Supplier Scorecards: How to Evaluate Chinese Suppliers Objectively

A supplier scorecard is a one-page weighted scoring sheet that rates each supplier on the same criteria: quality, delivery, cost, responsiveness, and compliance. Each criterion gets a weight, each supplier gets a 1–5 score per criterion, and the weighted total is the number you compare. Re-score every quarter, and decide by the trend, not just the total.

This works especially well for China sourcing, where price is easy to see and everything else is not. A scorecard forces you to put a number on communication speed, inspection results, and factory reality, instead of letting the cheapest quote win by default.

If you're still at the candidate-screening stage, start with the shortlist process in our guide to building a supplier shortlist. That article covers the one-time selection matrix. This one covers the tool you keep using after you've picked suppliers: the living scorecard you update as the relationship runs. A sourcing partner like CN Ally can fill in the hardest columns for you, from on-site verification to inspection results, but the scorecard itself is yours.

How a scorecard differs from a shortlist matrix

They look similar and people mix them up. The difference is when you use them and what feeds them.

A shortlist matrix is a selection tool. You build it when you're comparing five to eight candidate suppliers, before you've spent real money. Most of its scores are educated estimates: quote competitiveness, sample quality, communication speed, how the factory looked in a video call. You use it once, you pick a winner, and it's done.

A scorecard is an operating tool. You build it after you've started buying, and it runs for as long as the relationship does. Its scores come from evidence you accumulate: defect rates from incoming inspections, on-time delivery records, how many days it took to resolve a quality claim, whether the supplier hit the agreed FOB price. Because the inputs are real, the scorecard catches drift that the shortlist never could. A supplier who scored an 82 at selection can slide to 64 eighteen months later, and without a scorecard you usually notice only when a shipment fails.

Run both. The shortlist gets you a good starting decision; the scorecard keeps the decision honest.

What goes on the scorecard: six KPI categories

Keep the category list short. Three to five indicators that actually matter beat ten that nobody maintains; as one China sourcing practitioner puts it, agreeing on just 3 or 5 indicators is usually better than tracking 10 or 20 you will never maintain. For Chinese suppliers, six categories cover nearly everything worth tracking:

Category · Weight range · Example metrics

• Quality: 25–35% · Defect rate per shipment, AQL results, corrective-action closure time
• Delivery: 15–25% · On-time delivery rate, quoted-vs-actual lead time, re-schedule frequency
• Total cost: 15–25% · Unit price vs. quote, freight and duty variance, payment-term changes
• Responsiveness: 10–15% · Quote turnaround, reply time on issues, English clarity
• Compliance & verification: 5–15% · Valid business license, audit status, certification currency
• Flexibility & capacity: 5–10% · MOQ flexibility, rush-order capability, scalability for growth

Weights should add to 100%, and they should reflect your business. A private-label cosmetics brand lives or dies on quality, so quality gets 30% or more and cost gets less. A commodity accessory importer running tight margins can push cost to 25% and trim flexibility. There is no universal set; the weights are the strategy.

Two China-specific adjustments matter. First, give compliance and verification a real weight, not a token one. In domestic sourcing this category barely exists. For China, an expired business license or an unverifiable factory address is an existential risk, and it deserves to move the number. Second, don't over-weight unit price. Price in China is negotiable, volatile, and easy to undercut by cutting corners elsewhere. If price sits at 40% of your scorecard, the scorecard will just re-announce the cheapest quote every quarter.

The 1–5 scoring scale: write definitions, not vibes

A scorecard dies the day two people score the same supplier differently because "4" meant different things to each of them. Every point on the scale needs a written definition, at least for the categories you score subjectively. Hard data (defect rate, on-time percentage) converts to scores by thresholds you define. Judgment calls (responsiveness, cooperation) convert by rubric.

Here is a rubric that works well for China sourcing scorecards:

Score · Label · What it means in practice

• 5: Exceptional · Top-of-class. Zero defects in the last two quarters, or replies within hours with full answers, every time. Rare.
• 4: Strong · Exceeds requirements. Occasional small misses, fixed fast, no pattern. The kind of supplier you recommend.
• 3: Acceptable · Meets the requirement, nothing more. Ships on time mostly, answers within a day, defects within tolerance. Most decent suppliers live here.
• 2: Below par · Repeated misses or one serious failure without a convincing fix. Needs an improvement plan with dates.
• 1: Unacceptable · Fails the requirement. Counterfeit certificates, ghosting during a claim, defect rates that stop your line. Escalate or exit.

Two discipline rules keep scores honest. Score from evidence, never from memory: the delivery column should come from your shipping records, the quality column from inspection reports, the responsiveness column from your email timestamps. And score the supplier against the requirement, not against other suppliers. A 3 means "meets the bar," full stop. Relative scoring turns the scorecard into a ranking contest and destroys its meaning across quarters.

How to weight criteria for your business

The weight table above gives ranges. Choosing your exact numbers is the most consequential decision in the whole exercise, so don't inherit someone else's defaults blindly. Walk through three questions:

What kills us if it goes wrong? List your last two supplier disasters, or the two you most fear. For most importers from China, the answers are a quality failure that reaches customers, and a shipment delay that empties stock. Whatever categories cover those answers get the top weights. The scorecard's job is to warn you before the disaster, so it should lean toward the categories that cause disasters.

What differentiates suppliers in our category? If every supplier in your niche quotes within 5% of each other, cost is not a differentiator and weighting it heavily just adds noise. If quality swings wildly between suppliers, weight quality heavily. In China sourcing, communication and reliability often differentiate more than price, because price clusters and reliability doesn't.

How much do we actually know? Early in a relationship, you can't score flexibility or cost-trend reliably because you have one data point. Start with what you can measure (samples, quotes, verification documents, trial-order results), set the rest to small placeholder weights, and grow them as data arrives. A scorecard that demands data you don't have gets abandoned. One that grows with you gets kept.

Revisit the weights once a year. Businesses change, and a scorecard frozen at year one eventually measures the wrong things.

Mandatory gates: what a score can't fix

This is the most important design decision in your scorecard, and the one most templates skip. Some requirements should never be blended into a weighted average, because a high total can hide a fatal failure. A supplier scoring 88 overall that cannot prove it is a real, licensed company is not an 88. It is a no.

Keep a separate pass/fail list next to the scorecard:

• A verifiable business registration on China's official registry, with the license matching the name on the contract and bank account
• Factory status confirmed, or a conscious, documented decision to work with a trading company (know which one you have)
• Product compliance for your market: the certifications your country requires, current and authentic
• No sanctions or serious legal exposure you can't accept
• Payment destination matches the verified company, every time

A fail on any gate stops the process regardless of the weighted score. This sounds obvious, and it is, which is exactly why it needs to be written down: weighted averages are persuasive, and without gates, a charming sales rep and a low price can carry a supplier past problems that should have disqualified them. (The Dragon Sourcing scorecard guide makes the same structural point: a good model combines mandatory qualification, weighted scoring, risk assessment, and due diligence.) For the verification behind these gates, our supplier verification checklist and the business license verification guide walk through each check.

A worked example: scoring two Chinese suppliers

Theory is fine, but scorecards click when you see the arithmetic. Imagine you've narrowed your choice to two suppliers for a mid-volume private-label product, and you've agreed on these weights:

Category · Weight · Supplier A: 1–5 · Supplier A weighted · Supplier B: 1–5 · Supplier B weighted

• Quality: 30% · 3 · 0.90 · 4 · 1.20
• Delivery: 20% · 4 · 0.80 · 3 · 0.60
• Total cost: 20% · 5 · 1.00 · 3 · 0.60
• Responsiveness: 10% · 3 · 0.30 · 4 · 0.40
• Compliance: 10% · 2 · 0.20 · 4 · 0.40
• Flexibility: 10% · 4 · 0.40 · 3 · 0.30
• **Total: 100% · 3.60 · 3.50**

Supplier A is cheaper and faster. Supplier B scores better on quality, responsiveness, and compliance. The totals are close: 3.60 versus 3.50, and a naive reading says A wins.

Now apply the gates. Supplier A's compliance score of 2 reflects something real: during verification, its business license couldn't be cleanly matched to the contracting entity, and its claimed ISO certificate didn't verify. That is a gate failure, and the process stops. Supplier B wins at 3.50, not because 3.50 beats 3.60, but because B passes every gate and A doesn't.

This is why the gates section exists before the math section in every scorecard worth using. The weighted score compares the qualified. It never qualifies the unqualified.

One more lesson from the example: look at the category breakdowns, not just totals. If Supplier B's quality score had been built on one perfect sample and no production history, that 4 should have been flagged as low-confidence. Consider marking scores with confidence levels (high, medium, low) for the first two quarters. It reminds you which numbers are solid and which are provisional.

How often to re-score, and what the numbers should trigger

A scorecard you update once and file away is decoration. The cadence that works for most importers:

• Quarterly for active suppliers with regular shipments. Monthly is overkill for most small and mid-size importers; the data doesn't change fast enough to justify it, and the chore gets dropped.
• After every shipment for quality and delivery metrics, even if you only roll them up quarterly. Log the numbers while they're fresh; compile them on schedule.
• After any incident, immediately. A failed inspection or a missed vessel departure gets scored the day it happens, not three months later.
• Annually for the full review: weights, categories, gates, and whether the supplier still fits your strategy.

The scores have to trigger decisions, or scoring becomes theater. Set thresholds in advance:

• 4.0 and above, trend stable or rising: approved and preferred. Give them more volume, longer commitments, earlier involvement in new products.
• 3.0–3.9: approved with monitoring. Fine to keep ordering, but the weak categories get a written improvement expectation and a date.
• 2.0–2.9: develop or diversify. Put a corrective plan in writing, and start qualifying an alternative in parallel. Don't wait for the 1s to start a backup search.
• Below 2.0, or any gate failure: exit planning starts. Shift volume away, and don't place new development work with them.

Trends matter more than snapshots. A supplier sliding from 3.8 to 3.1 across two quarters is a louder signal than a supplier sitting at a steady 3.0. Keep the last four quarters on the sheet so the direction is visible at a glance.

Where the numbers come from when you have no history

The hardest part of the first scorecard is that half the cells are blank: no defect history, no delivery track record, no real responsiveness data. That's normal. Fill the gaps with the evidence you've already gathered, and mark those scores as provisional:

• Sample quality and revision rounds feed the first quality and responsiveness scores. See our sample sourcing guide for what to measure.
• Quote behavior feeds cost and responsiveness: how fast, how complete, how they handle pushback. Our quotation comparison guide shows what to look at.
• Verification documents feed compliance: license checks, registry lookups, certificate authentication.
• A factory audit feeds quality, compliance, and capacity in one pass, with a report you can score against. A factory audit or remote verification turns "they claim" into "we saw."
• Trial orders are the first real delivery and quality data. Size them to be informative but survivable.

Don't score what you can't evidence at all. Leave the cell blank with a note ("no data yet — fill after first shipment") rather than inventing a number. An honest blank beats a fabricated 3, because the blank tells you what to collect next.

Frequently asked questions

What should a supplier scorecard include?

At minimum: quality, delivery, cost, and responsiveness, each with a defined weight and a written scoring rubric. For China sourcing, add compliance and verification as a full category, not a footnote, plus a separate pass/fail gate list for requirements that must never be averaged away. Every metric needs a defined evidence source, or it won't get scored consistently.

How do you create a supplier scorecard in Excel?

One row per criterion, columns for weight, score, weighted score, evidence, and comments, plus a total row that sums the weighted scores. Add a second sheet for the pass/fail gates and a third for the quarterly history. Keep formulas simple: weighted score equals weight times score divided by 5, or just weight times score if you use the 5-point scale directly as we did in the worked example.

Supplier scorecard vs. supplier evaluation: what's the difference?

Supplier evaluation is the whole process of assessing suppliers; the scorecard is the tool that records it. Think of evaluation as the verb and the scorecard as the spreadsheet. Related tools include the shortlist matrix (a one-time selection aid used before you've spent money) and the factory audit (an on-site evidence source that feeds the scorecard's compliance and quality columns).

What is a good supplier score?

On a 1–5 weighted scale: 4.0 and above means a strong, preferred supplier; 3.0–3.9 means acceptable with monitoring; below 3.0 means improvement plans or exit planning. But the trend matters more than any single number. A supplier dropping from 3.8 to 3.1 across two quarters deserves more attention than one sitting steady at 3.0.

How often should you re-evaluate suppliers?

Quarterly for active suppliers, with quality and delivery data logged after every shipment and rolled up on schedule. Re-score immediately after any serious incident, and review the scorecard's own design (weights, categories, gates) once a year. A scorecard that never changes eventually measures the wrong business.

Can a small importer use scorecards, or is this just for big procurement teams?

Small importers arguably benefit more, because they can't absorb a bad supplier the way a large company can. You don't need enterprise software. One spreadsheet, six categories, quarterly updates. The value isn't in the tooling; it's in making the same decision criteria apply to every supplier, every quarter, instead of deciding by gut and price.

Your next move: build it before you need it

Don't wait for a supplier problem to start scoring. Build the scorecard this week, while nothing is on fire: six categories, your weights, written 1–5 definitions, the gate list. Score your current suppliers from whatever evidence you have, mark the low-confidence cells, and set your first quarterly review date.

The scorecard won't make decisions for you. It will make your decisions defensible, repeatable, and visible to everyone involved, including future you. And when a cheap quote tries to win on price alone, the numbers will ask the question your gut might skip: what is everything else costing you?

If you'd like help filling in the hardest columns, verification, factory audits, and inspection results, reach out at hi@cnally.com or through our contact page. The scorecard is yours to run; the evidence gathering is where a sourcing partner earns its place.