What Is the Amazon Inventory Performance Index (IPI) Score?

What the score actually measures

The Inventory Performance Index is a single number, 0 to 1000, that Amazon assigns to your seller account — not to an individual SKU — based on how well you're managing FBA inventory. You'll find it on the Inventory Dashboard in Seller Central. It rolls up four things Amazon tracks about your stock: how much is sitting in excess, how much is stranded with no live listing, how often your top sellers are actually in stock, and how fast inventory turns once it lands.

Amazon has never published the exact formula or the weighting between those four inputs. That's deliberate — sellers who know the exact math tend to game the inputs rather than fix the underlying inventory problem, which is the opposite of what the score is meant to encourage. What Amazon has confirmed is which levers move it, and that's enough to work with.

The four inputs, in plain terms

Amazon names four contributors on the dashboard itself, though it won't say how they're weighted against each other:

• Excess inventory — units sitting in an FC well past what your sales rate justifies. This is the one most sellers ignore because it doesn't look urgent until the storage bill or aged-inventory surcharge arrives.
• In-stock rate — the percentage of the last 30 days each ASIN was buyable, weighted toward your higher-velocity SKUs. A stockout on your best seller costs more score than one on a SKU that barely moves.
• Stranded inventory — units physically in a fulfillment center with no live, buyable listing attached. Usually a listing suppression, a category gate, or a variation error.
• Sell-through rate — units sold against average units on hand over a trailing window. It's the closest thing to inventory turns Amazon tracks internally.

None of these four move independently. Fix stranded inventory and your excess number often improves too, because the units that were stuck start selling again instead of aging in a bin.

A worked example: how a bad score actually happens

Say a SKU sells 400 units a month and you're carrying 3,000 units of average stock. Annualized, that's 4,800 units sold against 3,000 on hand — an inventory turn of 1.6. Common guidance puts a healthy turn somewhere between 6 and 12 a year, meaning you're carrying roughly four to seven times more stock than your sell-through supports. That's excess inventory, and it's the input sellers most often walk past because the units aren't costing anything dramatic on a per-day basis — the cost shows up later, as an aged-inventory surcharge, and as a lower score that then caps how much of your next PO you're allowed to send in.

Now add a second SKU where the listing got knocked into a suppressed state for three weeks. Those units are sitting in the FC, unsellable, incurring storage fees, dragging the stranded-inventory input down at the same time excess is already dragging on the other side. Two unrelated problems, one combined hit to a single account-level number — which is exactly why the score can drop sharply in a month where nothing about your actual sales performance changed.

The threshold keeps moving — don't trust a number you read on a blog

Amazon has changed the minimum IPI threshold for storage limits more than once — it's been set as low as 350 and as high as 500 in different periods, and restock limits attached to it have moved independently of the score threshold itself. Any specific number you read in an article, including this one, may already be stale by the time you act on it. Check your own Seller Central account for the current requirement rather than trusting a cached figure.

What doesn't change is the mechanism: fall under the threshold and Amazon caps how much inventory you can send into FBA for the SKUs affected, and adds storage limitations that get recalculated on a set cycle. That cycle is quarterly for the storage-limit determination itself, even though the score updates more often — which is the single most common source of "I fixed it and nothing happened" confusion.

When the score is bad news: what to check before you panic

A dropping score usually means one of three things, and they need different fixes. If restock limits just kicked in, go to the SKU level first — Amazon's own reports break out which ASINs are driving excess and which are stranded, and fixing the two or three worst offenders often moves the account number more than a broad cleanup across everything. If you already fixed the inputs and the score hasn't moved, remember the lag: the score recalculates on its own schedule, but the storage-limit decision tied to it is set quarterly, so a fix made mid-quarter may not show up as relief until the next cycle starts.

If you think the number itself is wrong — inventory Amazon shows as excess that you know already sold, or stranded units tied to a listing you fixed weeks ago — there's no formal appeal process for the score itself. What you can do is audit the underlying reports (Stranded Inventory, Excess Inventory, Inventory Age) against your own sales data and file a case through Seller Support if the discrepancy traces to a specific data error, like a shipment Amazon received but never checked into inventory correctly.

The mistake sellers make, including one we've made

The most common mistake is treating the score as the problem instead of the symptom. Teams create removal orders reflexively to shed excess inventory and pull the number up, without checking that the SKU being removed can't be re-sent to FBA for months afterward under Amazon's own removal restrictions — trading a temporary score dip for a longer-term stockout on a product that was about to turn.

We've made a version of this mistake ourselves: reacting to a single bad week in the score before checking whether it was one SKU with a data error rather than a real inventory problem. Across the accounts Full Circle has managed — more than $500M in managed revenue across 100+ brands — the pattern that actually holds is that the score is a lagging signal. The stockout that hurt your rank mid-campaign, or the reorder that shipped three weeks too late, already happened by the time the IPI number reflects it. Chasing the score without fixing the reorder timing behind it just means you'll be back here next quarter.

If the leak you're actually chasing is in the ad account — wasted spend propping up a listing that's about to go out of stock — that's a job for Dr. PPC, not an inventory fix.