Stock-Out Prediction: How Operations AGI Gives You 14 Days of Advance Warning
A stock-out is not a warehouse problem. It is a revenue problem that was created 14 to 21 days earlier, when someone should have reordered and did not. The reason they did not reorder is almost never negligence it is that the signal was buried. Sales velocity data lived in the OMS. Inventory data lived in the WMS. The calculation that would have revealed the problem required joining two systems, applying a sell-through model, and checking the output against supplier lead times a task that took 30 minutes if someone ran it, and happened fortnightly if the team was disciplined, monthly if they were not. By the time the stockout appeared on a dashboard, the reorder window had already closed. Operations AGI closes that window permanently by running the calculation continuously, on every SKU, across every warehouse and channel, and alerting 14 days before the stock-out date rather than on the day it occurs.
Operations AGI monitors live sales velocity by SKU, channel and warehouse location and projects stock-out dates based on current sell-through rates against available inventory. When the projected stock-out is within 14 days, a structured alert is generated with recommended reorder quantities. This piece explains the prediction model.
The Prediction Model: How It Works
Operations AGI's stock-out prediction model operates on a rolling sell-through velocity calculation at the most granular level available: individual SKU, by warehouse location, by sales channel. Velocity is calculated across three windows simultaneously 7-day, 14-day, and 30-day rolling averages weighted by recency and adjusted for identified seasonality patterns in the SKU's historical sales data. The multi-window approach is critical because it distinguishes between a genuine velocity increase (consistent across all three windows) and a spike (elevated in the 7-day window but not in the 14-day), which would otherwise trigger premature reorder recommendations.Available inventory is read directly from the connected WMS in real time not from a daily export or a weekly snapshot. For brands with multi-location inventory, the model tracks stock at the warehouse level and accounts for transfer lead times between locations when calculating the effective available stock for a given fulfilment zone. An SKU with 800 units in Delhi and 50 units in Mumbai, serving a customer base that is 60% in Maharashtra, has an effective available inventory figure that is different from its total inventory figure and the prediction model accounts for this.The stock-out date projection is calculated as: (current inventory at location) ÷ (weighted sell-through velocity) = days of cover remaining. When days of cover falls within the configurable alert threshold default 14 days a structured alert is generated. The alert is produced before the reorder needs to be placed, accounting for the SKU's configured supplier lead time. If a SKU's supplier lead time is 10 days and current days of cover is 14, the effective reorder window is 4 days. Operations AGI alerts on the 14-day mark so the procurement team has 4 days of decision time before the reorder becomes urgent.
What Goes Into the Velocity Calculation
Raw sales velocity units sold per day is the starting point but not the full picture. Operations AGI adjusts raw velocity for three factors that systematically distort naive sell-through projections. The first is promotional inflation: if the SKU ran a sale event in the trailing 7-day window, the velocity figure is inflated by the promotional uplift and should not be used to project forward demand at normal price. Operations AGI detects active promotional periods from the connected OMS and marketing calendar and applies a de-inflation factor based on the SKU's historical promotional vs. organic velocity ratio.The second factor is seasonal pattern adjustment. Operations AGI builds a SKU-level seasonality index from 12 to 24 months of historical sales data, identifying recurring patterns summer velocity increases for certain categories, pre-festival demand surges, post-season clearance depressions and adjusting the forward velocity projection accordingly. A winter clothing SKU showing a 14-day days-of-cover figure in October should be evaluated against the November velocity projection, not the October velocity, because the forward demand is structurally higher than the current trailing average.The third factor is channel mix shift. A SKU that is growing its share of marketplace sales relative to D2C website sales will have a different effective velocity than its total units sold figure suggests because marketplace fulfilment draws from a different inventory pool in multi-location setups. Operations AGI tracks velocity by channel and models the inventory draw implications of channel mix shifts, ensuring that reorder recommendations account for where the stock will be consumed, not just how much will be consumed.
The Alert Structure: What Procurement Receives
When Operations AGI generates a stock-out alert, it is structured for immediate action not as a notification requiring further investigation. The alert contains: the SKU name and ID, the warehouse location, the current inventory count, the weighted velocity figure, the projected stock-out date, the reorder lead time, the effective decision window, and a recommended reorder quantity. The recommended reorder quantity is calculated using an Economic Order Quantity model parameterised with the SKU's historical demand variance, the carrying cost estimate, and the configurable safety stock level for that product category.The alert also includes a 30-day demand forecast for the SKU, showing the projected daily demand with confidence intervals across the alert period. This gives the procurement manager visibility into whether the stock-out risk is driven by a temporary velocity spike that may normalise before the stock-out date, or by a sustained velocity increase that warrants a larger-than-standard reorder. The confidence intervals are particularly important for this decision: a wide confidence interval on a marginal stock-out projection is a different risk profile than a narrow confidence interval on the same projection.For SKUs where Operations AGI has historical data on supplier fulfilment reliability cases where the actual lead time has consistently differed from the stated lead time the alert incorporates a lead time adjustment factor. If a supplier states a 10-day lead time but has historically delivered in 13 to 14 days across 8 of the last 12 orders, the effective lead time used in the decision window calculation is 13 days, not 10. This prevents the common scenario where procurement teams rely on stated lead times and consistently experience stock-outs despite technically placing reorders within the nominal window.
Multi-SKU and Portfolio-Level Visibility
Beyond individual SKU alerts, Operations AGI generates a daily inventory health brief that gives the procurement and operations team portfolio-level visibility across all monitored SKUs. The brief is sorted by urgency decision window remaining and categorised into three tiers: critical (decision window under 5 days), urgent (5 to 10 days), and advisory (10 to 21 days). This tiered view allows the procurement team to allocate attention appropriately rather than treating all stock-out risks with equal urgency.The portfolio brief also surfaces correlated risk clusters: situations where multiple SKUs in the same product family, from the same supplier, or fulfiled from the same warehouse are simultaneously approaching stock-out. A correlated cluster often indicates a demand pattern shift a viral social media moment, a seasonal inflection, a competitor stock-out creating demand spillover that warrants a strategic response rather than a series of individual reorder decisions. Operations AGI flags these clusters explicitly so the procurement manager can assess whether the situation requires a supplier conversation about capacity rather than a standard reorder.
Integration with Supplier and Procurement Workflows
For suppliers integrated with SuperManager AGI through API or EDI connections, Operations AGI can initiate purchase orders autonomously when a stock-out alert is generated and the reorder quantity is within the procurement manager's pre-approved autonomy threshold. The PO is generated in the connected ERP, transmitted to the supplier, and logged in the operations dashboard. The procurement manager receives a confirmation notification with a 24-hour override window. For suppliers not yet integrated, Operations AGI generates a draft purchase order in the brand's standard format with all fields pre-populated from the SKU master and supplier database, ready for the procurement manager to review and send.Every autonomous or semi-autonomous reorder action is logged with the full decision chain the velocity calculation, the projected stock-out date, the recommended quantity, the model confidence level, and the approval or override action taken by the human. This audit trail serves two purposes: it allows the operations team to review and refine the model's accuracy over time, and it provides a documented record of inventory decision-making for organisations that require procurement governance at the finance or board level.