On-time delivery depends on which promise you measure
For Procurement heads, buyers and materials planners · Understand the problem
TL;DR — what you’ll learn
- Retain the original requested date and the supplier's confirmed commitment.
- Define whether performance is measured by order, line, receipt or quantity.
- Separate delivery, completeness, quality and response measures.
- Use the evidence to agree a specific improvement action rather than rely on a composite ranking.
The supplier's on-time score is 98%. Production still experiences shortages because the supplier repeatedly moves its promised dates and the report measures only the latest promise. The calculation is internally consistent. Its definition has stopped measuring the commitment the factory originally needed.
What should a food supplier performance scorecard measure beyond a single delivery percentage?
A supplier-performance scorecard should state the promise, event and denominator behind each measure. Requested-date performance and confirmed-date performance answer different questions. Keep both where useful, with quantity completeness and unresolved changes visible, so the review can distinguish planning gaps from missed commitments.
- 01Define the measuresPromise, receipt and quality scope
- 02Assemble the eventsAccepted changes and actual outcomes
- 03Review the denominatorMissing evidence and exclusionsIs the comparison defensible?YesDiscuss the findingsNoResolve inconsistent definitions before publishing a ranking.
- 04Discuss the findingsAgree corrective follow-up
Illustrative information flow. The article explains the evidence, exceptions and responsible decisions. Diagram: NextEdge Labs.
Decide which commitment the measure represents
The requested receipt date reflects the buyer's need. The supplier's first confirmed date reflects its accepted commitment. Later revisions show how that commitment changed. Preserve these dates rather than overwrite them. A useful report can show performance against more than one basis, clearly labelled.
The actual event also needs a definition: arrival at gate, completed receiving or another agreed point. Do not attribute internal release delays to supplier delivery unless the measure explicitly includes that responsibility. Use separate stage evidence when the business needs to investigate material readiness.
Choose the denominator before comparing suppliers
An order containing ten lines can be mostly delivered while one production-critical line remains missing. An order-level measure, line-level measure and quantity fill rate will describe that situation differently. State the rule and show the supporting count.
For an illustrative period with twenty due order lines, eighteen meet the defined date and complete-quantity condition. The line-level result is 90%. If one failed line is a critical ingredient, the consequence still needs separate visibility. The percentage cannot express every operational effect.
| Measure | Question |
|---|---|
| Requested-date performance | Did supply meet the date the business originally needed? |
| Confirmed-date performance | Did the supplier meet the defined accepted commitment? |
| Quantity completeness | Was the required quantity received under the stated rule? |
| Confirmation responsiveness | How quickly did an actionable reply arrive? |
Keep quality outcomes connected but distinct
A shipment can arrive on time and contain material that cannot be used under the approved quality process. Keep that outcome visible without mixing it into a vague delivery score. The quality record needs its own evidence, scope and investigation state.
Frappe's supplier-scorecard documentation illustrates criteria, weights and evaluation periods. Those features can organise a report, but the business still needs defensible definitions and source records. A weighted total should not replace the detail a buyer needs to understand a missed requirement.
Source: Frappe / ERPNext · Evaluation periods, criteria and weighting. Reference details
Show changed commitments and small samples
A supplier with only two deliveries has a different observation base from one with two hundred. Display counts with rates and avoid treating a small sample as a settled ranking. Show product mix, delivery frequency and any changes in measurement practice that affect comparability.
Track commitment revisions as their own information. Several revisions may indicate uncertainty even when the final delivery meets the latest date. Review whether changes were communicated early enough for the factory to respond. That can be more useful than arguing about one composite score.
Turn the scorecard into a purchasing review
Choose the repeated, evidenced issue and identify the operational response. It may involve earlier confirmation, smaller delivery lots, a different receiving arrangement or a qualified alternative source. Keep the agreed action, owner and review point linked to the relevant events through the organisation's normal supplier-management process.
Do not automatically block a supplier or cancel an order because a new scoring formula produces a low number. Those decisions need the business's approved authority and context. The report should support a review that can be explained from the evidence.
Start by reproducing the current score
Take a sample of recent due lines and reconstruct requested, confirmed and actual events. Compare the calculation with the current dashboard. Investigate missing dates, overwritten commitments and partial deliveries before selecting a new target.
Automation can maintain the event account and assemble the review pack. AI can classify correspondence or summarise recurring causes for review, with source links. The result should make the next supplier conversation more specific and reduce the time spent disputing the numbers.
Summary
A supplier-performance measure is only as useful as the promise and event it represents. Retained commitments, clear denominators and separate quality evidence keep the scorecard interpretable. Use it to identify an owned improvement action and review whether the underlying service changed.
What to do next
- Define the commitment date, actual event and denominator for the current delivery measure.
- Recalculate a sample including changed dates and partial receipts.
- Agree one evidence-based improvement action and retain its review outcome.
Recognise this problem in your operation?
Tell us where the records or handovers break down. We can discuss the process, data and software questions worth examining first.
Sources and further reading
Sources checked 15 October 2026. The worked examples and software evaluation suggestions are NextEdge Labs’ editorial analysis.
- Supplier ScorecardFrappe / ERPNext · Online product documentation, accessed October 2026
Product example of evaluation periods, criteria and weights. The article's metric definitions and numerical scenarios are original examples, not supplier ratings.
Visual credits
AI-generated illustration created for NextEdge Labs; it does not depict a client facility. Workflow diagram by NextEdge Labs. Icons from Lucide (ISC licence).
