. Miss it and the ocean sailing window for a late-October inbound slips away. A China-to-US-West-Coast ocean move, including customs clearance and drayage, routinely runs four to six weeks — longer right before peak. Every week the factory slips compresses every downstream step — expedited drayage, expedited receiving, none of it cheap.
Weeks 7–10 (October): split and inbound. This is the month that decides everything. FBA delivery appointments two to three weeks out are routinely fully booked before peak; miss your slot and you wait in line or pay up for premium truckload delivery. Walmart's WFS inbound needs advance appointments too. TikTok Shop is more flexible, but if you go viral and cannot replenish fast enough, the virality is wasted.
Weeks 11–13 (early November through the promos): hold the line. Inventory is in place. Now the job is watching stock levels, watching stockout alerts, and having emergency transfers ready. Check sellable inventory and inbound pipeline on every platform daily. If any SKU's coverage will not last through the promos, start moving units a week early.
The table looks simple. But I have watched too many sellers drag through the first two steps and discover in October that the factory has not shipped. At that point no strategy helps. The only options left are "pay more" and "accept it."
A back-of-the-envelope way to size the buy
How much to buy? Here is a rough-but-honest method, in three steps.
Step one: daily average per platform. Take this SKU's daily average sales on that platform during last year's promo window — say November 15 to December 15 — and multiply by this year's growth factor. How do you set the growth factor? Compare this SKU's average monthly sales on that platform from January through August this year against the same period last year. That number is real, not imagined.
Step two: add a cushion. Promo-period demand swings far more than everyday demand. My habit is to add two to three weeks of normal daily sales as safety stock. Why two to three weeks? Because once you stock out, the replenishment cycle into an FBA warehouse during peak starts at two to three weeks.
Step three: platform coefficients. The same SKU behaves completely differently on Amazon versus TikTok Shop. TikTok Shop demand is concentrated and spiky, with higher return rates — size the buy for the spike, but do not bet the whole business on it. Amazon is the base: the largest volume, the most stable. Walmart's WFS has been growing nicely, but its traffic mix differs from Amazon's, so copying your Amazon buy quantity over tends to produce overstock.
A worked example to show the logic (illustrative numbers, not a forecast): a SKU did 100 units a day on Amazon during last year's promos, and January–August sales are up 30% year over year. For a 30-day promo window: 100 × 1.3 × 30 = 3,900 units, plus a two-week cushion of 100 × 1.3 × 14 = 1,820, for roughly 5,700 units total. Crude, but every step has a basis — far more reliable than "I feel like we can sell ten thousand."
One more line item: fold in the return rate. For high-return categories like apparel and beauty, reserve a share of the buy for return-driven restocking, or you will hit the classic late-promo trap of "the system shows inventory, but nothing is actually sellable." Use your own store's historical return data. Industry averages will mislead you.
Stockout or overstock: pick your poison
This is the ultimate peak-season question, and my answer may surprise you: both hurt, but they hurt differently, and the right choice depends on the SKU.
Price the stockout first. A stockout costs three things: the promo-period sales you never make, the listing's ranking weight, and wasted ad spend. During the promos you are paying for ads; with no stock, the spend burns and converts at zero. Worse is the ranking damage — peak season concentrates the year's traffic, and stocking out then can knock a listing down for months afterward. One day of stockout routinely costs more than a month of storage on overbought units.
Now price the overstock. Overstock costs two things: tied-up cash and storage fees. Units aging in FBA warehouses attract long-term storage fees on an escalating scale — check Amazon's current published fee schedule for exact rates. Third-party warehouses charge per pallet or per cubic foot, which is cheaper, but the cash is still locked up.
So the tradeoff logic runs like this: for traffic drivers, hero SKUs, and promo headliners, overbuy. The cost of overstock is smaller than the cost of stocking out. For long-tail SKUs and unproven new products, underbuy — if they run out, they run out; replenishment can wait, while overstock is a real loss. Split your catalog into these two buckets and run completely different buy strategies for each. The classic mistake is one standard for every SKU — the fantasy of "no stockouts and no overstock" — which usually delivers both.
In the 2022 peak season I watched a client learn this the hard way. His hero SKU was bought "conservatively" and stocked out on day three of the promos. Emergency transfers from a Los Angeles 3PL into FBA — expedited freight plus appointment jumping — cost several times the normal inbound, and he missed the strongest week of the season. Meanwhile three long-tail SKUs were each bought at three months of cover; most of it was still sitting there after the holidays and had to be cleared at markdown. Hit on both ends, because he never segmented — one "feeling" for everything.
Where the freight actually goes
Getting goods to the US is only the first half. The second half is the split. I have seen too many sellers stack everything in one warehouse and then cry over cross-country transfers during the promos.
The basic rule: position by sales region, not by platform. West, Central, East — keep a share in each. FBA will distribute your units across its own network, but your 3PL safety stock should be regionalized too. Decide in advance which warehouse ships TikTok Shop orders and which ships Walmart orders. Do not figure it out mid-viral-moment.
Line up emergency transfer capacity before you need it. Find a transit warehouse within a two-hour drive of your main facility, agree on peak-season emergency transfer pricing and lead times, and write it down. Hunting for trucks and negotiating rates mid-promo puts you over a barrel. Every September I make clients put the contingency on paper: which warehouse, what price, how fast. Best case, you never use it. But you cannot not have it.
One veteran FBA tip: before peak, do not book your entire inbound into a single FBA warehouse in one shot. Split it into two or three waves, a week or two apart. The first wave guarantees you open the promos in stock; later waves flex with actual sell-through and can be rerouted. If a SKU underperforms, the later units can change destination or platform instead of sitting dead in one building. The appointment calendar always favors whoever moves first.
There is no perfect peak-season plan. There is only a plan whose costs you have priced in advance. Between stockout and overstock, know which one you chose — the worst outcome is choosing neither and eating both.
You can start right now: open a spreadsheet, run your top five SKUs through the rough method above, and compare the result against your current buy plan. Whichever SKU diverges the most is your biggest risk this peak season.
