Inventory Manager
Inventory is where Amazon sellers lose money in both directions. Stock out and you
lose rank and sales that do not come back. Overstock and you pay storage fees and
freeze cash. Restocking by feel produces stockouts and overstock at the same time.
This skill finds the reorder point and the order quantity that keep a product in
stock without burying cash in a warehouse, and projects the schedule 12 weeks
forward with capacity limits and seasonality baked in.
When to use this
- A seller does not have a reorder rule and restocks by feel.
- A product stocked out, lost rank, and the seller wants it never to happen again.
- Storage fees or long-term storage fees are eating margin.
- The IPI score is low or restock limits are biting.
- ASIN-level restock limits reactivated and the seller needs to fit within them.
- Stranded or aged inventory is sitting unsold.
- Q4 ramp planning where seasonal demand multiplies and lead times often slip.
- A new SKU stabilizing and needing its first forecast.
The framework. The Reorder Equation
Two numbers govern inventory: when to order and how much. Both come from sales rate
and lead time.
Reorder point. Order when stock on hand falls to:
Reorder point = (daily sales rate x total lead time in days) + safety stock
Total lead time is everything from placing the purchase order to the units being
live and sellable: production, freight, customs, Amazon receiving and check-in.
Sellers forget the receiving leg and stock out anyway.
Safety stock. A buffer for demand spikes and lead-time slippage:
Safety stock = daily sales rate x buffer days
Buffer days scale with volatility. a steady product with reliable supply needs ~14
days, a volatile or unreliable supply chain needs 21 to 30.
Order quantity. Enough to cover the cycle until the next reorder lands, balanced
against storage cost and cash. A common target is 60 to 90 days of cover. more than
that risks long-term storage fees and frozen cash, less risks stockouts on any lead-
time slip.
Running too lean is no longer only a stockout risk. Since April 2024 Amazon charges a
low-inventory-level fee on FBA standard-size units whose historical days of cover
stay low (the fee looks at both the short-term and long-term days-of-supply figures,
so chronically thin stock, not a brief dip, is what triggers it). It is a per-unit
surcharge that stacks on the normal fulfillment fee, which makes a too-tight reorder
quietly expensive. Treat the exact threshold and per-unit amounts as moving. Amazon
has adjusted both, so confirm the current numbers and your SKUs' status on the
Inventory Performance and fee pages in Seller Central. The practical takeaway. keep
enough days of cover that the fee does not trigger, which usually aligns with the same
60-to-90-day target the reorder equation already aims for.
The 12-week schedule. Days-of-Cover model
The reorder equation tells you when one PO fires. The schedule tells you when the
next three or four will. Two metrics drive it.
- DoC on hand = units on hand / daily sales rate. Watch daily.
- Reorder point as defined above.
Project DoC forward week by week for 12 weeks. apply any seasonal multiplier in the
weeks where a peak lands (see amz-seasonal-planning). When projected DoC dips below
the reorder point's days-equivalent, that is an order week. Compute order quantity
as target days of cover x daily rate, then apply the ASIN restock limit and the
account capacity limit as hard caps.
For peak windows, pull the order forward and increase the size to land stock before
the peak begins, not during it.
The IPI and FBA capacity limits
In March 2023 Amazon removed the hard IPI minimum that used to gate FBA storage and
replaced the old storage-volume and restock limits with a single monthly FBA
capacity limit measured in cubic feet. Capacity is now managed through Manage Your
Inventory Capacity in Seller Central, driven by recent sales, IPI, forecast demand,
and fulfillment-center availability together, not by a hard IPI floor that blocks
inbound shipments. IPI still matters as one input into how much capacity Amazon
grants, and a weak score still tightens the limit. it is driven by four levers: low
excess inventory, healthy sell-through, fixing stranded inventory, and keeping
in-stock rate high. Manage all four, and watch the capacity dashboard for the actual
limit on the account. Treat any specific IPI target you see quoted online as a moving
heuristic. confirm the current threshold and your account's status in Seller Central
rather than assuming a fixed number.
30-60-90 day IPI recovery plan
When a low IPI is tightening the capacity limit, recovery is sustained weekly action,
not a one-time push. Diagnose the single binding lever first (where the most points
are being lost), then work it on a phased schedule:
- Days 1 to 7. Diagnose and plan. Pull the four-lever breakdown from the inventory
dashboard. Open Manage Excess Inventory and the Stranded Inventory tool. Identify
the binding lever and build the weekly action list around it.
- Days 8 to 30. Hit the binding lever. Remove or liquidate excess, or fix stranded
listings, whichever is binding. Weigh removal versus liquidation per SKU, not flat:
removal costs a fee but frees capacity, liquidation loses some margin but recovers
cash. The score typically responds within 2 to 3 weeks.
- Days 31 to 60. Work the secondary levers. Push sell-through on the slow movers
with promotions and ad spend, and repair in-stock rate by fixing the reorder
schedule on the SKUs that keep dipping.
- Days 61 to 90. Stabilize and use the room. Weekly review to confirm the score
holds and the capacity limit loosened, then plan the restock that uses the recovered
capacity. Do not add inventory before the score recovers. more stock while IPI is
low worsens the excess metric and tightens the limit further.
IPI updates weekly, so review weekly across the whole 90 days and flag if either the
score or the capacity limit moves the wrong way.
Step by step
-
Collect inputs. Per SKU: daily or weekly sales rate (use last 90 days of
real sales, not historical average), current stock on hand, each leg of the
lead time (production, freight, customs, Amazon receiving), demand volatility,
any seasonality multipliers per week, current IPI, the account capacity limit,
and any ASIN-level restock limit.
-
Compute total lead time including the Amazon receiving leg.
-
Set safety stock from the volatility. larger buffer for spiky or seasonal
SKUs.
-
Compute the reorder point. Flag any SKU already at or below it as order now.
-
ASIN restock limit binding check. Compare the desired order quantity (target
DoC x daily rate) against the ASIN-level restock limit. If the limit is below
the desired quantity, the limit is binding. the SKU will run lean unless the
limit is raised. plan to request a capacity increase or to space POs more
tightly. Note this as a constraint in the output, not as a fix to ignore.
-
Project the 12-week DoC schedule. Apply seasonal multipliers per week.
Mark each order week and order quantity.
-
Set the order quantity for 60 to 90 days of cover, capped by the binding
restock limit and adjusted for upcoming seasonality. Pull peak-window POs
earlier and larger.
-
Scan for waste. Identify aged inventory approaching long-term storage fees
and any stranded inventory (in stock but not sellable), and recommend the fix:
markdown, removal, or relisting.
-
Check the IPI levers and name the weakest one.
-
Run the quality check, then deliver.
Output format
## Inventory Plan. [SKU or catalog]
### Per SKU
[SKU] . daily rate . on hand . reorder point . status [ok / order now] . order qty
. restock limit binding? [y/n]
...
### 12-week DoC projection (per priority SKU)
Week 1: [DoC] Week 2: [DoC] ... Week 12: [DoC]
Seasonal multipliers applied: [Nx in week X, ...]
### Order schedule
PO 1. Place by [date]. Quantity: [units]. Arrives in stock: [date]
PO 2. ...
### Waste scan
Aged inventory: [SKUs approaching long-term storage fees]
Stranded inventory: [SKUs in stock but not sellable]
### Constraints
Restock limit binding on: [SKUs]
Capacity increase request needed: [y/n]
### IPI
Current: [score] Weakest lever: [which] Fix: ...
### This week
[ordered by urgency]
Worked example
A SKU sells 25 units a day. Lead time: 30 days production, 25 days freight and
customs, 7 days Amazon receiving, total 62 days. Volatility is moderate, buffer 21
days. ASIN restock limit 3,000 units.
Safety stock: 25 x 21 = 525. Reorder point: (25 x 62) + 525 = 2,075 units. The SKU
is at 2,200 on hand. DoC = 88 days. Reorder point hits at roughly week 1-2. Order
quantity for 75 days of cover: about 1,875 units. ASIN restock limit binding check.
3,000 limit, 1,875 desired, not binding at this order, fine. A week-10 peak with a
3x multiplier needs an earlier larger PO at week 7. landed-before-peak schedule
shows PO1 at week 1 (1,875 units), PO2 at week 7 (3,000 units, restock limit now
binding for the peak, request an increase before week 5).
The seller had been reordering at 600 units on hand and stocking out during every
freight delay. The new schedule shows exactly when to place each PO and how much,
with no stockout risk and no excess.
Quality check
- Total lead time includes the Amazon receiving leg, not just production and freight.
- Daily rate is computed from real recent sales, not historical average.
- Safety stock scales with the SKU's demand volatility.
- Every SKU at or below its reorder point is flagged order now.
- DoC is projected weekly for 12 weeks, not as a snapshot.
- ASIN-level restock limit is checked as a binding constraint, not ignored.
- Account capacity limit is respected.
- Order quantity targets 60 to 90 days of cover.
- Seasonal pre-orders are placed early enough to land before the peak.
- Aged and stranded inventory are surfaced with a specific fix.
- The weakest IPI lever is named.
Common mistakes
- Forgetting the receiving leg. Units are not sellable the day they arrive at the
warehouse. that gap causes stockouts.
- Reordering by feel or on a fixed cadence. A monthly PO regardless of cover is
wrong both ways: stockouts in fast weeks, overstock in slow.
- Overordering. Burying cash and triggering long-term storage fees to feel safe.
- Ignoring the ASIN restock limit. Ordering 5,000 when the ASIN limit is 3,000
means 2,000 units stuck or expensive to send via STA capacity.
- Ignoring stranded inventory. Units in the warehouse that are not even listed as
sellable, earning nothing and costing storage.
- No seasonal multiplier. A baseline forecast through a peak guarantees a stockout.
- Letting IPI drift. A low IPI caps restock limits right when a product is taking off.
Built by Jay GPT Pro
Part of Amazon Pro Skills. Production-grade skills for serious Amazon sellers.
Free and open. Built by Jay Margaliot.
I share a new AI play for Amazon sellers every week, free, in my WhatsApp group.
Join here:
https://chat.whatsapp.com/ILX65p1yWcaIG3c9WGHpTY