Vendor Scorecard: Criteria, Weights and a Worked Example
A vendor scorecard rates a supplier on the handful of criteria that matter for that relationship, using a fixed scale and weights so that reviews are comparable from quarter to quarter and vendor to vendor. Pick four to six criteria, weight them to total 100, rate each from 1 to 5, and review the result with the vendor.
A scorecard turns the vague sense that a vendor is fine or is a problem into evidence you can use at renewal. It only works if the criteria are measured from data you actually have.
Vendors Covered in this Article
Disclosure: We may earn a commission if you buy through some links on this page. It doesn't change what we recommend.
Which criteria belong on a vendor scorecard?
Choose criteria that reflect why you hired the vendor. A software vendor and a logistics supplier shouldn't share the same list. Common categories to draw from:
- Quality: defects, rework and whether output meets the agreed standard.
- Delivery and reliability: on-time delivery, uptime for services, and consistency.
- Responsiveness: time to respond and resolve issues, and quality of communication.
- Cost: price against the contract, invoice accuracy and surprise charges.
- Risk and compliance: security, insurance, certifications and data handling.
- Relationship: proactive ideas, flexibility and ease of working together.
Keep to four to six. Every criterion needs a source, such as a ticket log, delivery record or invoice check, and a person who owns collecting it. A criterion with no data is an opinion.
How do you weight and rate the criteria?
Weights show what matters most. Say you're scoring a cloud service that runs customer-facing systems: reliability might carry 35 of 100, responsiveness 25, cost 20, risk and compliance 15 and relationship 5. For a one-off design supplier, quality and delivery might dominate instead.
Define the 1 to 5 scale in words for each criterion, so two reviewers give the same score:
- Unacceptable: repeated failures, or a breach of contract.
- Below expectations: misses often, needs follow-up.
- Meets expectations: does what was agreed.
- Exceeds: reliably better than the agreement.
- Exceptional: sets the standard, and rare.
For uptime, set the rating boundaries from the commitment in the contract. As a reference for what an availability target means, a 99.9 percent target allows about 8.76 hours of downtime a year, and 99.99 percent allows about 52.6 minutes1. Ask what your business can tolerate before you accept a target.
What does a worked example look like?
Use made-up numbers to see the calculation. Say a payment-processing vendor is rated 4 on reliability, 3 on responsiveness, 5 on cost, 4 on risk and 2 on relationship, with weights of 35, 25, 20, 15 and 5.
Multiply each rating by its weight and divide by five to turn the score into a 100-point scale: reliability is 4 times 35 divided by 5, or 28. Responsiveness is 15, cost is 20, risk is 12 and relationship is 2. The total is 77 out of 100.
Then set what the score means for action, and agree the bands before the review, not after. For instance, 85 and above means keep and consider expanding, 70 to 84 means keep with a note, 55 to 69 means put an improvement plan in place, and below 55 means start looking for alternatives. The bands here are illustrations. Choose yours to fit your tolerance.
How do you run the review?
Collect data in the week before, score it, and share the results with the vendor ahead of a short meeting. Discuss the lowest-scoring criteria first and agree on one to three specific actions with dates. Note what the vendor rates about you, too, since slow approvals or unclear requests on your side are often part of the problem.
Cadence follows risk: quarterly for vendors that are critical to delivery, twice a year for important ones, and at renewal for the rest. For critical vendors, track service-level performance continuously, not only in review meetings. Combine scorecard results with contract dates using the contract lifecycle guide, so a poor score triggers action before the auto-renewal date.
For a deeper model that groups vendors into tiers, read the vendor scorecard and tiering model, and if you're evaluating tools to run this, see this spend management tool comparison.
What mistakes make scorecards useless?
Watch for these:
- Too many criteria, so no one can tell what drives the score.
- Subjective ratings with no evidence, which vendors rightly dispute.
- Never sharing results with the vendor, so nothing changes.
- Using the same scorecard for every vendor, regardless of what matters.
- Scoring without consequences: no one acts on a low result.
- Changing the scale or weights each cycle, which makes trends meaningless.
The same discipline applies to internal scorecards, such as the leadership team scorecard and hiring's structured interview scorecard. To keep vendor data consistent from the start, use the vendor onboarding checklist.
What Good Looks Like
A good vendor scorecard has four to six weighted criteria, each with a data source and a defined 1 to 5 scale, and is reviewed with the vendor on a fixed cadence.
Building The Capability (5-Stage Skill Ladder)
How to Get Started
Disclosure: We may earn a commission if you buy through some links on this page. It doesn't change what we recommend.
Frequently Asked Questions
What is a vendor scorecard?
It's a structured rating of a supplier against a few weighted criteria, such as quality, delivery, responsiveness, cost and risk. Reviewing it regularly gives you evidence for renewals, improvement plans and vendor decisions.
How many criteria should a vendor scorecard have?
Four to six. More than that makes the score hard to interpret, and each criterion needs data behind it. Pick the ones that reflect why you hired the vendor.
How often should you review vendor scorecards?
Quarterly for critical vendors, twice a year for important ones and at renewal for the rest. Also review after a major incident or contract change.
Should you share scorecards with vendors?
Yes, in most cases. Sharing results gives the vendor a chance to fix problems and makes the review a working conversation. Keep the scoring evidence, so any disagreement is settled with data.
Sources
Where we quote a benchmark, we show its source. Other figures in this guide are estimates or general guidance, so check them against your own numbers.
- Allowed downtime per year by availability target. Google SRE Book, Table 1-1 Availability table, 2016.
Related Guides
Stopping a Vendor Contract From Auto-Renewing on You
A simple system for catching vendor contract renewal windows before they pass, so an evergreen clause never locks you into another term by default.
How to Build a Vendor Scorecard and Tiering Model
A simple way to sort vendors into tiers by how much risk and spend they carry, so review time goes to the suppliers that actually deserve it.
Procurify vs Coupa vs Ramp: Procurement & Purchasing Software Compared
Compare Procurify, Coupa, and Ramp for purchase order approvals, 3-way matching, corporate cards, vendor contract tracking, and spend compliance controls.
Leadership Team Scorecard: Build One in a Spreadsheet
Build a one-page leadership team scorecard: which 10 to 15 measures to track, how to set red thresholds, and how to review it in five minutes a week.
How to Build a Structured Interview Scorecard
Build a structured interview scorecard with competencies, behavioral questions and rating anchors, so hiring decisions rest on evidence, not impressions.
Vendor Onboarding Checklist: Vetting a New Supplier
How to onboard a new vendor: tier the risk, collect the right documents, verify payment details, set up the record, and review the vendor after 90 days.