Operations KPIs for Supply Chain and Planning Leaders

Operations KPIs show whether a supply chain delivers what customers want, at the right cost, with the right amount of inventory. The challenge is rarely a shortage of measures; most operations teams can produce hundreds. The challenge is choosing a small set that balances service, inventory and cost, so that improving one does not quietly damage another. This guide groups operations KPIs into service, inventory, forecast, supplier, warehouse and transportation, and cost measures, gives formulas, works through forecast accuracy, bias, weeks of supply and perfect order rate with checked arithmetic and offers advice on building a balanced set. It belongs to business analytics for finance and operations, follows KPIs for CFOs and KPI scorecards, and leads to leading vs. lagging indicators and OKRs vs. KPIs. Many of these measures feed the plan in demand planning and inventory optimization and the budget in budgeting and forecasting.

Service measures

On-time in-full (OTIF). The share of order lines delivered complete on the date promised, or within an agreed window. It is the headline measure of reliability. Define it with the customer: on time by order date or ship date, and how much tolerance is allowed.

Fill rate. Quantity shipped from stock divided by quantity ordered. It differs from OTIF because it ignores timing and is measured in units or lines. Choose one and be consistent.

Perfect order rate. The share of orders delivered on time, complete, undamaged and with correct documents. Because it multiplies several rates, it falls quickly. If on-time in-full is 94 percent, damage-free is 99 percent and documentation accuracy is 98 percent, the perfect order rate is 0.94 × 0.99 × 0.98 = 91.2 percent. A set of good-looking individual rates can hide a weaker overall result.

Order cycle time. Time from order to delivery. Track the median and the 90th percentile, as averages hide the late tail.

Inventory measures

Inventory turns. Cost of goods sold divided by average inventory. A company with $1.2 million of cost of goods sold and $300,000 of average inventory turns four times a year.

Days of inventory or weeks of supply. Inventory on hand divided by average demand per period. If 12,000 units are on hand and average weekly demand is 1,500 units, supply is 8 weeks. Compare it with the target cover by item, which depends on lead time and variability.

Stockout rate and stockout days. The share of items out of stock, or the days an item was unavailable. Weight by revenue so that a stockout of a top item counts more.

Excess and obsolete stock. Inventory above a defined cover or older than a set age, as a percentage of total. It shows the write-down risk. If $50,000 of a $200,000 inventory is older than 180 days, the aged share is 25 percent.

Inventory accuracy. The share of locations where system quantity matches the physical count. Low accuracy undermines every planning decision.

Forecast measures

Forecast accuracy. One common approach is weighted absolute percentage error (WAPE): the sum of absolute errors divided by the sum of actuals. Five items (hypothetical):

WAPE = 50 / 300 = 16.7 percent, so accuracy is 83.3 percent. Bias is (forecast - actual) / actual = (310 - 300) / 300 = +3.3 percent, which means the forecast was slightly high overall. A forecast can have low bias and high error (errors cancel) or the reverse, so track both. Measure at the level where decisions are made, such as item and location by lead-time horizon, and compare with a naive baseline. See machine learning for forecasting.

Forecast value added. The change in accuracy from each step in the process (statistical forecast, planner override, sales input) relative to a naive forecast. It shows which steps help and which add noise.

Supplier measures

Supplier on-time delivery. The share of purchase order lines received by the promised date. Supplier lead-time variability. The standard deviation of lead time, which drives safety stock. Quality acceptance rate. The share of received quantity that passes inspection. Purchase price variance. Actual price less standard price, multiplied by quantity. Together they show the reliability of supply and the cost.

Warehouse and transportation

Pick accuracy. Correct picks divided by total picks. Dock-to-stock time. Time from receipt to availability for sale. Order lines per labor hour. A productivity measure to pair with accuracy. Freight cost per unit or per order. Total freight cost divided by units or orders, tracked by lane and mode. Carrier on-time performance. Deliveries on time divided by total.

Cost and cash

Cash-to-cash cycle. Days of inventory plus days of receivables less days of payables. Total supply chain cost as a share of revenue. Includes freight, warehousing, inventory carrying and order processing. Inventory carrying cost. Inventory value times the annual carrying rate. At 20 percent a year, $300,000 of inventory costs $60,000 to hold. Reducing inventory by 10 percent saves $6,000 a year in carrying cost as well as releasing $30,000 of cash. See landed cost.

Building a balanced set

A set of operations KPIs should include at least one measure from each of three groups: service (OTIF, fill rate), inventory (turns, weeks of supply, aged stock) and cost (freight per unit, cash-to-cash cycle). Add forecast accuracy and bias as drivers. Pair measures that push in opposite directions: inventory turns with service level, speed with accuracy, cost with quality. Choose a manageable number: for a planning leader, perhaps eight to twelve KPIs with a handful at the top.

Data and definitions

Many operations KPIs are disputed because definitions differ. Agree on how to count partial shipments, backorders, cancelled lines and customer-requested delays. Document the measurement point for lead times (order date, confirmation date, ship date). Take data from system transactions, not manual logs, and reconcile counts regularly.

Common mistakes

• Measuring only cost or only service.
• Reporting averages that hide variability.
• Measuring forecast accuracy at the wrong level.
• Ignoring bias.
• Setting turns targets without service targets.
• Using different definitions across sites.

FAQs

What are the most important operations KPIs?

On-time in-full delivery, inventory turns, forecast accuracy and bias, supplier on-time delivery and total supply chain cost, balanced so that service, inventory and cost are all covered.

How do I calculate forecast accuracy?

Sum the absolute errors across items and divide by the sum of actual demand to get the weighted absolute percentage error. Accuracy is one minus that figure.

What is the perfect order rate?

The share of orders that are on time, complete, undamaged and correctly documented. It is the product of those individual rates.

How many KPIs should a planning team track?

A handful at the top, perhaps five to eight, with supporting measures underneath for diagnosis.

Why pair inventory turns with service level?

Because turns can be improved by cutting stock, which may hurt service. The pair keeps the trade-off visible.