Inventory Risk Alerting

Overview

This skill continuously evaluates inventory positions across SKU-location combinations to identify and classify overstock and stockout risks before they impact customer service or working capital. It computes forward-looking weeks of supply against demand forecasts, evaluates service level probability, and quantifies financial exposure from excess, slow-moving, and obsolete (SLOB) inventory. Alerts are severity-tiered and paired with specific corrective actions.

When to Use

Daily or weekly inventory health review across distribution network
When fill rate drops below target or customer complaints about availability spike
During monthly S&OP to highlight portfolio-level inventory risks
When new product launches require cannibalization risk monitoring on existing SKUs
Before seasonal peaks to ensure forward coverage meets demand surge
When CFO requests working capital exposure analysis on excess inventory

Required Inputs

Input	Description	Format
`inventory_snapshot`	Current on-hand, in-transit, and on-order quantities by SKU-location	Structured object
`demand_forecast`	Forward demand forecast (weekly, 12+ weeks)	Numeric array per SKU-location
`safety_stock_params`	Target service level, demand variability, lead time variability	Structured object
`lead_times`	Supplier and replenishment lead times by SKU	Numeric (days)
`product_lifecycle`	Product status (active, end-of-life, seasonal, new launch)	Enum
`cost_data`	Unit cost, carrying cost rate (% of COGS per annum)	Numeric
`sales_velocity`	Trailing 4-week and 13-week average daily sales	Numeric

Methodology

Step 1: Weeks of Supply Calculation

Weeks_of_Supply = (On_Hand + In_Transit + On_Order - Backorders) / Weekly_Demand_Forecast

Classify against category-specific targets:

Risk Zone	WoS vs. Target	Classification
Critical Stockout	< 1 week	Immediate action required
Stockout Risk	< Safety Stock coverage	Elevated risk
Healthy	Within target range (±20%)	No action
Overstock Warning	> 1.5× target	Monitor and reduce
Critical Overstock	> 2.5× target or > 26 weeks	Markdown/liquidation candidate

Step 2: Safety Stock & Service Level Projection

Calculate safety stock using the standard formula:

Safety_Stock = Z × √(LT × σ_d² + d² × σ_LT²)

Where:

Z = service level Z-score (e.g., 1.65 for 95%, 2.33 for 99%)
LT = average lead time in days
σ_d = standard deviation of daily demand
d = average daily demand
σ_LT = standard deviation of lead time

Project forward service level:

Projected_Fill_Rate = P(Demand ≤ Available_Inventory) over review period

Flag when projected fill rate drops below target service level within the lead time horizon.

Step 3: Excess & Obsolescence (E&O) Exposure

Quantify financial risk from overstocked inventory:

Excess_Qty = max(0, On_Hand - (Forecast_Demand × Forward_Weeks + Safety_Stock))
Excess_Value = Excess_Qty × Unit_Cost
Carrying_Cost_Exposure = Excess_Value × (Carrying_Cost_Rate / 52) × Estimated_Weeks_to_Clear

Apply aging buckets:

0-30 days excess: Manageable through demand pull-forward or transfer
31-90 days excess: Consider promotional markdowns
91-180 days excess: Aggressive markdown or secondary channel liquidation
180+ days excess: Write-off candidate, provision in financial reserves

Step 4: Velocity-Based Risk Scoring

Compute a composite risk score (0-100) per SKU-location:

Risk_Score = w1 × Stockout_Probability + w2 × Excess_Severity + w3 × Lifecycle_Risk + w4 × Demand_Volatility

Default weights: w1=0.35, w2=0.25, w3=0.20, w4=0.20

Lifecycle risk multipliers:

End-of-life SKU with >8 WoS: 2.0× multiplier on excess severity
New launch SKU with <2 WoS: 1.5× multiplier on stockout probability
Seasonal SKU past peak: 1.8× multiplier on excess severity

Step 5: Alert Generation & Prioritization

Generate alerts prioritized by:

Revenue at risk (stockout alerts): Lost_Sales = Unmet_Demand × Selling_Price × Fill_Rate_Gap
Capital at risk (overstock alerts): Excess_Value + Projected_Carrying_Cost + Markdown_Exposure
Customer impact: A-class SKUs and key accounts prioritized

Output Specification

inventory_risk_report:
  generated_at: "2026-02-07T08:00:00Z"
  portfolio_summary:
    total_skus_monitored: 4500
    stockout_risk_count: 127
    overstock_risk_count: 342
    healthy_count: 4031
    total_excess_value: 2450000
    projected_lost_sales: 890000
  critical_alerts:
    - alert_id: "STOCK-001"
      severity: "critical"
      type: "stockout_imminent"
      sku_id: "SKU-78901"
      location: "DC-EAST-02"
      weeks_of_supply: 0.4
      projected_stockout_date: "2026-02-10"
      revenue_at_risk: 125000
      recommended_action: "Expedite PO-44521 (ETA 2/12) or transfer 500 units from DC-WEST-01 (3-day transit)"
      root_cause: "Demand spike from competitor OOS driving substitution"
    - alert_id: "EXCESS-001"
      severity: "high"
      type: "overstock"
      sku_id: "SKU-34567"
      location: "DC-CENTRAL-01"
      weeks_of_supply: 34.2
      excess_value: 89000
      carrying_cost_monthly: 1480
      recommended_action: "Initiate 20% markdown via promotional calendar or transfer to high-velocity locations"
      root_cause: "Forecast over-bias of +40% during Q4 promo that underperformed"

Analysis Framework

ABC Classification for Alert Thresholds

Class	Revenue Share	Stockout Threshold	Overstock Threshold	Review Frequency
A	Top 80% revenue	< 2 WoS	> 8 WoS	Daily
B	Next 15% revenue	< 1.5 WoS	> 12 WoS	Weekly
C	Bottom 5% revenue	< 1 WoS	> 16 WoS	Bi-weekly

Key Performance Indicators

Inventory Turns: COGS / Average_Inventory_Value — target varies by category (grocery: 15-25, general merch: 4-8)
GMROI: Gross_Margin / Average_Inventory_Cost — target > 2.0 for healthy ROI
Days of Supply: Average_Inventory / Average_Daily_COGS — complement to WoS for financial view
Fill Rate: Units_Shipped / Units_Ordered — target 95-98.5% for A-class items
Perfect Order Rate: Orders delivered on-time, in-full, undamaged, with correct documentation

Examples

Example 1 — Stockout Risk Alert

"ALERT [CRITICAL]: SKU-78901 (Premium Coffee 12oz) at DC-EAST-02 has 0.4 weeks of supply remaining with weekly demand of 1,200 units. No open POs. Revenue at risk: $125,000 over next 3 weeks. Action: Place emergency PO for 5,000 units with expedited shipping, estimated 4-day lead time. Alternative: Transfer 2,000 units from DC-WEST-01 where WoS = 6.8."

Example 2 — Overstock Alert with E&O Quantification

"ALERT [HIGH]: SKU-34567 (Seasonal Sunscreen SPF-30) at DC-CENTRAL-01 has 34.2 WoS (target: 8 WoS). Excess quantity: 8,900 units valued at $89,000. Monthly carrying cost: $1,480. Product is post-season with declining demand. If not actioned within 60 days, projected markdown loss is $44,500 (50% of excess value). Recommend immediate 20% promotional markdown and cross-DC rebalancing."

Guidelines

Run stockout risk detection daily for A-class items; weekly for B/C-class
Always include the financial impact in alerts — units alone do not drive executive action
Factor in in-transit and on-order inventory to avoid false stockout alerts
For seasonal items, use seasonal demand forecast (not trailing average) for WoS calculations
Distinguish between planned promotional overstock (intentional forward buy) and unplanned excess
Cross-reference stockout alerts with substitution availability — if a direct substitute is in stock, lower severity
Account for minimum order quantities (MOQs) when recommending reorder actions
Include lead time in stockout horizon — alert must fire before lead time window closes
Track alert-to-action conversion rate to calibrate thresholds and reduce alert fatigue

Validation Checklist

Inventory snapshot is current (within 24 hours for daily alerts)
Demand forecast aligns with the latest consensus forecast from S&OP
Safety stock parameters reflect current service level targets by category
Lead times are updated to reflect actual recent supplier performance, not contractual
E&O calculations use current standard cost, not historical purchase price
Seasonal classification is current — items are flagged pre-season, in-season, and post-season
Alert deduplication logic prevents repeated alerts for the same SKU-location-risk within the suppression window
Financial exposure totals reconcile with inventory valuation reports from finance
Recommended actions are feasible (check supplier capacity, transfer lane availability, markdown approval authority)