Factory evaluation for first-time suppliers: part 2

Key takeaways

1. Treat first-time suppliers as unknown quantities and build the capability picture from scratch rather than from samples and promises.
2. Verify the business basics first: identity, scope, and whether the site manufactures or mainly trades, before judging any process in depth.
3. Probe sampling and test discipline closely, since defined plans consistent with acceptance sampling logic separate controlled sites from improvising ones ^[1].
4. Read findings against the destination market's regulatory context, keeping in view how international principles on technical barriers to trade shape conformity expectations ^[2].
5. Convert the report into an approval decision with named conditions, owners, and recheck points instead of a vague overall impression.

Every buyer remembers a first order from a new supplier that went wrong in an avoidable way. The cartons arrive, the goods look close to specification, and then the field failures begin — wrong components, missed tests, mixed lots. Investigation usually finds that nobody ever checked whether the new source could repeat its golden samples. A first-time evaluation exists to prevent exactly that class of surprise.

New suppliers deserve a different posture from known ones. With an established source, the evaluator verifies that a known-good state still holds. With a stranger, the evaluator must build the picture from scratch: who this business is, what it actually makes on site, how it controls its processes, and whether its records describe reality. Trust is extended in proportion to evidence, not enthusiasm.

This second part of the first-time supplier guide moves from preparation to judgement. It covers how deep evidence review should go with an unknown source, which checks carry the most weight, and how buyers should turn findings into approval decisions with conditions. The emphasis throughout is on observable facts over reassuring narratives.

Unknown sources need evidence, not stories

A new supplier's sales presentation is a marketing document, not a capability statement. Golden samples may have been built by engineers, run slowly, and checked piece by piece; volume production will be built by operators, run at pace, and checked by sampling. The evaluation bridges that gap by observing normal work and testing whether records describe it honestly.

Start with identity and scope. Does the business licence cover manufacturing, does the site match the registered address, and do the processes on the floor match the claimed product range? Trading companies can be legitimate partners, but a buyer who thinks it qualified a factory when it actually qualified an office will misjudge every risk that follows. An early walk arranged with the help of UTS keeps these basic checks in a consistent order so nothing structural is skipped.

References and history help but never substitute for observation. Other buyers' experiences, sample quality, and responsiveness all inform the visit, yet each says more about intentions than about the process that will build the buyer's own orders. The report should therefore weight observed facts above supplied narratives and say plainly where observation was limited.

Where to dig deepest with strangers

With no history to lean on, record review with a new supplier should go deeper than usual. Ask for completed records across several recent lots that resemble the buyer's products, not a single showcase file. Watch a changeover if one occurs, observe the checks operators actually perform at pace, and compare what you see with what the procedures require. Differences observed on day one usually describe the everyday state.

Sampling and testing deserve special attention because they are easy to stage and hard to fake under questioning. Ask which rule sets each sample size, where it is written, and what happens on failure; then check that recent records follow the stated rule. The underlying logic of acceptance sampling, with its defined plans and accept numbers, is what turns checking from a gesture into protection ^[1]. Sites that can name their rules and show records that follow them earn confidence quickly.

Calibration and test equipment form the third deep-dive area. Unlabelled instruments, expired stickers, and borrowed gear all cast doubt on every number the site reports. Because a new relationship has no reservoir of trust, measurement doubts weigh heavier here than with a known source. Note instrument identities, status, and certificate currency directly rather than accepting summaries.

Signals that separate promising from risky

Reading findings against market duties

A new supplier's capability only matters relative to what the destination market demands. Products bound for strictly regulated markets need tighter controls, fuller technical files, and more disciplined testing than simple goods for lenient ones. The evaluation should therefore map findings against the applicable product rules rather than grading the factory in a vacuum.

The international principles governing technical regulations and conformity procedures explain why those market duties exist and how suppliers are expected to demonstrate compliance ^[2]. Evaluators do not need to lecture on the framework, but their reports read better when findings connect to it: this gap matters because that market requires this evidence. That linkage turns a list of observations into a sourcing argument.

For buyers, the practical step is a short market-requirements sheet attached to the approval decision. List each product family, its destination markets, the key standards and labelling duties, and where the evaluation found the evidence thin. New suppliers with thin evidence in high-duty cells need conditions, closer inspection, or a delayed start. Suppliers whose gaps sit in low-duty cells can often proceed with scheduled rechecks.

Turning the report into a decision

1. Sort findings by consequenceSeparate findings into blockers that prevent approval, conditions that allow limited volume, and observations to watch. A report with twenty undifferentiated points produces delay; a sorted list produces a decision.
2. Name owners and datesEach blocker or condition needs an owner at the supplier, a concrete action, and a date. Vague commitments to improve quality by next quarter have no place in an approval record.
3. Set the monitoring planState which shipments will be checked, which checks apply, and what triggers escalation while trust is being built. Tighter inspection is the bridge between conditional approval and full confidence.
4. Recheck narrowly, then expandReturn only to the points that blocked or conditioned approval, verify the evidence, and record the new state. Focused rechecks cost less and close faster than repeated full visits.

Frequently asked questions