Measurement is the prerequisite for improvement. Supply chain leaders who cannot precisely define, calculate, and contextualize their key performance indicators are operating by feel in an environment that demands precision. This article defines the twelve essential supply chain KPIs, explains the business logic behind each one, provides exact formulas, establishes industry benchmarks, and explains how each metric connects to diagnostic and corrective action.
These metrics are not a complete catalog - supply chains generate hundreds of measurable data points. They are the twelve that consistently differentiate high-performing supply chains from average ones, and the twelve that experienced supply chain leaders monitor first when diagnosing performance problems.
For guidance on how to display these metrics effectively, see Dashboards and Reporting. For information on the data systems that produce these figures, see Data Sources.
1. Inventory Turnover
Inventory Turnover measures how many times a company sells and replaces its inventory over a given period. It is the single most important indicator of inventory efficiency and working capital utilization.
Formula:
Inventory Turnover = Cost of Goods Sold / Average Inventory ValueWhere Average Inventory Value = (Beginning Inventory + Ending Inventory) / 2
Interpretation: A turnover of 8 means the company cycles through its entire inventory eight times per year, or approximately every 45 days. Higher turns generally indicate better efficiency, lower carrying costs, and less risk of obsolescence - but turns that are too high can signal stockout risk.
Benchmarks by industry:
- Grocery and food distribution: 12-25x
- Consumer electronics: 6-12x
- Apparel and footwear: 4-8x
- Industrial manufacturing: 3-6x
- Automotive parts: 8-15x
Diagnostic value: When turns decline, investigate category by category. A company-wide decline often masks SKU-level or warehouse-level problems. Segment by product family, warehouse location, and customer channel to isolate the root cause. Rising turns with rising stockout rates indicate you may have cut inventory too aggressively.
2. On-Time In-Full (OTIF)
OTIF is the percentage of customer orders delivered at the correct time and in the complete quantity. It is the primary supply chain metric for customer service level and contractual compliance, and many major retailers impose financial penalties for OTIF failures.
Formula:
OTIF % = (Orders Delivered On-Time AND In-Full / Total Orders) × 100Note: OTIF is conjunctive - a delivery that is on time but short-shipped does not qualify. Both conditions must be satisfied simultaneously.
Component metrics:
- On-Time Delivery Rate = (Orders Delivered On or Before Requested Date / Total Orders) × 100
- In-Full Rate = (Orders Shipped Complete / Total Orders) × 100
Benchmarks:
- Best-in-class: 95-99%
- Industry average: 80-90%
- Walmart OTIF penalty threshold: below 95% triggers fines of 3% of the invoice
Diagnostic value: Decompose OTIF failures by root cause - supplier delays, warehouse picking errors, carrier performance, and demand forecast errors each require different remediation. Track OTIF separately by customer segment, distribution center, and carrier to isolate systemic from situational failures.
3. Order Fill Rate
Order Fill Rate measures the percentage of customer demand satisfied from available stock at the time of order, without backordering or substitution.
Formula:
Order Fill Rate = (Units Shipped from Available Stock / Units Ordered) × 100Alternatively calculated at the order line level:
Line Fill Rate = (Order Lines Shipped Complete / Total Order Lines) × 100Distinction from OTIF: Fill Rate measures stock availability at the moment of order; OTIF measures delivery performance against the committed date. Both matter, and they diagnose different problems. A high fill rate with poor OTIF points to logistics execution failures. A low fill rate points to inventory positioning or demand forecasting failures.
Benchmarks:
- Best-in-class: 97-99%
- Industry average: 85-95%
- Below 85% typically indicates systemic inventory positioning problems
Diagnostic value: Segment fill rate by SKU velocity tier (A/B/C analysis), by warehouse, and by customer class. A-tier SKU stockouts represent the highest revenue impact and should trigger immediate root cause analysis. Chronic C-tier shortages may indicate that the SKU portfolio needs rationalization rather than replenishment optimization.
4. Perfect Order Rate
Perfect Order Rate is the strictest customer service metric, measuring the percentage of orders that are delivered on time, in full, damage-free, and with accurate documentation.
Formula:
Perfect Order Rate = (On-Time %) × (In-Full %) × (Damage-Free %) × (Documentation Accurate %) × 100Example calculation:
- On-Time: 96%
- In-Full: 94%
- Damage-Free: 99%
- Documentation Accurate: 98%
- Perfect Order Rate: 0.96 × 0.94 × 0.99 × 0.98 = 87.6%
Why the multiplication matters: Perfect Order Rate exposes the compounding effect of multiple moderate failures. A company with 95% performance on each of four dimensions achieves only an 81% Perfect Order Rate. This forces leadership to recognize that supply chain excellence requires near-perfect execution across all dimensions, not just one.
Benchmarks:
- Best-in-class: 90-98%
- Industry average: 75-90%
Diagnostic value: Use perfect order decomposition to prioritize improvement investment. The component with the lowest individual rate is typically the highest-leverage improvement opportunity. Damage rates that trend upward signal packaging, handling, or carrier problems. Documentation errors often indicate system integration failures between order management and warehouse management systems.
5. Days of Inventory on Hand (DOH)
Days of Inventory on Hand converts inventory value into time units, expressing how many days the current inventory would last at the current rate of sales. It is the most intuitive measure of inventory coverage and risk.
Formula:
Days of Inventory on Hand = (Average Inventory Value / Cost of Goods Sold) × Number of DaysOr equivalently:
DOH = 365 / Inventory TurnoverInterpretation: A DOH of 45 means the current inventory would cover 45 days of demand at current run rates. Optimal DOH varies by supply chain design: a company sourcing from domestic suppliers with short lead times can hold 15-20 days. A company sourcing from overseas with 60-day ocean transit times must hold significantly more.
Benchmarks by industry:
- Fast-moving consumer goods: 15-30 days
- Consumer electronics: 30-60 days
- Industrial equipment: 60-120 days
- Fashion/apparel: 60-90 days (with high seasonality risk)
Diagnostic value: Track DOH at the SKU, category, and warehouse level. Rising DOH in a flat or declining sales environment signals purchasing decisions are disconnected from demand signals. Falling DOH with rising stockout rates signals demand is outpacing replenishment - investigate whether demand forecasts are current and whether lead times have extended.
6. Cash-to-Cash Cycle Time
Cash-to-Cash Cycle Time measures the number of days between when a company pays its suppliers and when it collects payment from its customers. It is the primary supply chain metric for working capital management and directly impacts a company’s financing requirements.
Formula:
Cash-to-Cash Cycle Time = Days Inventory Outstanding + Days Sales Outstanding - Days Payable OutstandingWhere:
- Days Inventory Outstanding (DIO) = (Inventory / COGS) × 365
- Days Sales Outstanding (DSO) = (Accounts Receivable / Revenue) × 365
- Days Payable Outstanding (DPO) = (Accounts Payable / COGS) × 365
Interpretation: A positive cycle time means the company is financing its customers and inventory with its own capital. A negative cycle time (common in retail) means customers pay before the company pays suppliers - a structural working capital advantage. Amazon operates at approximately negative 30 days.
Supply chain levers:
- Reduce DIO: improve inventory turns through better demand forecasting and replenishment
- Extend DPO: negotiate extended supplier payment terms
- Reduce DSO: tighten customer credit terms and collection processes
Benchmarks:
- Best-in-class manufacturers: 30-60 days
- Average manufacturers: 60-100 days
- Best-in-class retail: negative 10 to 0 days
7. Supplier Lead Time
Supplier Lead Time is the elapsed time from purchase order submission to goods receipt. It is the primary determinant of safety stock requirements and procurement planning horizons.
Formula:
Supplier Lead Time = Date of Goods Receipt - Date of Purchase Order SubmissionFor planning purposes, use average and standard deviation, not just average:
Lead Time Variability = Standard Deviation of Lead Time / Average Lead TimeWhy variability matters more than average: Safety stock calculations depend on lead time variability. A supplier with an average lead time of 21 days and a standard deviation of 2 days requires far less safety stock than a supplier with the same average but a standard deviation of 7 days. Supplier lead time reporting that reports only averages systematically understates inventory requirements.
Benchmarks:
- Domestic suppliers: 5-15 days (best-in-class), 15-30 days (average)
- International suppliers: 30-60 days (ocean), 5-10 days (air)
- Target lead time variability coefficient: below 0.2
Diagnostic value: Track lead time trends by supplier and SKU. Rising lead times signal capacity constraints, quality holds, or financial distress at the supplier. Suppliers with deteriorating lead time consistency should be flagged in supplier risk reviews before they cause stockouts.
8. Freight Cost Per Unit
Freight Cost Per Unit normalizes transportation spend to a per-unit metric, enabling meaningful comparison across lanes, carriers, modes, and time periods.
Formula:
Freight Cost Per Unit = Total Freight Cost / Total Units ShippedAlso useful:
Freight Cost as % of Revenue = (Total Freight Cost / Revenue) × 100
Freight Cost Per Pound = Total Freight Cost / Total Pounds ShippedMode benchmarks (approximate, varies by lane):
- Less-than-truckload (LTL): $0.08 - $0.20 per pound
- Full truckload (FTL): $0.04 - $0.10 per pound
- Parcel (small package): $0.30 - $1.50 per pound
- Ocean container: $0.01 - $0.05 per pound (when markets are stable)
Diagnostic value: Rising freight cost per unit can signal mode shift (from FTL to LTL or parcel), declining shipment sizes, lane changes, or carrier rate increases. Decompose by lane, mode, carrier, and shipment size to identify the primary driver. Consistently small LTL shipments that could be consolidated into FTL represent a straightforward cost reduction opportunity.
9. Warehouse Capacity Utilization
Warehouse Capacity Utilization measures the percentage of usable storage space actually occupied by inventory. It is the primary metric for warehouse network planning and operational efficiency.
Formula:
Warehouse Utilization % = (Occupied Storage Locations / Total Available Storage Locations) × 100Or by volume:
Warehouse Utilization % = (Cubic Feet Occupied / Total Usable Cubic Feet) × 100The utilization paradox: The theoretical maximum utilization is 100%, but operational maximum is approximately 85-88%. Beyond 85%, picking efficiency degrades, put-away becomes difficult, safety incidents increase, and the warehouse effectively becomes slower. A warehouse running at 95%+ capacity is operationally stressed even if it looks efficient on paper.
Benchmarks:
- Optimal operating range: 75-85%
- Below 70%: potential for consolidation or subleasing
- Above 85%: operational stress, investigate expansion or SKU rationalization
Diagnostic value: Track utilization by zone, aisle, and storage type (rack, floor, cold storage). Imbalanced utilization - one zone at 95% while another runs at 60% - signals slotting problems that cause picker inefficiency. Seasonal utilization peaks that exceed 90% are predictable capacity crises that should trigger advance planning for overflow solutions.
10. Stockout Rate
Stockout Rate measures the percentage of SKU-location combinations that have zero available inventory during a given period, often while customer demand exists.
Formula:
Stockout Rate = (SKU-Location Combinations with Zero Inventory During Demand Period / Total Active SKU-Location Combinations) × 100An alternative formulation captures lost demand:
Stockout Rate = (Units of Demand Not Fulfilled Due to Zero Inventory / Total Units Demanded) × 100Benchmarks:
- Best-in-class retail: below 2%
- Average retail: 5-8%
- Above 10%: systemic inventory positioning or forecasting failure
Revenue impact calculation:
Lost Revenue from Stockouts = Units of Unmet Demand × Average Selling Price × (1 - Substitution Rate)Where Substitution Rate is the fraction of customers who buy an alternative product rather than abandoning the purchase.
Diagnostic value: Correlate stockout patterns with demand forecast accuracy, supplier lead time variability, and replenishment trigger settings. Frequent stockouts on A-tier SKUs are the highest-priority failure mode. Stockouts that follow a seasonal pattern without corresponding safety stock adjustments indicate a forecasting process that does not incorporate seasonality correctly.
11. Return Rate
Return Rate measures the percentage of sold units that are returned by customers, a critical metric for retail, e-commerce, and consumer goods companies.
Formula:
Return Rate = (Units Returned / Units Sold) × 100Financial return rate:
Return Rate (Value) = (Value of Returned Goods / Total Sales Value) × 100Benchmarks by channel:
- Brick-and-mortar retail: 8-10%
- E-commerce general: 20-30%
- Apparel e-commerce: 30-40%
- Electronics e-commerce: 15-25%
Return reason codes: The analytics value of return rate is multiplied by structured return reason capture. Returns due to “wrong item shipped” are a warehouse accuracy problem. Returns due to “defective product” are a quality problem. Returns due to “did not meet expectations” are a product description or photography problem. Each cause requires a different response.
Diagnostic value: Track return rates by SKU, supplier, warehouse origin, and customer segment. High return rates from specific warehouses indicate picking accuracy problems. High return rates from specific suppliers indicate product quality or specification problems. Rising return rates after product changes indicate the change created customer dissatisfaction.
12. Supply Chain Risk Index
The Supply Chain Risk Index is a composite metric that aggregates multiple risk signals into a single score for monitoring supply chain vulnerability. Unlike the preceding metrics, there is no universal formula - organizations construct it based on their specific risk profile.
Common component dimensions:
- Supplier concentration risk: percentage of spend with single-source suppliers
- Geographic concentration risk: percentage of suppliers in high-risk geographies
- Lead time buffer coverage: safety stock as a multiple of lead time variability
- Financial health of critical suppliers: based on credit ratings or payment behavior signals
- Logistics disruption exposure: percentage of volume on congested or volatile lanes
- Demand forecast error: MAPE (Mean Absolute Percentage Error) across the SKU portfolio
Example composite formula:
Risk Index = w1 × Supplier Concentration Score + w2 × Geographic Risk Score + w3 × Lead Time Buffer Score + w4 × Supplier Financial Health Score + w5 × Logistics Exposure ScoreWhere weights (w1 through w5) are calibrated to the organization’s historical loss experience from each risk category.
Benchmarks: Because this metric is organization-specific, benchmark against your own historical baseline and set threshold alerts for index scores that exceed the 75th percentile of historical readings.
Diagnostic value: The Risk Index is most valuable as a leading indicator - it should rise before disruptions occur, not after. Weekly or monthly tracking enables early intervention: diversifying supplier base before a concentration becomes a crisis, building buffer stock before a predicted logistics disruption, or qualifying alternative suppliers before a financial health signal becomes a failure event.
For guidance on visualizing these KPIs in operational dashboards, see Dashboards and Reporting. To understand the data systems that generate these figures, see Data Sources. For the analytical methods that optimize against these metrics, see Techniques and Models.