Amazon Europe inventory planning: reorder points, split stock and cross-border risk

1. The short answer

Build the plan in this order:

1. define the sellable unit, store and stock location;
2. measure demand using a base and downside case;
3. record every lead-time stage from purchase order to available stock;
4. separate predictable lead time from the risk buffer;
5. decide which inventory states are sellable, in transit, reserved, returned or unusable;
6. calculate reorder point and order quantity;
7. test central stock against split stock; and
8. set alerts, owners and a 30, 60 and 90-day review cadence.

If you cannot explain how long it takes for an approved purchase order to become sellable inventory, you do not have a reorder point yet.

2. 1. Define the unit and location

Use one plan for one SKU, condition, store and stock route.

Do not mix one unit sold in Germany with a case pack purchased from a supplier. Convert the inputs to the same sellable unit before calculating.

3. 2. Build the end-to-end lead-time record

The lead time starts when you commit to replenishment and ends when inventory is available for sale at the chosen route.

Record the stages separately:

These are illustrative assumptions. Replace them with your own evidence. If a stage is not measurable yet, use a conservative assumption and label it as unknown.

The landed-cost guide owns customs and purchase-to-warehouse cost construction. F6 uses its timing and route facts without rebuilding the customs calculation.

4. 3. Separate demand rate from the safety buffer

Use a demand rate that matches the decision.

Possible inputs include:

• average daily or weekly units;
• recent 30-day and 90-day sales;
• seasonal peak rate;
• promotions or advertising that changed demand;
• stock-out periods that make sales look lower; and
• the destination store's separate demand pattern.

Do not use one blended Europe average if the stock is split by country. A store with eight weekly units can need a different reorder trigger from a store with 40 weekly units, even when the total looks stable.

Safety stock is the protection against uncertainty. Explain what it covers:

• demand volatility;
• supplier variation;
• customs or freight delay;
• receiving and inspection delay;
• a promotional peak; or
• a service promise that cannot tolerate a stock-out.

Do not hide every unknown inside safety stock. A large buffer can make a weak route look safe while tying up cash and storage.

5. 4. Worked example with illustrative assumptions

The following example is invented. It is a planning demonstration, not a demand forecast.

Assume one SKU in one store:

• base demand: 3 units per day;
• downside demand: 4 units per day;
• full replenishment lead time: 44 days;
• base safety stock: 30 units; and
• downside safety stock: 50 units.

The calculation is:

The seller currently has 180 sellable units available and 60 units in transit. If the next 44 days consume 132 base units, projected sellable stock at the next receipt is 108 units before any return or damage adjustment. That is below the base reorder point of 162. The route should trigger a replenishment review now.

In the downside case, the gap is larger. The seller may need to reduce the order quantity, increase the purchase timing or review the service promise. The model does not automatically approve a purchase order.

6. 5. Count inventory states correctly

Amazon's current inventory management guidance describes tools for monitoring inventory, shipments and replenishment. Seller Central also distinguishes inventory that is available, in transit, reserved, stranded or unfulfillable. Review the current inventory help for the applicable store and policy.

For planning, keep these rows separate:

Do not count every unit shown in a dashboard as sellable stock. A unit that cannot be bought by the customer does not protect the customer promise.

7. 6. Central stock versus split stock

Split stock can improve service. It can also create duplicated safety stock, more storage, more country questions and slower rebalancing.

Central stock may suit you when

• demand is still uncertain;
• one route can meet the delivery promise;
• the product is slow-moving or expensive;
• cash is limited;
• customs and stock-country work is not yet settled; or
• the seller needs one place to inspect and reconcile returns.

Split stock may suit you when

• demand is proven in more than one market;
• local service materially changes the customer proposition;
• the fulfilment programme requires or supports local placement;
• stock can be seen by country and condition;
• the seller can fund duplicated safety stock; and
• VAT, import, product and returns owners are named.

Use the EFN, Pan-European FBA and local inventory comparison for the network decision. Use F2 for the stock-location map. F6 should not decide tax treatment from the inventory diagram.

8. 7. Model cash and aged-inventory exposure

The reorder point is only half the decision. The order quantity can create a larger risk than the stock-out it prevents.

Record:

• supplier minimum order quantity and case pack;
• deposit and balance timing;
• freight and customs cash timing;
• storage and aged-inventory cost;
• return and unsellable-stock reserve;
• exchange-rate exposure;
• removal or disposal cost; and
• the contribution lost when cash is tied up in slow stock.

The contribution-margin guide should show the storage, removal, returns and slow-stock assumptions. Do not create a second margin model inside the reorder sheet.

9. 8. Reorder alerts and exception log

Set alerts for:

• projected sellable stock falling below reorder point;
• lead time moving beyond the model;
• demand changing by the agreed percentage;
• stock in transit missing the receiving date;
• returns or unsellable units exceeding the reserve;
• stock split by country changing;
• contribution falling below its guardrail; and
• a programme, supplier, importer or warehouse change.

The alert does not always mean “buy more”. It may mean reduce the range, change the route, delay a market, expedite a shipment or stop replenishment.

10. 9. 30, 60 and 90-day review cadence

Day 30

Check demand rate, stock states, receiving delay, return rate, listing availability and actual per-unit costs. Correct obvious data errors.

Day 60

Compare the base and downside cases with actual sales. Review supplier reliability, customs timing, stock age and cash tied up. Re-score central versus split stock if the service promise changed.

Day 90

Decide whether to maintain, change or stop the route. Update the reorder point, safety stock rationale, order quantity, stock location and owner map.

Recalculate immediately when a supplier, route, packaging, programme rule, stock country or customer promise changes. Annual review is not enough for a live route with moving inputs.

11. 10. When to stop replenishing

Stop or reduce replenishment when:

• the downside demand case no longer supports the stock commitment;
• contribution falls below the agreed guardrail;
• returns or unsellable stock exceed the reserve;
• product evidence or listing status is unresolved;
• stock age reaches the exit limit;
• a new country creates unresolved tax or compliance exposure; or
• the supplier minimum order is larger than the route can fund.

The right action may be to sell through, move stock, reduce the SKU range, renegotiate the case pack or stop the product. For a new-market route, also use when not to expand to another Amazon marketplace to test whether more inventory is the right decision at all. A reorder formula should make those decisions visible.

12. The practical conclusion

Inventory planning is a chain of assumptions.

Demand, production, inspection, freight, customs, receiving, availability, returns, stock location and cash all affect the date when a reorder is needed. Make each one visible. Keep central and split-stock cases separate. Use Amazon's tools as inputs, not as a replacement for your own route record.

The best reorder point is not the most precise number. It is the number whose assumptions you can explain and update when the facts change.