How much inventory should I bring for a product launch?
Enough to cover the launch period plus your lead time at a conservative demand estimate, with the option to release more from a holding location if the SKU takes off. The goal is to stay in stock through the launch without committing your whole first production run to FBA before you know the product sells.
Should I send my whole first order into FBA at launch?
Usually not. Send a launch quantity into FBA and hold the rest at a 3PL, AWD, or your own warehouse. That keeps FBA storage costs and stranded-inventory risk down on an unproven SKU while letting you replenish quickly if demand confirms. If the product flops, the held units never paid FBA fees and can be redirected or liquidated more cheaply.
T. Brian Jones is co-founder and CTO of Inventory Hero. He leads the engineering behind its Amazon data pipeline, demand forecasting, and the AI platform that lets sellers talk to their live inventory, sales, and supplier data in plain language.
How do I protect against a launch that does not sell?
Order conservatively, stage the inventory so most of it is not in FBA, and watch early sales daily so you can stop or slow the next order. The asymmetry favors caution: a brief stockout on a new SKU is recoverable, while a warehouse of a product nobody wants is trapped cash and ongoing fees.
Read article
For an Amazon product launch, bring enough inventory to survive the launch period plus your first reorder at a conservative demand estimate, stage it so most of it is not sitting in FBA yet, and plan the first reorder before you go live. The short version: a launch is a bet on an unproven SKU, so you size for survival, not for a hopeful peak, and you keep your options open by holding reserve stock outside FBA. Below is how much to bring, how to stage it, and how to limit the downside.
The launch quantity question is really a forecasting question, covered in new product demand forecasting: build a conservative estimate from analogs and benchmarks, then cover the launch window plus your lead time at that rate. What is specific to inventory is the asymmetry of the bet. A brief stockout on a new SKU with little rank to lose is cheap and recoverable. A warehouse full of a product that flops is months of fees and trapped cash.1 So you bias low and keep the ability to reorder fast.
Even when your first production order is large, you do not have to send all of it into FBA at once. Split it:
A launch quantity into FBA. Enough to cover the launch period and stay in stock through the first reorder cycle, with safety stock for the launch spike.
The rest held outside FBA. Keep the balance at a 3PL, in AWD, or in your own warehouse, ready to replenish quickly.
Staging does three things on an unproven SKU. It keeps your FBA storage cost and stranded-inventory risk down while the product is still proving out. It avoids tying up a tight restock limit with inventory that may not sell. And if the launch flops, the held units never paid FBA fees and can be redirected, bundled, or liquidated far more cheaply than units stuck in a fulfillment center.
The portable rule: send roughly six to eight weeks of conservative velocity into FBA, and hold the rest. That covers the launch and the first replenishment cycle without committing the whole run to fulfillment-center fees.
You produce 2,400 units for launch. Your conservative forecast is 12 units a day, and replenishment from a domestic 3PL into FBA takes about 14 days:
Allocation
Units
Rationale
Into FBA at launch
~600
About 7 weeks of cover at 12/day, tighter if the launch spikes
Held at 3PL
~1,800
Released as demand confirms, ~14-day replenish
At 600 units (50 days at the conservative 12/day rate) you cover the launch and bridge to the first 3PL replenishment. If the SKU sells at 20 a day you push more in quickly; if it sells at 4 a day, only ~600 units are exposed to FBA fees instead of the full 2,400.
The held units are not all equal. Amazon Warehousing and Distribution (AWD) replenishes into FBA fast, often a few days, so if you are on AWD you can safely hold a larger share outside FBA and trickle it in. A standard domestic 3PL replenishes more slowly (commonly two weeks or more), so you stage a bit more into FBA up front to cover the longer refill, but a 3PL is often easier to route liquidation through if the launch flops, since you skip the AWD-to-FBA transfer. Match the split to your replenishment speed.
Launch demand is inflated. Ads, launch coupons, and the ranking honeymoon push early sales above the rate the SKU will hold once the promotions stop and rank settles. If you reorder to the launch spike, you overbuy for a peak that is not real. A practical haircut is 30 to 50 percent off launch-week velocity if you ran aggressive PPC and a launch coupon. Discount the early sales velocity toward that sustainable rate, using only the in-stock days, before you size the next order.
Because your lead time is long, you will commit to the next order while demand is still proving out. From a staging view, that is when you decide how fast to pull the held 3PL units into FBA, and you can size that tranche in the FBA restock calculator. The full treatment of how to read the early signal and time that reorder lives in new product demand forecasting; for launch staging, the point is simply to keep the reserve outside FBA until that first real read confirms the SKU is working.
Launch inventory is a staged bet: bring a conservative quantity, send only a launch portion into FBA, hold the rest outside it, discount the launch spike, and plan the first reorder in advance. That keeps you in stock through the launch while capping the downside if the product does not land. For the wider system, see Amazon inventory forecasting and restock planning.