A supplier performance scorecard helps a retailer judge what happens after the purchase order is placed. Initial verification remains important, but a verified supplier can still deliver late, send incomplete quantities or take too long to resolve a problem. Tracking actual results makes the next purchasing decision more informed than relying on memory or the lowest quotation.
This guide sets out a practical scorecard for a small wholesale buying operation. It covers metric definitions, evidence, sample weights and review conversations. The numbers are hypothetical examples, not industry benchmarks or ratings of any real supplier. Use the framework to improve decisions and supplier relationships, while adjusting the measures to your product category and operational needs.
Start with the decisions the scorecard should support
Decide how the scorecard will be used before collecting numbers. You may want to choose a preferred supplier, allocate the next order, identify a recurring quality problem or decide when a backup source is needed. Each purpose requires slightly different evidence. A single total score cannot answer every purchasing question, especially when one failure can have a much larger impact than several minor delays.
Separate initial checks from ongoing performance. Identity, relevant documents, product suitability and commercial terms belong in onboarding. Delivery reliability and claim resolution require actual trading history. Rawhub’s supplier verification guide provides a starting checklist for a first order. The scorecard then adds evidence from completed transactions rather than replacing those initial checks.
Define on-time delivery before measuring it
Choose whether the promise refers to dispatch or arrival. A supplier may dispatch on time while the goods reach the store after the required date. Both facts can be useful, but they are different metrics. Record the original confirmed date, any agreed revision and the actual event date. Avoid rewriting the original promise after a delay just to make the score appear better.
For an order-based measure, on-time delivery rate equals orders arriving within the agreed window divided by eligible delivered orders, multiplied by one hundred. If eighteen of twenty comparable orders arrive on time, the rate is ninety percent. Define how early arrivals, split shipments and buyer-requested delays are handled. Apply the same rule throughout the review period so comparisons remain meaningful.
Track order completeness with a separate fill rate
An order that arrives promptly may still leave the shelf empty if key lines are missing. Fill rate can be measured by units, order lines or complete orders. Pick the method that answers your operational question and name it clearly. A unit-based rate should not be compared directly with another supplier’s line-based rate as if the two figures mean the same thing.
Suppose an order requests one hundred units and ninety-two acceptable units arrive in the first shipment. The illustrative first-shipment unit fill rate is ninety-two percent. If the remaining eight arrive later, record that separately rather than erasing the initial shortage. Also capture which items were missing. Losing a fast-moving core SKU may matter more commercially than the same unit shortage in a slow-moving optional line.
Measure quality using accepted and rejected units
Define what counts as a defect for each product category. The definition may involve damage, incorrect variant, unacceptable workmanship or a mismatch with agreed specifications. Inspect consistently and retain appropriate evidence. Record the delivery reference, affected quantity and reason. Without a clear standard, one employee may reject goods that another accepts, producing an unreliable supplier comparison.
A simple acceptance rate is accepted inspected units divided by total inspected units, multiplied by one hundred. State whether every unit was checked or only a sample. Do not present a sample inspection as complete inspection. For critical concerns, escalate the issue immediately through the relevant business process; a high average score should never be used to ignore a serious product problem.
Compare the real buying cost
Invoice price is only one part of supplier performance. Freight, additional handling, shortages, rework and agreed credits can affect what the buyer ultimately pays for usable stock. Use a consistent cost basis across suppliers. Rawhub’s landed-cost guide explains why the headline unit rate can be a poor comparison when delivery and other purchasing costs differ.
Keep recoverable claims distinct from confirmed losses. If a supplier has agreed a credit but has not issued it, record an open claim instead of assuming the money has arrived. Avoid adding speculative lost sales to a precise cost total unless you can explain the assumptions. It can be better to report service disruption separately than to create an impressive-looking cost figure from weak evidence.
Include response and resolution times
Measure more than the speed of the first reply. A quick acknowledgement followed by weeks without action does not resolve a shortage. Record the date an issue was raised, the first useful response, the agreed action and the closure date. Define closure as a verified outcome, such as replacement receipt or an issued credit, rather than simply a promise to investigate.
Categorise issues so urgent and routine requests are not mixed unfairly. A missing invoice copy is different from an incorrect bulk shipment. Where the buyer has not supplied necessary evidence, record that dependency as well. The goal is an accurate account of the process. Fair records make the review conversation more productive and help both sides identify the step that actually caused delay.
Build a transparent weighted score
For an illustrative scorecard, assign thirty percent to delivery reliability, thirty percent to quality, twenty percent to fill rate, ten percent to cost performance and ten percent to issue resolution. These weights are a starting example, not a recommended standard for every business. A perishable category or a product with strict specifications may need a different emphasis and separate acceptance gates.
Convert each metric to a clearly defined score on the same scale before applying weights. If delivery scores ninety, quality ninety-five, fill rate eighty-five, cost eighty and resolution seventy, the weighted result is eighty-eight point five out of one hundred. Keep the component scores visible. Two suppliers can have the same total while one has a quality problem and the other has a delivery problem.
Do not let averages hide critical failures
Set a few review triggers outside the weighted total. Repeated wrong products, unresolved serious complaints or a material mismatch with agreed specifications may require immediate action regardless of the average. Document who reviews such events and what evidence they need. The scorecard should support responsible judgement rather than act as an automatic approval system.
Also show the volume behind each score. A supplier with two successful orders has a different evidence base from one with two hundred. Mark new suppliers as having limited history and avoid ranking them confidently against established partners. Review trends over time as well as the latest month. One unusual event deserves investigation, while a repeated pattern usually requires a more structured corrective plan.
Keep a simple evidence table
Record supplier name or internal reference, purchase order number, product line, promised date, received date, ordered quantity, accepted quantity and issue status. Link supporting documents through your normal controlled filing process. Limit access to people who need the information. Avoid mixing private employee details or unrelated customer information into a purchasing performance sheet.
Assign an owner for each data field. Receiving staff may record quantities and condition, purchasing may record agreed dates, and accounts may confirm credits. Reconcile obvious inconsistencies before the monthly review. If the same shipment appears twice, the supplier’s score and your stock figures can both become misleading. A short, reliable table is more useful than an elaborate dashboard fed by incomplete records.
Hold a review that produces an action
Share the relevant evidence with the supplier and allow factual corrections. Focus on specific orders and definitions rather than broad claims such as “your service is always poor.” Ask what caused the issue, which corrective action is proposed and when you should expect evidence of improvement. Confirm the action in the normal business record so the next review has a clear reference point.
An illustrative action could be: “For the next four dispatches, confirm packed quantities by SKU before pickup and flag any shortage before invoicing.” This is more measurable than asking the supplier to be careful. At the next review, compare results against the agreed action. Recognise improvements as well as failures; a fair process encourages useful cooperation instead of defensive reporting.
Use the scorecard to plan supply continuity
A strong score does not eliminate concentration risk. If one supplier provides every critical item, an unexpected disruption can still affect the store. Identify which products need a backup source and what qualification that source requires. Test alternatives through a controlled order rather than waiting until the main supplier fails during a busy selling period.
When allocating orders, consider capacity, product fit, landed cost and recent performance together. Do not move all volume based on a single small monthly score change. Discuss the commercial impact of changes with the relevant team. Buyers comparing options can use structured supplier bid comparisons alongside the scorecard, because a quotation and actual execution history answer different questions.
Introduce the process in thirty days
During the first week, choose a small group of active suppliers and agree the metric definitions internally. In the second week, collect recent order evidence and identify missing fields. During the third week, calculate sample results and check them against the underlying records. Correct definitions that produce misleading outcomes before presenting a total score as authoritative.
In the fourth week, hold one review and assign a limited number of actions. Keep the first version manageable. Add complexity only when it changes a decision or reveals a problem the current measures miss. A supplier performance scorecard succeeds when it improves purchasing choices, reduces repeated mistakes and gives both parties a clear record of what needs to happen next.
Frequently asked questions
How often should a supplier scorecard be reviewed? Monthly reviews can suit regular purchasing, while low-volume categories may need a longer window to collect meaningful evidence. Serious issues should be addressed when they occur. Do not wait for the next reporting date merely because a formal review has been scheduled.
Can the lowest-price supplier still be the best choice? Yes, if the product, delivery and service meet the business’s needs at a competitive total cost. The scorecard does not penalise low prices. It makes other relevant outcomes visible so a purchasing decision is based on more than one attractive number.
Is specialist software necessary? Not for a small operation with manageable order volume. A well-maintained spreadsheet and organised evidence can support the process. Consider software when data volume, multiple locations or approval requirements justify it, and evaluate whether it can preserve your metric definitions and audit trail.
