Supplier Scorecard
The Numbers Behind the Relationship

Why Two Distributors Get Two Different Answers From the Same Factory

Ask three distributors of the same printer brand how that supplier performs and you will get three incompatible answers, each delivered with conviction. One reports a founder who answers messages at midnight and ships in nine days. Another reports four-month lead times and a support desk that closes tickets without replying. Both are describing the same plant, at different order sizes and different moments in the quarter. Supplier quality is not a property of the factory; it is a property of the relationship and the measurement discipline the buyer brings to it.

The practical consequence is that a scorecard has to be built before the second order, not after the first dispute. A distributor who arrives at a price negotiation with six months of documented on-time performance, defect rates and claim resolution times negotiates from a different table than one who arrives with a feeling. This guide sets out the nine metrics worth tracking, the audit that produces them on a first visit, and the traffic-light thresholds that turn a spreadsheet into a decision to expand, hold or replace a supplier.

The Six Numbers That Actually Predict Trouble

Most distributor scorecards fail because they measure too much. A dashboard with forty columns gets filled for two months and then abandoned. The six figures below cover the failure modes that cost money in a distribution business, and each can be produced from documents the supplier already issues.

On-time delivery against the acknowledged date is the first, and it must be measured against the date the supplier confirmed, not the date the customer wanted. A supplier that acknowledges twelve weeks and ships in eleven is reliable; one that acknowledges eight and ships in eleven is a supplier that will cost you a customer. Track the percentage per shipment and the worst single miss, because one 30-day miss in a season can empty a launch slot.

Incoming defect rate in DPPM is the second. For a printer containing roughly 900 to 1,400 parts, a 1,000 DPPM incoming rate means one machine in a thousand arrives with a fault, which sounds acceptable until you price the field visit. Distributors who handle their own first-line service typically find that a machine arriving faulty costs 4 to 12 times the margin on that unit to resolve between diagnosis, swap, freight and the customer conversation.

Claim resolution time is the third and the most frequently fudged. Measure from your written claim to the arrival of a working replacement or credit note, in calendar days, and split it into acknowledgement, decision and remedy. A supplier that acknowledges within 24 hours and settles in 45 days is easier to plan around than one that acknowledges in a week and settles in ten days only when chased.

The remaining three are spare parts fill rate within 30 days on a defined list of the twenty parts you actually consume, documentation accuracy measured as the share of shipments whose commercial invoice, packing list and conformity documents need no correction, and engineering response time for a technical question that requires the factory rather than the sales desk. That last one is the hardest to measure honestly, because the sales channel will absorb two or three questions before the factory ever sees one. Submit a question that only a design engineer can answer, log the date, and see what comes back.

Measure all six monthly, over a rolling twelve. A single month tells you about the quarter; twelve months tells you about the supplier. The pattern to look for is a supplier that is stable at 92 percent and one that oscillates between 100 and 70, because the second will destroy your own delivery promises twice a year and your customers will attribute the failure to you, correctly. Where a stable baseline matters more than a peak, the commercial terms that support it are covered in our guide to distribution agreement contract terms.

The First Visit: What to Look at in Four Hours

A factory audit that runs for two days produces a beautiful report and no decisions. Four hours on site, spent in the right sequence, produces enough to score a supplier and to know what to write into the agreement. The sequence matters more than the length, because the parts of a plant that matter are the ones that are visible without a prepared tour agenda.

Start at the goods-in area, not the showroom. Incoming inspection is where a factory reveals whether it trusts its own supply chain. Look for a defined inspection standard posted at the bench, a quarantine area with a physical boundary, and records of what was rejected in the last month. A goods-in area with no rejection records either has a miraculous supply chain or no inspection, and the second is far more likely. Ask to see the last ten nonconformance records and check whether the same part number appears repeatedly, which is the signature of a supplier problem that has been tolerated rather than solved.

Then the line, and specifically the station where calibration happens. Every printer requires a bed mesh, an extruder calibration and a motion check before packing. Whether that step is a documented procedure with a recorded result, or a technician's judgement, determines whether the machine you receive in month nine matches the one you received in month one. Ask for the calibration record of a machine built in the current week, and compare its numbers to the specification sheet.

• Goods-in: posted inspection standard, physical quarantine, ten recent nonconformance records, repeat part numbers.
• Line stations: documented calibration procedure with a recorded result, not a judgement call.
• Burn-in: duration in hours, what is printed, what is measured, and where the record lives.
• Packing: foam specification, carton drop standard, and whether the pack has been tested or inherited.
• Traceability: can the factory connect a serial number to its build date, its component lots and its test record in under ten minutes?
• Spares: the physical stock of the twenty parts you consume, counted, not described.

Then the burn-in and packing area. A printer that has run for two hours has proven that it powers on. A printer that has printed a defined test part for six to twelve hours, been measured, and been recorded has proven something about the population. The questions are the duration, the test artefact and the acceptance limits. In packing, ask whether the carton has been drop tested and to what standard, and whether the foam specification is written down or copied from whatever the last customer accepted.

Scoring: Turning Observations Into a Weighted Number

A scorecard is useful only if the weights reflect what costs you money in your market. A distributor selling into education cares about documentation and spares; one selling into a manufacturing customer cares about defect rate and engineering response. The weights below are a reasonable default for a distributor of finished printers with a first-line service obligation, and they should be adjusted once you have a year of your own claim data.

Score each dimension from 1 to 5 against written anchors, not against a feeling. A 5 on delivery means 98 percent or better against the acknowledged date with no miss beyond five days. A 3 means 90 to 95 percent with one miss beyond ten days in the period. A 1 means below 80 percent or a miss that stopped a customer's production. Write the anchors into the scorecard document itself; the act of defining what a 3 looks like is where most of the value is created, because it forces the sales and service sides of your own business to agree on what they mean.

Weight the total and sort suppliers into three bands. Above 4.0 is expand: increase allocation, consider exclusivity, invest in joint forecasting. Between 3.0 and 4.0 is hold: keep the volume, fix the specific weak dimension with a written improvement request and a review date. Below 3.0 is replace: begin qualifying an alternative before the next buying cycle rather than during a crisis. The cost of switching suppliers in a hurry, including re-certification and customer re-approval, is typically three to five times the cost of switching on a planned schedule.

A scorecard with no consequences is a report. Attach it to something the supplier feels: the annual volume review, the payment terms renegotiation, or the decision on which new model you take first. Distributors who share the scorecard openly with their supplier, including the weak numbers, usually find that the weak numbers improve within two quarters, because most suppliers have never been told which of their failures actually reaches the customer.

The Four Phrases That Should Trigger a Second Look

Every supplier conversation contains claims that are unfalsifiable and therefore worthless as evidence. Four recur often enough to be worth memorising, because each has a specific question behind it that converts the claim into a fact.

"Our quality is very good." The question behind it is: what is your DPPM, measured how, and who measures it. A supplier that quotes a number and then explains the measurement has a system; one that repeats the adjective does not.

"We can do any tolerance you need." The question is: at what cost, at what yield, and what is your demonstrated capability on this feature. For a printed part, dimensional capability depends on material, orientation and geometry, and a supplier who does not ask about all three is quoting a wish. Our guide to SPC, Cpk and gauge R&R for printed parts sets out how to test the claim with twenty parts.

"We have CE and RoHS." The question is: which certificate, issued by whom, with which standards and which report numbers, and can I have the PDF today. Certificates for a printer model are specific to that model and that revision; a certificate for last year's chassis does not cover this year's power supply. The paperwork side is covered in our GPSR and UKCA compliance guide.

"Delivery is 30 days." The question is: 30 days from what, in which month, for which quantity, and what is the acknowledged date on the order confirmation. A supplier's quoted lead time is a marketing figure; the acknowledged date on a signed order is a commitment, and only the second one belongs in a scorecard.

What a Distributor Should Actually Sell

The commercial logic of a scorecard is not procurement hygiene. It is the raw material for the two offers a distributor can make that a marketplace cannot: a delivery promise your customer can plan around, and a service response that has a number attached. A distributor who knows that supplier A runs at 96 percent on time and supplier B at 84 percent can route the urgent order to A and the price-sensitive order to B, and can quote a lead time to the customer that is true.

That is a margin story. The distributor who says "six to eight weeks" without knowing which supplier will ship is guessing, and the customer eventually learns it. The distributor who says "six weeks from order confirmation, and here is what happens if it slips" is selling risk reduction, which is a different product from a printer at a price.

Pair the scorecard with a written improvement request. When a supplier scores 3.2 overall with a 2 on claim resolution and a 4.5 on delivery, the request is precise and therefore answerable. Broad dissatisfaction produces defensive emails; a specific number with a date produces a plan. Review it at the same meeting as pricing, so the supplier understands the two are connected, and give the improvement a quarter before it affects volume. Our guide to should-cost analysis and teardown covers the cost side of the same conversation.

Finally, keep the scorecard inside your own organisation honest. The most common failure is a distributor who rates a supplier highly because the relationship is pleasant, while the service desk quietly absorbs a 12 percent failure rate. Score the numbers, not the account manager, and let the numbers drive the annual review.

← Back to Blog