Building a supplier scorecard for landed cost, lead time and defect rate

Two suppliers quote for the same product. Supplier A comes in 8% cheaper on unit price. Supplier B quotes higher but ships on a shorter, more consistent lead time and has a lower historical defect rate. Buyers who compare only the invoice line pick Supplier A every time, and a good number of them regret it within two or three purchase orders, once the cost of expediting, rework, and stockouts starts showing up in places the purchasing spreadsheet never looked.

A supplier scorecard exists to stop that mistake before the purchase order is issued, not after the container has cleared and the defects have been found. The idea isn't complicated: you're building the true price of sourcing from a given supplier, and unit price is only ever one input into that number. Landed cost, lead time reliability, and defect rate combined tell you what a supplier relationship will actually cost you, and a scorecard is simply the structure that keeps you comparing suppliers on the same three axes every time, rather than reverting to price whenever the sourcing decision gets rushed.

Key Takeaways

• Unit price is not landed cost. Freight, duty, insurance, financing days, and inspection costs routinely add 15-40% on top of the ex-works or FOB price, and that markup varies by supplier and lane.
• Average lead time hides the number that actually costs you money: lead time variability. A supplier with a longer but consistent lead time is often easier to plan around than one with a shorter but erratic one.
• Defect rate should be scored as a rate against units received, not as a count of complaints, so it stays comparable across suppliers shipping different order sizes.
• A weighted scorecard (for example 40% landed cost, 30% lead time reliability, 30% defect rate) turns three incomparable metrics into one number you can use to rank suppliers.
• Re-score suppliers on a fixed cycle, not only after a problem. Scorecards that only get updated after a bad shipment always look better than the supplier actually performs.

Why price alone hides the real cost

A quoted unit price is the easiest number to compare and the least complete one. It tells you nothing about how the goods will actually move from the supplier's factory to your warehouse, what condition they'll arrive in, or how much cash you'll have tied up in transit and in safety stock built to cover an unreliable delivery date. Two suppliers with identical unit prices can produce very different total costs once you add freight mode, packaging quality, customs documentation accuracy, and historical fill rates into the picture.

The scorecard approach forces those hidden costs into the open by scoring three things separately before combining them: what the goods actually cost once they're sitting in your warehouse (landed cost), how reliably the supplier hits the date they promised (lead time), and how much of what arrives is actually usable (defect rate). None of these three is optional to track if you're sourcing at any real volume, and none of them is visible from the quote alone.

Landed cost: the number that actually matters

Landed cost is the ex-works or FOB unit price plus every cost required to get the goods usable in your warehouse: international freight, marine or air cargo insurance, customs duty, port and terminal handling charges, inland trucking, and any inspection or compliance cost specific to that supplier or product category. Industry estimates commonly put total landed cost at 25-50% above the FOB price for a typical import, with ocean freight from Asia to major destination markets alone adding roughly 25-35% before duty and handling are even added (DocShipper, "Landed Cost: Definition & Calculation Guide", retrieved 2026-09-08); the 15-40% range used in this guide's Key Takeaways sits inside that band, and the exact figure for any one shipment depends heavily on product category, freight mode, and destination duty rate. For a UAE importer, this typically means totaling the supplier's price in its billing currency, the freight quote for the relevant mode and lane, the applicable customs duty rate, and any brokerage or clearance fees, then converting everything to a single per-unit figure.

The reason this matters for a scorecard specifically is that landed cost varies by supplier even when the ex-works price is identical. A supplier located near a major port with efficient export documentation adds less friction cost than one that requires longer inland trucking to reach a port, or one whose paperwork errors regularly trigger customs delays and demurrage. Score landed cost per unit, not per order, so that supplier comparisons hold regardless of order size, and run each candidate supplier's numbers through the landed cost calculator before you finalize a sourcing decision rather than after the first shipment has already landed. That gives you a real per-unit figure to put into the scorecard instead of the ex-works price you started with.

Lead time: the average is the wrong number to watch

Most buyers track average lead time and stop there. That's a mistake, because two suppliers can both average 30 days and produce completely different amounts of operational pain. A supplier who delivers between 27 and 33 days every time is easy to plan safety stock around. A supplier who delivers in 18 days on one order and 45 on the next averages to the same 30 days but forces you to carry far more buffer inventory to avoid a stockout, which is a real cost even though it never appears on a supplier invoice.

Score lead time on two components: the average time from purchase order to goods received, and the variability around that average, expressed as the range between your fastest and slowest recorded shipments from that supplier over your last several orders. A supplier scoring well on average lead time but poorly on variability should not automatically outscore one with a longer but tighter delivery window, particularly for products where a stockout is expensive or reputationally costly. Track this per supplier over a rolling set of actual orders, not the lead time quoted at onboarding, since quoted and actual lead times diverge more often than buyers expect.

Defect rate: turning quality into a comparable number

Defect rate should be calculated as defective or non-conforming units divided by total units received in a given period, not as a raw count of complaints or returns. A count of five defect reports means something completely different for a supplier shipping 500 units a month versus one shipping 50,000, and only the rate version of the number is comparable across suppliers of different sizes.

Where you can, break the defect rate down by defect type: cosmetic, functional, and packaging-related defects usually carry very different costs to your business, since a functional defect on a high-value item can mean a full write-off while a cosmetic defect might be sellable at a discount. If you run incoming inspection, sample a consistent percentage of each shipment rather than inspecting some suppliers more thoroughly than others; an inconsistent inspection regime will make careful suppliers look worse than sloppy ones simply because more of their shipments got checked closely.

Building the scorecard: weights, thresholds, and cadence

Once you have landed cost, lead time reliability, and defect rate as separate, comparable numbers per supplier, combine them into a single weighted score so you can actually rank candidates rather than eyeballing three separate spreadsheets. A common starting weighting is 40% landed cost, 30% lead time reliability, and 30% defect rate, though the right split depends on your business: a company selling on razor-thin margins may weight landed cost more heavily, while one selling into a market with high return costs for defective goods may weight defect rate closer to 40%.

Set a minimum threshold on each individual metric as well as the combined score. A supplier scoring very well on landed cost but failing a hard defect-rate threshold shouldn't be masked by a good blended number; some businesses can't absorb a 6% defect rate no matter how cheap the landed cost is. Re-run the scorecard on a fixed cycle, quarterly is reasonable for most import volumes, using actual shipment data rather than supplier-reported figures, and re-score every active supplier at the same time so the comparison stays apples to apples. Scorecards that only get revisited after something goes wrong tend to flatter the supplier who caused the problem, because the review happens exactly when attention (and sometimes a corrective action plan) is highest, which is not representative of ordinary performance.

Worked example: Supplier A quotes $10.00/unit FOB with a landed cost of $12.60 (26% markup), a 30-day average lead time with a tight 27-33 day range, and a 1.5% defect rate. Supplier B quotes $9.20/unit FOB but a landed cost of $12.90 (40% markup, driven by a longer inland trucking leg), a 30-day average lead time swinging 18-45 days, and a 4% defect rate. On landed cost alone the two suppliers are within 2.4% of each other, close enough that lead-time variability and defect rate should decide the ranking: at a 40/30/30 weighting, Supplier A scores higher on all three components and wins the scorecard, even though Supplier B's raw FOB quote looked 8% cheaper going in.

Frequently asked questions

How many suppliers do I need before a scorecard is worth building?

A scorecard earns its keep as soon as you're sourcing the same product category from more than one supplier, or even from a single supplier where you're deciding whether to add a second source. Below that, informal tracking is usually enough; the value of a scorecard is in comparing options, not in monitoring a sole source in isolation. Once you're at the point of actively comparing and adding suppliers, that's usually also the point to step back into the broader trading growth strategy planner, since supplier diversification and volume decisions tend to move together as a business scales its trading operations.

Should landed cost or defect rate carry more weight?

It depends on your margin structure and what a defect actually costs you. If your margins are tight and a defective unit can be discounted and still sold, weight landed cost more heavily. If a defect means a full write-off, a customer complaint, or a compliance issue, weight defect rate higher even if it means accepting a more expensive supplier.

How do I score a new supplier with no shipment history yet?

Score them provisionally using the numbers a similar-tier supplier in your existing base historically achieved, run a smaller trial order to generate real landed cost, lead time, and defect data, and then convert the provisional score to an actual one once you have at least two or three real shipments to draw on. Don't let a strong quote substitute for real performance data in the scorecard.

How often should the scorecard be updated?

Quarterly is a reasonable default for most import volumes, though high-frequency or high-risk categories may warrant a monthly review. The key is consistency: update every active supplier on the same schedule using the same data window, so a supplier isn't scored well simply because the review landed after their best month.

The bottom line

A supplier scorecard is not about finding the cheapest supplier; it's about finding the supplier whose true, all-in cost of doing business with them is lowest, and unit price is only one input into that number. Landed cost, lead time reliability, and defect rate, scored consistently and combined with weights that reflect your actual margin and risk tolerance, will surface suppliers that a price-only comparison would have missed, and will flag expensive ones that looked cheap on the quote.

This guide was reviewed and verified on September 6, 2026.