A supplier scorecard is a structured way to rate vendors against the same criteria. You choose criteria such as quality, delivery and cost, give each a weight, score each supplier from 0 to 100, and add up the weighted scores into one comparable number that supports sourcing and review decisions.
How is a weighted supplier score calculated?
Multiply each criterion score by its weight, then add the results. Weights must total 100 percent.
Weighted score = sum of (weight x score) / 100, across all criteria, with weights adding up to 100.
Each criterion gets a score from 0 to 100, where higher is better. For cost, higher means better value. For compliance and risk, higher means lower risk. The supplier scorecard calculator applies this formula to five criteria and compares two suppliers side by side.
Which criteria should a scorecard include?
Most scorecards cover five areas. Keep the list short enough that each criterion can be measured with real data.
| Criterion | What it captures | Typical evidence |
|---|---|---|
| Quality | Product meets specification | Defect or rejection rate, returns, corrective actions |
| Delivery | Goods arrive on the agreed date and in full | On-time and in-full percentage, lead time variation |
| Cost | Price and total cost of ownership | Landed cost, payment terms, price stability |
| Responsiveness | Communication and problem solving | Response times, quality of corrective action |
| Compliance and risk | Certifications, financial health, ethics | Audit results, certificates, financial checks |
Add criteria that matter to your business, such as sustainability or innovation, only if you can score them consistently. Delivery performance can be measured with the fill rate and OTIF calculator, and a fair cost comparison should use total landed cost, which the landed cost calculator estimates.
How do you set the weights?
Base weights on what a failure would cost you. If a late delivery stops a production line, delivery should weigh more than responsiveness. If a defect creates safety or recall risk, quality should carry the most.
Agree the weights with the people who depend on the supplier, such as operations, quality, finance and the buyer, and set them before looking at any supplier’s results. Weights chosen after seeing the outcome tend to justify a favourite rather than measure the supplier. Different categories can have different weights: a commodity item may weigh cost heavily, while a critical component weighs quality and compliance more.
Worked example
These figures are illustrative. Weights are quality 35, delivery 25, cost 20, responsiveness 10 and compliance 10. Supplier X scores 90, 80, 70, 60 and 100. Supplier Y scores 80, 90, 85, 80 and 70.
- Supplier X = 0.35 x 90 + 0.25 x 80 + 0.20 x 70 + 0.10 x 60 + 0.10 x 100 = 31.5 + 20 + 14 + 6 + 10 = 81.5.
- Supplier Y = 0.35 x 80 + 0.25 x 90 + 0.20 x 85 + 0.10 x 80 + 0.10 x 70 = 28 + 22.5 + 17 + 8 + 7 = 82.5.
- Y leads by one point.
Now test the weights. Move them to quality 45, delivery 20, cost 15, responsiveness 10 and compliance 10.
- Supplier X = 40.5 + 16 + 10.5 + 6 + 10 = 83.0.
- Supplier Y = 36 + 18 + 12.75 + 8 + 7 = 81.75.
The ranking flipped. A one-point gap is not a decision, and the example shows why: the result depends on how much quality matters compared with delivery and cost. When one supplier wins under every reasonable weighting, the choice is solid. When the lead changes with the weights, the scorecard is telling you the suppliers are close and other factors should decide.
How do you turn raw data into 0 to 100 scores?
Convert each measure to the same scale using a rule written down before scoring. Without one, two reviewers will score the same supplier differently.
- Percentages such as on-time delivery: use the percentage directly, or map ranges to scores, with a band for each level of performance.
- Defect rates: define a target and a limit, and score linearly between them.
- Cost: one common method is lowest total landed cost divided by this supplier’s total landed cost, times 100. The cheapest supplier scores 100.
- Subjective items: use a short rubric with defined levels, such as 0, 50 and 100, and examples of what each means.
Common mistakes that distort rankings
- Too many criteria, which dilutes important ones and encourages scores nobody can back with data.
- Weights set after seeing the results.
- Scoring cost by unit price only, so a cheap supplier with high freight, duties or rework wins unfairly.
- Letting one person score subjective criteria alone.
- No must-have gates. A required certification, minimum financial strength or legal compliance should be a pass or fail test outside the score, so a high total cannot hide a critical failure.
- Comparing suppliers from different categories on the same scorecard.
- Reviewing too rarely, so the card reflects last year, or too often, so noise looks like change.
How often should suppliers be reviewed?
Review on a fixed schedule that fits the data. Quarterly works for suppliers whose delivery and quality data accumulate quickly, and annual reviews suit lower-volume or lower-risk vendors. Share results with the supplier along with the evidence, and agree improvement actions for any criterion that falls short. A scorecard that nobody sees does not change behaviour. Related guidance is on the procurement and suppliers hub, and the scorecard can be rebuilt any time in the supplier scorecard calculator to test new weights.
Where analytics and AI can help
Analytics reduces the effort of collecting delivery, quality and invoice data from several systems, and it can refresh scores more often than a manual review. AI-based tools can also monitor news, financial filings and other signals for supplier risk. The weights, the gates and the final judgement stay with people, since a model cannot decide what a late shipment costs your business.
Common questions
Which criteria should a supplier scorecard include?
Most cover quality, delivery, cost, responsiveness and compliance or risk. Add criteria that matter to your business, such as sustainability or innovation, and keep the list short enough to measure well. Each criterion should have a data source and a defined way to turn it into a score.
How are the weights chosen?
Base them on what a failure would cost you. If late deliveries stop production, weight delivery higher. Agree the weights with the people who depend on the supplier before scoring, and test how the ranking changes when weights shift slightly.
Can a scorecard replace judgement?
No. A scorecard organises evidence and makes comparisons consistent, but strategic fit, relationship quality and exposure to disruption still need human judgement. Use the score as one input, alongside gates, site visits and conversations with the people who work with the supplier.
How many suppliers should be on a scorecard?
Score the suppliers that matter most by spend or by risk, not every vendor. A small group of critical and high-spend suppliers deserves regular scoring, while low-value suppliers can be reviewed by exception. Spreading effort too thin lowers data quality across the board.