with very different costs
Inventory management has two failure modes. Both are expensive. Most brands manage for one and get burned by the other.
Stockout caused by reactive reordering
When a reorder is triggered by a human noticing that stock is low, it is already too late for the replenishment to arrive before the stockout. FBA lead time plus production lead time plus transit time means the reorder must happen weeks before the stock hits zero — not when it hits zero.
Overstock driven by demand overestimation
Ordering ninety days of inventory based on an optimistic sales projection produces FBA storage fees, long-term storage charges, and capital tied up in product that is not moving. The cost of overstock is less visible than a stockout but equally real.
No seasonality adjustment in the reorder model
A reorder model built on trailing average sales applies the same demand assumption to Q4 as it applies to Q2. Categories with seasonal demand patterns require a forecasting model that accounts for seasonality — otherwise Q4 stockouts and Q1 overstock are structurally inevitable.
demand-calibrated replenishment system.
Each component addresses a different input to the replenishment decision. Together they produce a reorder recommendation that accounts for demand, lead time, and cost simultaneously.
Sales Velocity & Demand Modeling
Rolling sales velocity by SKU, adjusted for seasonality, promotional lift, and advertising spend changes. The demand model is the foundation — every other calculation depends on it being accurate.
Lead Time Integration
Production lead time, transit lead time, and FBA processing time combined into a total lead time model for each SKU and supplier. The reorder point is calculated from this total lead time, not from a fixed rule applied uniformly.
Safety Stock Calculation
Statistical buffer stock calculated from demand variability and lead time variability for each SKU. Safety stock prevents stockouts caused by demand spikes or transit delays without requiring excess inventory across the entire catalog.
Automated Reorder Alerts
ARIA monitors FBA stock levels against the demand model and lead time and generates a reorder alert at the correct reorder point. The alert includes the recommended order quantity, the projected stockout date without action, and the supplier lead time to confirm.
Calibrate continuously.
A replenishment model built from assumptions degrades as conditions change. Ours is built from actuals and recalibrated monthly.
Demand Baseline
We pull twelve to twenty-four months of sales data by SKU and build the demand model: trailing velocity, seasonality index, and promotional lift adjustment. For SKUs with limited history, we use category benchmarks until sufficient data accumulates.
Lead Time & Reorder Point Calculation
Supplier lead times are documented and combined with FBA transit and processing time to calculate total lead time for each SKU. Reorder points and safety stock levels are calculated from the demand model and lead time model together.
System Configuration & Ongoing Monitoring
ARIA is configured with the reorder parameters for each SKU. Alerts go out when stock hits the reorder point. The model is reviewed and updated monthly as actual demand and lead time data replaces projections.
that runs before the stockout.
Every alert comes with the recommended action, not just the warning.
SKU-level demand model with seasonality and promotional adjustments.
Total lead time by SKU and supplier, integrated into the reorder calculation.
Calculated reorder points and safety stock levels for each SKU.
ARIA alerts configured for each SKU at the correct reorder point.
Demand and lead time model updated monthly with actuals.
Inventory forecasting module fully integrated with ARIA performance reporting.
Operations & Platform includes additional services that compound on this one.
Reorder before
the rank starts to fall
We work through referrals. If you have been referred, send us a message and we will model your current reorder points against actual lead time and show you where the gaps are.