Sellers tend to treat the Inventory Performance Index as a compliance metric — a number Amazon wants you to keep above a line, for reasons that are Amazon’s business rather than yours. So it gets checked when a warning email arrives and ignored the rest of the quarter.
That framing is expensive. A low IPI score does not cost you a penalty fee. It costs you three separate things at once, and the third is the one that does real damage: it caps how much inventory you can hold, which caps how much you can sell, which drives down the sell-through rate that determines your IPI. It is a self-reinforcing loop, and it takes weeks to climb out of.
This article puts numbers on the cost.
What IPI is
The Inventory Performance Index is a score from 0 to 1,000 that Amazon assigns to your FBA account, updated weekly on a rolling 90-day window. It measures how efficiently you convert stored inventory into sales.
Amazon does not publish the formula, but the four components are visible in Seller Central:
- Excess inventory — units held well beyond forecast demand
- Sell-through rate — units shipped relative to average units held
- Stranded inventory — units in the network with no active listing
- In-stock rate — availability of replenishable ASINs
A note on the threshold
The minimum threshold has sat at 400 for most of the recent period, having moved between 350, 400 and 500 in earlier years. Amazon reserves the right to adjust it quarterly based on fulfillment network capacity and demand, and industry reporting through 2026 has been inconsistent — some sources report 400, some report a drop to 350, and at least one reports a substantially lower figure alongside a changed scoring scale.
The practical conclusion is not to plan against any published number. Read your own Capacity Monitor in Seller Central, and hold a buffer rather than sitting a few points above whatever the current line is.
Two structural changes matter more than the threshold itself. In mid-2025 Amazon reduced storage allowances from six months of forecast sales to five, and reactivated ASIN-level restock limits on individual SKUs. The second is the one sellers underestimate: account-level capacity can look acceptable while an individual SKU is capped, which means your bestseller can be throttled because of problems elsewhere in your catalogue.
The three costs, separated
Cost 1: Storage overage charges
Fall below the threshold and Amazon assigns a capacity limit in cubic feet. Inventory stored above that limit attracts an overage charge on top of standard monthly storage. This is the most visible cost and usually the smallest.
Cost 2: Blocked restocks and stockout losses
This is where the money goes. A restock limit doesn’t reduce your margin — it removes revenue entirely, and it removes it from your best SKUs, because those are the ones you were about to reorder.
Illustrative math on a single SKU:
- Sells 240 units/month at $32, contribution per unit $11
- Restock limit blocks a shipment; the SKU is out of stock for 18 days
Units lost: 240 × (18 ÷ 30) = 144
Contribution lost: 144 × $11 = $1,584
That is one SKU, one stockout. But the loss doesn’t end when stock returns:
- Lost organic rank during the stockout, requiring paid spend to recover
- Lost Buy Box share to competitors who stayed in stock
- A sell-through rate that just got worse, because units shipped fell while average inventory held elsewhere in the catalogue didn’t
Cost 3: The reinforcing loop
Sell-through is units shipped divided by units held. A stockout reduces the numerator. Excess inventory on your slow SKUs keeps the denominator high. So the stockout your restock limit caused pushes your IPI down further, tightening the limit again.
Because the score runs on a rolling 90-day window, poor performance keeps weighing on it long after you start fixing things. Recovery from a low starting point is generally a six-to-ten-week project, not a weekend one.
Putting a quarterly figure on it
A seller with 40 SKUs, $220,000 quarterly revenue and a score that has dropped below threshold:
| Cost | Basis | Quarterly estimate |
|---|---|---|
| Storage overage | Charges on inventory above the assigned limit | $600 – $1,800 |
| Stockout contribution loss | 3 SKUs × ~2 weeks out of stock | $4,000 – $7,000 |
| Rank recovery ad spend | Incremental PPC to restore organic position | $1,500 – $3,000 |
| Forced liquidation of excess | Clearing aged stock below cost to raise the score | $2,000 – $6,000 |
| Total | $8,100 – $17,800 |
The storage overage — the only line that looks like a penalty — is under 10% of the total. Everything else is opportunity cost and remediation, which is why treating IPI as a fee problem understates it substantially.
Note the fourth line. Clearing aged inventory at a loss is a common and often correct fix, but it is a real cost, not a free lever. Sellers do report meaningful score movement within a few weeks of clearing stock over 180 days old through deals and coupons — at whatever margin that liquidation gives up.
The diagnostic order that actually works
The four components are not equally fixable, and they are not equally fast. Work in this order:
- Stranded inventory. Fastest fix available and pure upside — these are units you own, sitting in the network, with no listing to sell them through. Resolving a listing error converts dead cubic feet into sellable stock in days.
- Excess inventory on your worst SKUs. Removal, liquidation or aggressive discounting. Painful, effective, and it also stops the aged-inventory surcharges you were already paying.
- In-stock rate on replenishables. Requires reorder discipline rather than a one-off action.
- Sell-through. A trailing outcome of the other three. It cannot be fixed directly, which is why chasing it first wastes weeks.
Best practices
- Check the Capacity Monitor weekly, not quarterly. Amazon’s formal evaluations happen several weeks before quarter end, so the quarter you’re graded on is already partly written by the time you look.
- Hold a buffer above the threshold, not a margin of one point. Given that the threshold can move, sitting at 410 is not safe.
- Set a days-of-cover ceiling per SKU rather than reordering on a fixed schedule. Excess inventory is the component you control most directly.
- Treat aged inventory as a decision with a deadline. The cost of holding a slow SKU rises on a schedule; deciding at 240 days is far more expensive than deciding at 120.
- Audit stranded listings monthly. They accumulate silently through suppressed listings, pricing errors and category changes.
- Model the cost of a stockout per SKU before you need to. Knowing that SKU 14 costs $53/day out of stock changes how you prioritise limited capacity.
- Consider AWD or a 3PL as an upstream buffer for bulk inventory, keeping FBA lean while retaining replenishment ability.
Tools like sellerboard track inventory levels against sales velocity and show net profit per SKU, which is what lets you decide which excess inventory to clear — the useful question is not which SKU has the most units, it is which units are costing more to hold than they will ever contribute.
Common mistakes
- Treating IPI as Amazon’s problem. The score is a proxy for capital efficiency. A low score usually means real money is tied up in stock that isn’t moving.
- Reacting only when the warning email arrives. By then the 90-day window already contains the damage.
- Reordering bestsellers while ignoring the slow tail. The tail is what’s suppressing the score, and it will keep capping the bestsellers.
- Liquidating indiscriminately to raise the score. Clearing a SKU that was going to sell at full margin next quarter is a worse outcome than the overage fee.
- Assuming the published threshold applies to you. It moves, and third-party reporting on it conflicts.
- Forgetting ASIN-level caps. Account-level capacity looking fine does not mean your individual bestseller can be restocked.
- Not counting rank recovery. The stockout ends; the ad spend to rebuild position does not end with it.
FAQ
What is the minimum IPI score? It has most commonly sat at 400, but Amazon adjusts it and third-party reporting through 2026 is inconsistent. Check your Capacity Monitor in Seller Central for the figure that applies to your account.
How often does the score update? Weekly, on a rolling 90-day window. Formal quarterly evaluations determine capacity for the following quarter.
Does a low IPI cause a penalty fee? Not a penalty as such. It triggers capacity limits, and inventory above those limits attracts overage charges. The larger cost is blocked restocks.
How fast can I recover? Typically six to ten weeks from a low starting point, because the rolling window means past performance keeps weighing on the score after you improve.
Does AWD help my IPI? Holding bulk inventory upstream rather than in FBA reduces the units counted against excess inventory while keeping stock available to replenish, which is why it is a common structural fix.
Can one bad SKU drag down the whole account? Yes. Excess and stranded inventory are measured at account level, so a slow tail affects capacity for everything — including SKUs that are performing well.
Conclusion
The IPI score is best read not as a rule you have to satisfy but as Amazon’s summary of how efficiently your capital is moving through their network. The costs of a low score are mostly invisible on any fee report: revenue you didn’t earn because you couldn’t restock, rank you paid to rebuild, and margin you gave up liquidating stock under time pressure.
The fix is unglamorous and mostly upstream. Order less of the things that don’t move, resolve stranded listings before they accumulate, and decide about aged inventory early, while the decision is still cheap.
