Stockout Costs: Formulas, Examples, and Proven Ways to Prevent Lost Sales
Stockout costs are the measurable financial impacts when demand exceeds available inventory. For ecommerce operators and inventory managers, understanding stockout costs helps set service levels, size safety stock, choose reorder points, and reduce lost sales and penalties.
Stockout costs explained
Definition and scope
Stockout cost is the total economic impact when an item is unavailable to fulfill demand. It includes direct losses (e.g., lost margin, expediting) and indirect effects (e.g., churn, brand damage).
Scope should be set per SKU-location and per channel. Measure both immediate event costs and longer-term impacts such as customer lifetime value erosion.
Direct vs indirect components
Direct components:
Lost gross margin on unfulfilled demand
Backorder handling and expediting costs
Retailer or marketplace penalties (e.g., OTIF, late/short shipment fees)
Extra picking, split shipments, and customer service time
Refunds, discounts, and returns related to delays
Indirect components:
Customer churn and reduced repeat purchase rates
Lower average order value due to substitutions or partial shipments
Brand and search ranking impacts (ecommerce) from OOS signals
Store labor inefficiency and planogram/shelf integrity issues
Stockout vs backorder vs on-shelf OOS
Stockout: Demand occurs but on-hand available-to-sell is zero at the point of fulfillment.
Backorder: Demand is captured with a commitment to ship later; revenue may be delayed but not fully lost.
On-shelf OOS: Store has inventory in the backroom/DC but shelf is empty; the customer still experiences an OOS even if inventory exists in the network.
These distinctions change the model: lost sales vs backorder recovery, and where to intervene (store execution, DC allocation, or supply).
Why stockouts happen
Forecast error and demand spikes
Under-forecasting and positive bias in promotions or new launches.
Demand volatility from seasonality, weather, media, and competitor actions.
Parameter drift when demand patterns change faster than forecast updates.
Lead time variability and supplier risk
Unstable supplier lead times from capacity, quality, or logistics constraints.
Port congestion, customs delays, and carrier variability.
Single-sourcing and long inbound lead times magnifying risk exposure.
Process/data issues and parameter drift
Stale reorder points, MOQs that overconstrain replenishment, or incorrect pack sizes.
Data quality problems: mismatched SKUs, late receipts, phantom stock, and inaccurate on-hand.
Governance gaps: no regular review of service levels, safety stocks, or lead-time assumptions.
Promotions, seasonality, and new items
Demand uplift misestimated or cannibalization ignored.
Slow assortment rationalization creating internal competition for inventory.
New-item demand lacks history; proxy modeling and guardrails are needed.
More time, More Sales
How to calculate stockout costs
Choose a modeling approach: lost sales or backorder
Pick a primary approach by SKU/channel:
Lost sales model: Use when customers are unlikely to wait (fast-moving retail, impulse buys).
Backorder model: Use when demand is captured and fulfilled later (MTO, B2B, subscription).
Hybrid: Split demand into recoverable vs lost, based on observed substitution and wait rates.
Decision rule (practical):
High price, specialized B2B, or contract orders → mostly backorder model.
Low switching cost, broad substitutes, or promotional retail → mostly lost sales model.
Key variables and formulas
Define:
p: unit price
c: unit cost
m = p − c: unit gross margin
Qshort: units short (unfulfilled initial demand)
r: recovery rate (fraction of Qshort eventually shipped later)
b: backorder cost per recovered unit (handling/expedite/discount)
pen: external penalty or chargeback
svc: customer service/time costs
churn: monetary impact of lost future contribution from churned customers
Lost sales model:
Units lost = (1 − r) × Qshort
Direct cost = Units lost × m
Total cost ≈ Units lost × m + r × Qshort × b + pen + svc + churn
Backorder model:
Direct cost = Qshort × b + pen + svc
Margin timing impact may be modeled with discount rate if material.
Quick Qshort methods:
Event-based: Qshort ≈ demand during OOS window − available alternative supply.
Probabilistic: expected shortage per cycle from service-level design (use fill-rate formulas or simulation).
Step-by-step numeric example
Scenario (D2C SKU):
Average daily demand = 80 units; two-day OOS → demand = 160 units.
On-hand at OOS start = 10 units; Qshort = 150 units.
Margin m = $12/unit.
Recovery r = 0.35 (35% customers accept backorder or buy later).
Backorder cost b = $3/unit.
Penalties pen = $0 (D2C storefront).
Customer service time cost svc = $200 for the event.
Churn impact: 5% of lost customers would have bought again once next month; expected future contribution per repeat $10; of the 97.5 units lost, assume 30% would have been repeat-capable customers → churn units = 29.25 → churn = 29.25 × $10 = $292.50.
Calculations:
Units lost = (1 − 0.35) × 150 = 97.5
Lost margin = 97.5 × $12 = $1,170
Backorder costs = 0.35 × 150 × $3 = $157.50
Event service cost = $200
Churn = $292.50
Total stockout cost ≈ $1,170 + $157.50 + $200 + $292.50 = $1,820
Interpretation: Even a short, two-day stockout can exceed $1,800 for a single SKU. Larger assortments or marketplace penalties can drive this higher.
Estimating indirect costs (churn, brand, penalties)
Churn: Estimate using pre/post repeat rate or cohort drop for customers exposed to OOS. Apply expected future contribution (margin × projected repeats).
Brand/search impacts: For ecommerce, model a temporary conversion rate dip on product pages and paid search inefficiency; apply a conservative window (e.g., 1–2 weeks).
Penalties: Many retailers assess short/late shipment chargebacks. Typical structures include a percentage of the invoice (e.g., low single-digit percent) or a per-order fee. Use your contract terms to set pen.
Use scenario bands (low/base/high) to avoid false precision.
Quick estimator when data is limited
When data is sparse, approximate:
Stockout cost per day ≈ daily revenue × gross margin rate × at-risk share × unrecovered share + daily penalties + CS/time cost.
Where at-risk share is the portion of traffic/orders unable to substitute, and unrecovered share is 1 − r.
Example:
Daily revenue $10,000, margin 40%, at-risk 60%, unrecovered 70%, penalties $0, service $100/day.
Cost/day ≈ 10,000 × 0.4 × 0.6 × 0.7 + 0 + 100 = $1,780.
Service levels and risk trade-offs
Probability of stockout and CSL vs fill rate
Cycle service level (CSL): Probability you do not stock out in a replenishment cycle.
Fill rate (FR): Proportion of demand fulfilled immediately.
They are related but not equal. You can have high CSL with modest FR if shortages, when they occur, are large. Choose targets per ABC class and channel.
Typical target ranges:
A items: CSL 98–99.5%; FR 97–99%+
B items: CSL 95–98%; FR 95–98%
C items: CSL 90–95%; FR 92–96%
Safety stock with demand and lead-time variability
If demand ~N(μd, σd) per period and lead time L is constant:
Safety stock = z × σd × √L
If lead time varies with mean μL and std dev σL:
Safety stock = z × √(μL × σd² + μd² × σL²)
Choose z from CSL target (e.g., z ≈ 1.65 for 95% CSL; 2.33 for 99% CSL).
Reorder point:
ROP = expected demand during lead time + safety stock = μd × μL + SS
Newsvendor intuition and cost curves
Use the critical fractile to set CSL when you can price the trade-off:
Critical fractile = Cu / (Cu + Co)
Cu (underage cost) ≈ stockout cost per unit (lost margin + penalties)
Co (overage cost) ≈ holding cost per unit (capital, storage, obsolescence)
Map CSL to z and compute holding vs stockout costs to visualize the efficient frontier. For items with high Cu and low Co, target higher CSL.
Measuring impact and KPIs
OOS rate, fill rate, OTIF, backorder rate
OOS rate: time- or opportunity-weighted percentage of demand windows where on-hand is zero.
Fill rate: shipped or picked immediately ÷ ordered demand.
OTIF: on-time, in-full compliance to customer/retailer commitments.
Backorder rate: backordered units ÷ ordered units.
Lost sales rate: unfulfilled and unrecovered demand ÷ total demand.
Track by SKU-location-channel and roll up to category and vendor.
Data sources and instrumentation
Order history, shipment lines, and cancellation reasons from your OMS/ERP.
On-hand, receipts, and adjustments from WMS/ERP.
POS and shelf scans (retail) for on-shelf availability.
Supplier ASNs and lead-time actuals for variability.
Web analytics (ecommerce) for OOS page events and conversion impacts.
For deeper guides and playbooks, see our inventory management blog.
Target ranges and alert thresholds
Alerts when daily OOS rate > 2% for A items or > 5% for B items.
Escalate when fill rate dips below 97% for A items or 95% overall.
Flag suppliers with lead-time CV (σL/μL) above 0.3 for mitigation.
Trigger review if actual penalties exceed 0.5% of monthly revenue.
Set thresholds by business cycle and refine quarterly.
Strategies to reduce stockouts
Forecasting improvements and bias control
Monitor MAPE and forecast bias (mean percentage error) by SKU-class.
Use hierarchical forecasting to borrow strength from category and seasonality.
Separate baseline from promotion/price uplift; build event libraries.
For new SKUs, use analogs and cap uncertainty with conservative CSL early on.
ABC-XYZ policies and dynamic safety stock
ABC by contribution; XYZ by demand variability.
Assign CSL and review cadence by class (e.g., A/X highest CSL and weekly tune).
Recalculate safety stock dynamically as σd and σL change.
Apply demand censoring corrections during OOS periods to avoid underestimating demand.
Reorder points, MOQ, and buffer placement
Compute ROP = μd × μL + SS; round to pack sizes logically, not mechanically.
Challenge MOQs that cause long cover spikes and subsequent OOS.
Place buffers where variability is greatest or postponement is feasible.
Multi-echelon: position safety stock upstream to pool risk, with downstream minimums for shelf availability.
Supplier reliability and lead-time reduction
Score suppliers on OTIF, lead-time variability, and quality.
Use dual-sourcing for critical A/X items.
Reduce lead-time variability with better ASNs, slotting, and capacity reservations.
Consider safety time when SS is impractical due to constraints.
Omnichannel allocation and substitutions
Distinguish on-shelf OOS vs DC OOS; fix shelf replenishment and scanning to protect sales.
Allocation rules: protect ecommerce or key accounts with service-level reservations.
Enable smart substitutions with affinity rules; measure recovery vs margin dilution.
For partial fulfillment, prefer consolidating shipments when delays are short; otherwise communicate ETAs early to avoid cancellations.
To evaluate methods and templates that support these workflows, explore our free inventory planning tools.
Scenario planning and sensitivity analysis
Stress tests for demand and lead-time shocks
Demand surge tests: +20%, +50% for 2–4 weeks.
Lead time shocks: +2× average and +σL to +3σL scenarios.
Supplier disruption: 0 receipts for 14–30 days; evaluate alternative sourcing.
What-if cost curves and breakeven analysis
Plot expected stockout cost vs CSL using your Cu/Co assumptions.
Identify breakeven CSL where marginal holding cost equals marginal stockout reduction.
Run channel-specific curves; ecommerce often warrants higher CSL due to immediacy and penalties.
Contingency playbooks
Expedite triggers by SKU-class and order window.
Allocation switches during shortages (e.g., protect top 20% accounts).
Substitution activation and messaging templates.
Price/assortment levers to modulate demand during constrained periods.
Implementation roadmap
30/60/90-day plan
Day 0–30:
Baseline KPIs: OOS rate, fill rate, OTIF, backorder rate by SKU-location.
Clean data: remove phantom stock, mark OOS-censored periods.
Classify ABC-XYZ and set initial CSL targets.
Identify top 50 SKUs by stockout cost exposure.
Day 31–60:
Implement dynamic safety stock and recalculated ROPs.
Tune forecasts; add promotion overlays and bias controls.
Launch supplier scorecards and lead-time measurement.
Pilot omnichannel allocation and substitutions on high-impact SKUs.
Day 61–90:
Automate parameter updates weekly; add exception-based alerts.
Expand to long-tail SKUs with simplified policies (min/max).
Institutionalize monthly S&OP review of CSL, penalties, and cost curves.
Document contingency playbooks and drill twice per quarter.
Roles, governance, and change management
Roles: demand planner, supply planner, merchandising, operations, and IT/data.
Governance: monthly parameter review, quarterly CSL reset by ABC-XYZ, and supplier summits.
Change: start with a narrow pilot, publish wins, and standardize.
Integrations (ERP/WMS/forecasting) and automation
Integrate order, inventory, and supplier data to refresh μd, σd, μL, σL.
Automate ROP/SS calculation and push to ERP/WMS; log overrides.
Add OMS rules for allocation, backorders, and substitutions with ETA logic.
Create alert thresholds for OOS risk, lead-time drift, and penalty exposure.
Examples and benchmarks
Retail and CPG
Example:
Grocery promotion misses forecast by 25%. Two-day on-shelf OOS on a high-velocity SKU.
Daily demand 1,200 units; margin $0.60; limited substitution due to brand loyalty.
Event cost estimate:
Qshort ≈ 2,400 units; recovery r = 0.25; lost units = 1,800
Lost margin = 1,800 × $0.60 = $1,080
Shelf execution cost and labor: $300
Potential retailer penalties for short ship vary by contract; include a placeholder if applicable.
Prevention: higher CSL for promo weeks, store execution checks, and DC-to-store allocation guardrails.
Benchmarks:
Many retailers target on-shelf availability above 95%, with top items higher.
Chargebacks can be a percentage-based or fixed-fee structure; verify contracts to model pen accurately.
Manufacturing and MRO
Example:
MRO spare part with low demand (5/week), long lead time (6 weeks), high downtime cost if OOS.
Underage cost Cu includes downtime and service penalties; overage cost Co is moderate holding.
Critical fractile suggests very high CSL (e.g., 99%+), justifying higher safety stock or vendor stocking agreements.
Benchmarks:
Fill rate expectations for critical spares commonly exceed 98% due to downtime risk.
Dual-sourcing and consignment inventory can materially reduce effective stockout cost.
D2C ecommerce
Example:
Fast-moving accessory SKU, lead time 10 days with σL = 3 days, margin $8.
Safety stock with μd = 70/day and σd = 20/day:
SS ≈ z × √(μL × σd² + μd² × σL²)
For 98% CSL (z ≈ 2.05): SS ≈ 2.05 × √(10×400 + 4900×9) ≈ 2.05 × √(4,000 + 44,100) ≈ 2.05 × √48,100 ≈ 2.05 × 219.4 ≈ 449 units
ROP = μd × μL + SS = 70×10 + 449 = 1,149 units.
If you operate on Shopify, consider how platform-native data and order timing affect forecasting and replenishment; tools that surface short-term risk can improve availability; such as the Verve AI Shopify inventory forecasting.
WooCommerce stores often see long-tail variability and batchy supplier MOQs; parameter automation and exception alerts help target the right SKUs. See our WooCommerce inventory forecasting plugin to help with these issues.
Soft next step
Run the quick estimator on your top 20 SKUs to quantify exposure, pick target CSLs by ABC-XYZ, then tune safety stocks and ROPs. Review impact after two replenishment cycles and adjust with scenario tests.
FAQs
How do I estimate stockout cost when I have limited data?
Start with the quick estimator: daily revenue × margin rate × at-risk share × unrecovered share + penalties + service cost.
Use conservative ranges for recovery (e.g., 20–40%) and at-risk share (50–80%).
Validate with one or two real events to calibrate assumptions.
What is the difference between lost sales and backorder cost models?
Lost sales: Revenue and margin are forfeited; focus on units lost × margin plus indirect effects.
Backorder: Revenue is delayed; add handling, expediting, discounting, and potential cancellations. Many businesses use a hybrid with a recovery rate.
Is fill rate the same as service level?
No. CSL is the probability of no stockout in a cycle; fill rate is the percentage of demand fulfilled immediately. They move together but are not identical.
What service level targets make sense by ABC class?
A items: CSL 98–99.5%
B items: CSL 95–98%
C items: CSL 90–95%
Adjust by channel and penalty exposure; ecommerce and key accounts may warrant higher CSL.
How do promotions and seasonality change stockout costs?
Higher uplift increases Cu (underage cost), justifying elevated CSL and safety stock.
Use event-based forecasts, pre-build inventory, and allocate to protect promotional windows.
How do substitutes affect stockout cost calculations?
Substitutes reduce unrecovered demand. Estimate recovery via observed cross-SKU substitution.
Include margin dilution if substitutes carry lower margin or require discounts.
How do I set safety stock when lead time is highly variable?
Use SS = z × √(μL × σd² + μd² × σL²).
If σL is large, prioritize lead-time reduction, dual-sourcing, or safety time in addition to SS.
How often should I recalculate stockout costs and inventory parameters?
Recompute safety stock and ROPs at least monthly for A/B items; quarterly for C items.
Re-estimate cost assumptions and CSL targets quarterly or after major disruptions.