Storage fee forecasting starts before inventory arrives

Storage fees feel like a month-end surprise only when the operation waits until month end to look. The inputs are visible earlier: units on hand, cubic feet, sell-through, inbound quantity, seasonal multipliers, and aged inventory exposure.

The forecast should start when the purchase order is approved, not when the inventory reaches Amazon. A shipment that looks profitable on unit economics can become weak if too much cube arrives too early.

Forecast cube days

Unit count is not enough. Model cubic feet multiplied by expected days in FBA. A small expensive item and a bulky low-margin item may have the same unit count, but their storage economics are completely different.

For each SKU, forecast:

• starting units and cube
• inbound units and expected receiving date
• daily demand by recent velocity and seasonality
• expected ending units
• projected cube days
• aged inventory date

Add operational decisions

Forecasts are useful only if they produce actions. If a SKU will cross an aged-inventory threshold, decide whether to discount, coupon, remove, bundle, or stop replenishing. If a slow SKU is bulky, consider FBM or smaller FBA replenishment batches. If Q4 storage multipliers matter, move the decision date earlier.

Review forecast accuracy

Every month, compare forecasted storage fees with actual charges. Was the miss caused by demand, receiving delay, catalog cube, or forgotten inbound inventory? Tag the reason and improve the next forecast.

Storage fee control is rarely one dramatic move. It is a habit of connecting cube, velocity, and replenishment timing before Amazon sends the bill.

Forecast by decision date

A forecast that arrives after the team can act is only an explanation. Build storage views around decision dates: purchase order approval, inbound shipment creation, promotion planning, aged-inventory review, and removal cutoff. Each date should show what happens if the team does nothing and what changes if the team discounts, delays, removes, or replenishes less.

This framing helps buyers and operators see storage as part of the buy decision. A bulky SKU that looks profitable at landed cost may be unattractive if it needs three months of FBA storage before selling through. A smaller SKU with lower gross margin may be healthier if it turns faster and uses less capacity.

Treat forecast misses as data

Every forecast miss should improve the next model. If demand was wrong, adjust the velocity window or seasonality assumption. If receiving was late, update the inbound lead time. If cube was wrong, fix catalog dimensions. If a planned promotion did not happen, add promotion execution as a risk factor instead of pretending the forecast was purely mathematical.

The best storage process is not a perfect prediction. It is a learning loop that makes each buy, shipment, and liquidation decision more grounded than the last one.

Segment by product behavior

Storage risk is different for replenishable, seasonal, launch, and cleanup SKUs. A replenishable SKU needs a stable ordering rhythm. A seasonal SKU needs a sell-through deadline. A launch SKU needs a limited test quantity. Cleanup inventory needs a recovery plan rather than another replenishment cycle.

Forecasting should reflect those roles. If every SKU uses the same target days of cover, the team will overstock some products and understock others. Segment first, then set storage thresholds and action dates that match the product's job in the catalog.

Include removal economics

A forecast should show the cost of waiting and the cost of acting. Removal, liquidation, discounting, and holding all have different cash outcomes. When a SKU is projected to age, compare those options early enough that the team can choose deliberately instead of reacting after fees have already accumulated. The earlier the comparison happens, the more options remain open.