Since 2025 your restock cap is set per SKU, which means your ad spend ceiling is too.
A stockout on your own site is a gap in the calendar. On Amazon it resets the listing.
Ranking slides, the Buy Box goes to whoever is in stock, Subscribe and Save subscribers get handed a competitor, and the campaigns funding that ASIN either keep spending against a listing nobody can buy or get paused and give up their placement history.
None of it reverses when the stock lands. You rebuild it.
And something changed in 2025 that most inventory advice hasn’t caught up with. Amazon cut storage allowance from six months of forecasted sales to five, and brought back restock limits at the individual SKU level. Account capacity and SKU capacity are now separate constraints, so you can hold plenty of one and none of the other, on the exact ASIN your campaigns are funding.
At a Glance
Table of Contents
Five things to take away
- A stockout resets organic velocity. The lost sales are the smaller cost.
- Losing a Subscribe and Save subscriber costs the rest of their lifetime value, not one order.
- Since 2025 restock caps run at ASIN level, so account capacity tells you nothing about the SKU you’re advertising.
- Storage allowance dropped from six months of forecasted sales to five in mid-2025.
- An IPI score below 400 triggers storage limits. IPI is Amazon’s inventory performance score, and it’s the result of four inputs you already control.
Quick Answer:
Amazon inventory management is the planning, replenishment and day-to-day control of stock across FBA, where Amazon ships for you, and FBM, where you ship yourself. A stockout on a top ASIN costs more than the lost sale. It resets organic rank, hands over the Buy Box, and forces you to rebuild velocity from zero. Since 2025, restock limits work at ASIN level, not account level, so the budget you can defend behind a SKU is capped by how much of it you can resupply.
Definition
Three constraints get treated as one thing, and they’re separate.
Account storage capacity is the total volume you can hold in FBA, now set at roughly five months of forecasted sales, down from six in mid-2025.
ASIN-level restock limits cap how many units of a single SKU you can send in. Amazon brought them back in 2025. You can have account capacity available and still be blocked on one product.
IPI, the Inventory Performance Index, is a score from 0 to 1,000 that decides whether storage limits apply at all. Below 400, they do.
The first two are quantities and the third is a gate. Mix them up and you end up with room in the warehouse and nothing to sell.
1. What a stockout costs after the stock returns
The lost sales are the visible cost, and the smaller one.
Ranking doesn’t pause, it decays. A single day out of stock can drop a top product’s rank noticeably, and three or more days compounds it sharply. Amazon reads a fulfillment failure as a sign the listing matters less, and the position you built is the position a competitor now holds.
The Buy Box goes immediately. Nearly all marketplace sales go through it, so losing it on a reseller listing removes the path to purchase entirely.
The advertising keeps running or stops, and both are expensive. Left running, it spends against a listing nobody can buy. Paused, the campaign loses its placement history and has to re-earn it at a higher cost per click when stock returns.
So the real cost is the rank you paid to build and now have to buy back. That’s a different number from the missed orders, it lands in a different report, and nobody connects the two.
Getting ahead of it means forecasting before the stockout, not reacting after it. Predictive modeling for demand forecasting is where that starts, and it sits alongside pricing, because the hidden costs of stagnant pricing feed the same profitability question.
What to do
- Measure the recovery window, not the outage. How long rank took to come back is the number that sizes the cost.
- Pull the ad data for the outage and the month after. Look at cost per click and conversion across the whole account, because the effect spreads beyond the affected ASIN.
- Decide the pause rule before you need it. Whether campaigns stop or keep running during an outage should be policy, not a call made under pressure.
What to avoid
- Treating restock as recovery. The stock coming back is where the rebuild starts.
- Reading the outage cost as lost revenue only. The advertising cost outlives it.
2. Why Subscribe and Save SKUs cost more
A stockout on a Subscribe and Save hero SKU stops costing you revenue and starts destroying recurring lifetime value.
Amazon doesn’t just pause the order. After repeated stockouts it cancels the subscription outright and suggests a replacement. That replacement is a competitor.
Churn is a ranking signal. A spike tells Amazon the listing is unreliable, which weakens visibility beyond the immediate cancellation and turns the stockout into an ongoing penalty, not a one-time hit.
The financial gap is every order they haven’t placed yet. Losing a subscriber costs the rest of that stream, handed to whoever Amazon suggests next.
The earned placements go too. Subscribe and Save badges, the frequently repurchased carousel and brand recommendations are all earned through a record of staying in stock with low churn. A stockout breaks the signal that earned them, and winning them back takes longer than losing them.
What to do
- Rank your Subscribe and Save ASINs by active subscriber count, then protect them in that order.
- Set the safety stock on subscription SKUs above everything else. The downside is lopsided.
What to avoid
- Treating a subscription SKU like any other in the allocation queue. Nothing else in the catalog carries a cancellation cost.
3. Segmenting and protecting what earns
A profit-first inventory strategy doesn’t mean keeping everything in stock. It means protecting what earns.
Heroes against long tail
A small share of ASINs generates most revenue. Losing a long-tail SKU for a week is an annoyance, while a day without a hero SKU can wipe out months of growth.
That split is the foundation for every other decision here. Revenue alone is the wrong way to sort, though. A deep read of SKU economics on Amazon shows which products earn hero status on net profit, not topline.
Sorting by contribution margin assumes the catalog record tells you which item is which. On a wide assortment it often does not. Orphaned children, variations nobody has audited in a year and parents holding items that stopped selling for reasons nobody wrote down all distort the ranking before you apply it. Amazon catalog management straightens the record so the segmentation means something.
Amazon first, DTC as the shock absorber
Hero SKUs on Amazon should never stock out. The carrying cost of extra safety stock is small next to the cost of an outage.
When inventory tightens, your own site takes the squeeze, because you can take backorders and build waitlists there. On Amazon, the algorithm hands your customers to someone else.
Rules, written down
Set safety stock higher on heroes than the standard your other channels use, and set it in days of cover, not units.
Trigger reorders earlier on heroes, with the buffer sized for supply chain variance and FBA check-in delay, not the average case.
Write the allocation rule before you need it. When inventory is constrained, hero ASINs get priority, and a rule decided in advance survives the week it’s tested.
What to do
- Sort by contribution margin, not revenue. A high-revenue SKU with thin margin isn’t a hero.
What to avoid
-
- Setting one safety stock policy across the catalog. It overspends on the tail and underprotects the top.
- Deciding allocation during the shortage. That’s when the loudest SKU wins, not the most valuable one.
|
+1,119%
Profit Growth
Levtex Home
|
+586%
Profit Growth
Karat
|
+833%
Revenue Growth
Crazy Dog
|
4. IPI and the 2025 FBA constraints
What IPI is, and what it isn’t
IPI runs from 0 to 1,000, and the threshold is 400. Below it, storage volume gets capped and inbound shipments get restricted.
IPI is an output, not a lever. It’s calculated from four inputs you already control, and the first two carry the most weight.
Excess inventory percentage. Stock sitting beyond what demand justifies.
Sell-through rate. How fast units turn into cash.
Stranded inventory percentage. Units in FBA that aren’t attached to a buyable listing.
In-stock rate. Whether your popular ASINs stay available.
You can’t manage IPI directly. You manage those four and IPI follows.
The 2025 changes, and why they matter more than the score
Storage allowance fell from six months of forecasted sales to five in mid-2025.
ASIN-level restock limits came back. They used to work at account level. Now a single SKU can be capped while the account still has room.
That second change is the one to plan around. Total capacity tells you nothing about whether the product your campaigns are funding can be resupplied. So the most expensive mistake in the account is a specific one, funding a campaign for a SKU you can’t restock.
Every dollar of that spend builds rank on a listing that’s about to go unbuyable, in a position a competitor takes while you wait for capacity.
The recovery order
Stranded inventory first. Usually the fastest win, and often fixed the same week.
Excess inventory second. Removal orders or steep discounts on anything aged past 90 days.
Sell-through third. Advertising, pricing and replenishment working together on the products that turn.
Recovery from a low score takes weeks, not days, and Amazon runs a formal evaluation about six weeks before each quarter ends. That evaluation sets the next quarter’s limits, so the work has to start well ahead of the quarter you care about.
Fees tie into all of it. Amazon’s low-inventory fee and related fee changes hit margin and IPI at the same time.
What to do
- Check the score weekly and the constraint monthly. They move on different clocks.
- Read restock capacity at ASIN level before you set campaign budgets. Since 2025 that’s the binding number.
- Hold a buffer above 500, not 400. Amazon can adjust the threshold quarterly.
What to avoid
- Planning next quarter’s budget against the capacity you have now. The quarterly evaluation may cut it.
Decision grid
Each inventory signal, and the advertising decision it should trigger.
If nobody owns row two, the account keeps funding campaigns for products that can’t ship.
How Adverio Helps
Adverio is an agency that runs Amazon, Walmart and Target for consumer brands, measured on contribution margin against a baseline set before the work starts.
On inventory, that means reading restock capacity at item level before the media plan is set, not discovering the limit after the budget’s committed.
Inventory sits inside a broader Amazon account management system, so subscription risk and IPI exposure get read alongside the ad budget, not after it.
Frequently asked questions
What IPI score do you need to avoid storage limits?
An IPI of 400 or above. Below that, Amazon caps storage volume and restricts inbound shipments. Amazon can adjust the threshold quarterly based on fulfillment network capacity, so a buffer above 500 protects you against a change you didn’t see coming.
How long does it take to recover a low IPI score?
Weeks, not days, and the timing matters more than the duration. Amazon runs a formal evaluation roughly six weeks before each quarter ends, and that evaluation sets the next quarter’s limits.
Start recovery inside the evaluation window and you carry the restriction through the next quarter, however fast the score moves.
Can you have storage capacity and still be unable to restock a product?
Yes, and since 2025 it’s common. Restock limits came back at ASIN level, so a single SKU can be capped while the account has plenty of room.
Check the SKU, not the account, before you commit budget to a campaign that depends on it.
Should campaigns pause during a stockout?
Both options cost something, which is why it should be a policy, not a decision made under pressure. Running spends against a listing nobody can buy. Pausing gives up placement history and raises the cost of getting back in.
The shorter the expected outage, the stronger the case for holding position. The longer it runs, the harder it is to justify spending through it.
Closing
Before the next budget meeting, pull restock limits for your top 20 ASINs and cap each campaign to what that SKU can resupply.
References
Platform rules move. Figures were checked on 4 September 2026 and should be confirmed before you budget.
- Amazon Seller Central, Inventory Performance Index and FBA capacity. https://sellercentral.amazon.com/
- Amazon Seller Central, FBA restock limits and Capacity Manager. https://sellercentral.amazon.com/
- Amazon Seller Central, low-inventory-level fee. https://sellercentral.amazon.com/



