Sign in

Blog · Board and management reporting · Supply chain

KPIs for supply chain teams: ten measures that matter, each with its formula and the export it comes from

The ten KPIs a supply chain function should run on, each with its formula, the export it comes from and what it tells you: supplier OTIF against each supplier's own baseline, lead time variability, customer OTIF under each customer's rule, forecast bias by SKU and site, inventory ageing, single-source exposure, days of cover against target, expedite cost, supplier deterioration watch, and cause of service misses. Also the three measures most teams miss, the figures to drop, the identities, and who owns what.

The short answerA supply chain function should run on ten measures: supplier OTIF against each supplier's own baseline; lead time variability per supplier and site; customer OTIF under each customer's own rule; demand forecast bias by SKU and site; inventory ageing by site; single-source exposure by part and site; days of cover against target; expedite cost by cause; a watch list of suppliers deteriorating on any measure for two consecutive months; and the cause split of service misses. They come from the purchase order and receipt files, the order and shipment files, the forecast and actuals, the inventory file, the supplier master and the freight ledger. The three most often missed are lead time variability, because average lead time sets the plan and variability sets the safety stock; forecast bias, because accuracy hides a plan that is always high in one place; and the supplier's own baseline, because a group benchmark lists the supplier that has not changed and misses the one that has.

A supply chain is a set of promises between suppliers, sites and customers. The measures that matter show which promises are slipping against their own history, how much uncertainty the inventory is paying for, and why the misses happen.

The ten measures

# Measure Formula Export What it tells you
1 Supplier OTIF against own baseline Receipts on time and in full ÷ receipts due, per supplier and site, against the supplier's trailing 12 months Purchase order and receipt files Suppliers that have changed
2 Lead time variability Interquartile range, and 90th percentile, of actual lead time, per supplier and site Purchase order and receipt dates The number behind the safety stock
3 Customer OTIF, customer's rule Orders on time and in full by the customer's definition ÷ orders due Order and shipment files; customer terms The score the customer sees
4 Forecast bias Mean of (forecast − actual) ÷ actual, by SKU and site Forecast snapshots; actuals Plans that are always high or low in one place
5 Inventory ageing Stock value with no movement in 90, 180, 365 days, by site Inventory file; movement history Cash in stock that is not moving
6 Single-source exposure Spend, and parts, with one qualified supplier ÷ total, by site Supplier master; purchase orders Where one failure stops a line
7 Days of cover against target Stock ÷ average daily demand, by SKU class, against target Inventory; demand Too much and too little, at once
8 Expedite cost by cause Premium freight and expedite fees, by cause: supplier late, forecast miss, order change Freight ledger; reason codes What unreliability costs
9 Supplier deterioration watch Suppliers worse than own baseline by more than the margin for two consecutive months, ranked by spend Receipts; quality file The call list for procurement
10 Cause of service misses Customer OTIF misses by cause: stock-out, pick, carrier, order data, customer date Shipment and warehouse files Who owns each fix

Every one of these is computed per account, per site and supplier, and in total, and every one carries an identity that must hold before the table is shown.

The three most supply chain teams miss

Lead time variability. Scorecards show the average. Planners live with the spread.

Forecast bias. Accuracy is reported as one percentage and the bias inside it cancels out.

The supplier's own baseline. One target for all suppliers is simple to explain and wrong for most of them.

A worked line

A site reports forecast accuracy of 82 percent. By SKU, thirty items are forecast an average of 18 percent high every month at this site and nowhere else. Those thirty hold $640,000 of the site's $1.1 million of stock aged over 180 days. The accuracy figure gave no hint, because other items were forecast low by a similar amount.

What to drop

Average lead time, alone. Always with the spread.

Forecast accuracy as one number. Bias by SKU and site is the actionable form.

Supplier OTIF against a single group target. Use each supplier's own history, with the group figure for context.

The identities

Table Must hold
Receipts Receipts due = on time and in full + late only + short only + both
Customer OTIF Orders = OTIF + late only + short only + both
Inventory Opening stock + receipts − issues ± adjustments = closing stock; aged bands sum to total
Expedite Expedite cost by cause sums to the freight ledger premium lines

A table whose identity fails is a table with a row missing or counted twice. It is not shown until it is fixed.

Who owns what

Measure Owner Reviewed
Supplier deterioration watch; supplier OTIF Procurement and supplier quality Monthly
Customer OTIF; cause of misses; expedite cost Operations director; site heads Weekly to monthly
Forecast bias; days of cover; ageing Demand and supply planning Monthly
Lead time variability; single-source exposure Chief operating officer Quarterly

A measure with no owner is a metric, not a KPI; see KPI versus metric versus measure.

Go deeper

The short version

Ten measures from receipts, shipments, forecasts and the inventory file. Compare each supplier with itself, read the spread and the bias, and split every miss by cause. Covirage computes all of them from the exports supply chain teams already produce, files only, with the definitions stated and the identities checked. See Covirage for supply chain teams.

Questions people ask

Why variability and not average lead time?

Two suppliers with a 20-day average are different if one delivers in 18 to 22 days and the other in 10 to 35. The second needs far more safety stock. The spread, as an interquartile range or a percentile, is the number behind the safety stock calculation, and it is rarely on the supplier scorecard.

What is forecast bias?

The signed error: forecast minus actual, averaged over periods. Accuracy treats a miss of plus ten and minus ten alike. Bias shows a SKU forecast 15 percent high at one site every month, which builds ageing stock there. It is correctable once seen, and it cancels out in any aggregate accuracy figure.

Why measure a supplier against its own baseline?

Because suppliers differ by category and geography. One that has always shipped 85 percent on time, and still does, has not changed, and the price reflects it. One that shipped 99 and now ships 94 has changed. A group average flags the first and praises the second.