Supplier evaluation · capability criteria · evidence · trial orders · scoring

How to Evaluate a Corporate Merch Supplier

In custom merchandise the cheapest opening offer is frequently the least capable one, and a price list is a poor guide to whether a supplier can actually do the work. Capability is method-specific, scale-specific, and evidenced rather than asserted, which means it cannot be read off a catalogue or inferred from a unit price. The evaluation that works is a gate followed by a score: a short list of binary criteria that eliminate a supplier who cannot do the job at all, and a weighted rubric applied only to the suppliers who survive. This guide sets out the four capability blocks worth testing, the specific evidence to request for each, how to structure a trial order so it tests something real, and the handful of conditions that should disqualify a supplier outright rather than merely lower its score.

Request a supplier pack Read all guides

Why a price list is not an evaluation of a merchandise supplier

The cheapest opening offer in this category is often the least capable one, and the reason is structural rather than accidental. A low per-unit figure at a given volume normally means volume-purchased blanks and a standard decoration method. The supplier who can print onto a specific substrate, hold a colour reference against a named standard, work to a stated tolerance on a rigid good, or repeat a specification identically two years running is buying capability with that margin. A price list is not a proxy for any of it. A buyer who narrows the field on unit price ends up comparing offers that differ in the two capabilities the programme most depends on, and the cheaper offer is simply the one that omitted them.

Capability resists reading from a catalogue for three separate reasons, and each one defeats a different kind of shortlisting. It is method-specific: a supplier with heavy screen capacity may hold no embroidery setup at all, while a supplier with strong embroidery may subcontract every printed item, which means the schedule and the quality control on that portion of the order sit outside their building. It is scale-specific: running 250 units and running 5 000 units are different operations with different bottlenecks, and a supplier excellent at small runs is not automatically strong at volume. And it is evidenced rather than asserted: the quality of what a supplier offers as proof is itself a signal, because a supplier that can name the machine, the blank source, the tolerance and the inspection step is describing a process, while one that writes high quality in a summary paragraph is describing an intention.

This is why the evaluation should run as a gate before it runs as a score. The gate asks a small number of binary questions: can this supplier perform the required decoration at all, at the required volume, on the specified material, and will it accept the inspection and remedy terms the programme depends on? Binary criteria are cheap to test and expensive to discover late, which makes them the right place to spend the first round of questions. The score then ranks only the suppliers that survived, on the things that differ in degree rather than in kind. Reversing that order — scoring everything and folding the disqualifiers into a weighting — produces a spreadsheet in which a supplier missing the single method the programme needs can still finish at the top, which is precisely the outcome the exercise exists to prevent.

Decoration methods: Named per method with in-house or subcontracted stated for each; not one phrase covering the whole offer
Blank sourcing: Named grades and sources; a category leaves the substitution decision entirely with the supplier
Capacity at volume: Evidence of throughput at the quantity the programme needs, not a general claim about production capacity
Quality process: How a nonconformity is raised, who inside the supplier owns it, and what a rework involves
Commercial posture: Whether delivery, inspection and remedy terms are stated before the quote is requested or only after
Reference evidence: Projects of comparable specification and volume, described in enough detail to be checked rather than merely listed

Four capability blocks worth testing, and the evidence that proves each

Requests for evidence are easy to make and easy to answer badly. The useful version names the artefact rather than the quality. Below are the four blocks that separate suppliers in practice, each with the specific thing to ask for; a supplier who answers all four in detail has told you more than one who answers all four in adjectives.

Decoration capability, in-house or declared out

Ask what is run on the supplier’s own equipment, method by method, and what is passed to somebody else. The answer matters for three reasons. Schedule: work that leaves the building sits outside the supplier’s control and outside their queue. Quality: a subcontracted step is a point at which your specification meets somebody who has never read it. Accountability: when a defect appears on a subcontracted step, the supplier who answers for it is the one that did not perform the work. Push one step further and ask whether they will name the subcontractor. Suppliers who subcontract are entirely normal and there is nothing wrong with it — but a supplier who will not say where the work goes has asked you to accept a risk they have not described, and that is a different proposition from one who says plainly that embroidery goes to a named partner and carries an extra week. Also establish who owns artwork preparation. Where the supplier prepares it, they are accountable for the file being production-ready; where the buyer does, every subsequent question about why a colour did not match is answered slowly.

Material and blank sourcing depth

Ask for the grade rather than the category, every time. A supplier who can name a fabric weight, a blend, a stainless grade, a coating or a paper grammage is demonstrating that they hold options at that level. A supplier who answers mid-weight cotton has described a shelf, and a shelf means the substitution is theirs to make. The behavioural question is sharper than the technical one: when the specified grade is unavailable, what does this supplier do? There are three honest answers. They hold the production slot and substitute, having told you the date cost. They substitute and inform, having absorbed the schedule impact. Or they substitute silently, having decided the difference would not be noticed. Only one of those three is acceptable, and you cannot tell which you have been offered until you ask the question in advance. On a repeat programme this single answer predicts more about the next order than the whole of the first one, because a drifting specification that nobody flags becomes the specification within a year.

A quality process with a named owner

Ask what happens when something is wrong, and then keep asking until you reach a name. The useful answers describe a sequence rather than a sentiment: how a nonconformity is reported to the supplier, who receives it, what evidence they ask for, whether the remedy is rework, replacement or credit, and how long they will take to decide. A supplier that answers with a quality policy document has given you a policy. A supplier that answers by describing the last time a customer reported a colour mismatch and what was done about it has given you a process. Two specific commitments are worth requesting in writing, because both are routinely available and both are routinely assumed. The first is agreement to a defect classification — an agreed list of what counts as critical, major and minor, fixed before production rather than negotiated after it. The second is an inspection step at a point the buyer controls. Neither is expensive, both change behaviour, and a supplier willing to agree both is telling you something about how it handles the orders where nobody is watching.

Commercial and communication discipline

Capability that cannot be communicated is not usable by a buyer who has to plan around it. The signals are unglamorous and reliable. Does the quotation break itself into the layers a buyer can check — blank, decoration, setup, packaging, freight — or does it present a single number that cannot be interrogated? Does it state what is excluded, including artwork preparation, samples, delivery to a named address and anything to a regional destination? Does it answer a technical question in writing without requiring a call? And does it raise a risk itself — a tight production window, a material with a long availability lead time, an artwork file that will not separate cleanly — before the order is placed rather than afterwards? The last of these is the one buyers underrate. A supplier who volunteers the problem in week one is demonstrating that their communication reaches you; a supplier who mentions the same problem in week six is demonstrating that their communication only runs when prompted, which is a fact worth knowing before you depend on it for a dated programme.

Turning capability into a shortlist that can actually be chosen from

A workable rubric has five blocks scored against named evidence on a small scale, and the scale should describe the evidence rather than the impression. Zero means nothing was offered. One means an assertion. Two means a description. Three means a specific, checkable answer. Scoring the evidence rather than the relationship is what keeps the exercise auditable six months later, because a reader can return to the answer that earned the mark. Weight the blocks by what the programme actually depends on rather than treating all five as equal: an annual staff apparel programme leans on repeatability and colour consistency, a single client gift set leans on presentation and packaging, and an event kit leans on capacity and on a production window that will not move. Set two thresholds rather than one — a minimum overall score, and a minimum in each individual block. A single threshold permits a total built almost entirely on one strength, which is a description of a supplier who is excellent at the wrong thing.

Disqualification belongs in the same document, stated as plain conditions and applied before the scoring, not as penalties folded into it. A supplier who cannot perform the required decoration method at all, who will not state the material grade, who will not agree a defect classification, who will not accept an inspection point before release, or whose terms transfer every risk to the buyer without offering a corresponding remedy, is out regardless of how well they score elsewhere. Each of those can be perfectly strong on all five blocks and still be unusable, which is exactly why they cannot be expressed as a weighting. Writing the exclusions down in advance also protects the evaluation from its own most common failure, which is that the criteria quietly move once a favourite has been identified and every subsequent question is asked only of the others.

Running a trial order instead of a longer conversation

A trial is an experiment, and an experiment has one variable. Design it around the single decision that is still open — the decoration method where the tactile result matters more than the photograph, the substrate grade where weight changes how the item is perceived, the closure or fastener on a bag — and hold everything else against an already-approved specification. A trial that varies material, decoration, packaging and timing simultaneously returns a verdict rather than a diagnosis, and a buyer holding a verdict has to run the whole thing again to find out what to change.

Quantity is a design decision too. Run enough units for the decoration process to settle into its steady state, since the first units off a screen or the first minutes of an embroidery run are not representative, and few enough that the fixed setup is not the dominant line on the invoice. Inspect the result against the written specification rather than against a general impression of quality: colour against the named reference, decoration position and dimensions against the approved proof, substrate against the stated grade, packaging as specified, and the count against the size or recipient split. Keep that inspection note whether the trial passes or fails, because the record of a failure is the part that tells you whether to stay with the supplier.

Decide in advance what each outcome means, because that decision is almost impossible to make once the result is in hand and everyone involved has a view. If the trial passes, the specification is locked and the supplier’s record opens with a pass on a defined specification rather than on a general impression. If it fails, the failure has to carry a name — material, decoration, finish, tolerance or schedule — and an unnamed failure reliably produces a second trial with the same variables and a shorter supply of patience. Whichever way it goes, keep the artefact: the trial unit, its inspection note and a photograph become the reference object that every later order is checked against, which is arguably worth more than the trial itself.

Tell the supplier what the trial is. A run described as a sample is understood as a courtesy offered before an order, and a courtesy is not built to be measured. A run described as an evaluation, with the specification to be written from its result and the outcome recorded either way, gets treated as one — which is also the only way to ensure the trial is not quietly optimised, since a supplier expecting a verdict will have an incentive to show the specification at its best rather than at its most representative.

Questions buyers ask when qualifying merch suppliers

How many suppliers should we evaluate for a one-off programme?

Four is usually enough to make the comparison real, and three is the practical floor. Below three there is no ranking, only a choice between one option and ignorance. Above four the evidence requests start to crowd the calendar without improving the decision, because the suppliers who answer carefully are the ones already in the set. Run the gate against all four, then score only the survivors, and record the gate result for every supplier including the excluded ones. The excluded list is worth more than the shortlist at this stage, because it is the record that stops the same supplier being reconsidered from scratch in six months because nobody could remember why it was passed over.

What should a supplier show as evidence that they can handle our volume?

Something specific to the quantity and the method rather than a general statement about production. In practice that means the shift pattern and the equipment they would schedule the run on, the throughput they commit to for that method, where the blanks come from and how long those take to arrive, and what they would do if the volume moved by half. The last question is the diagnostic one: a supplier who has thought about a halving has a cost model and a schedule model, and one who has not is pricing the run you described and nothing beyond it. Where the programme depends on the supplier holding a slot across several orders rather than running a single batch, that is the question to press hardest, because it is a different operational commitment from filling one order on time.

When is a trial order worth the extra cost and lead time?

When the programme is a repeat one and the specification is still open. The first production run of an ongoing programme is the cheapest trial available, because it is not extra — it is the first order, done deliberately, with the evaluation written down before it is placed rather than reconstructed afterwards. A separate paid trial only earns its cost in two situations: when the specification has options that cannot be resolved on paper, such as a decoration method where the tactile result matters more than the photograph, or when a supplier is being qualified for capacity reasons on a programme far larger than a pilot can prove. What a trial must avoid is being used to test everything at once, because a run that tests material, decoration, packaging, timing and colour simultaneously produces a verdict rather than a diagnosis, and the buyer then knows the order failed without knowing which part to change.

Is Corp-Merch affiliated with, endorsed by, or an official supplier to the brands, agencies, or institutions referenced on this page?

No. Corp-Merch is an independent B2B merchandise supplier. Any mention of a manufacturer, a brand, a standards body, a carrier or an institution is informational and implies no partnership, sponsorship, endorsement or official-supplier relationship. Nothing on this page is legal, tax or financial advice, and any criterion, score or exclusion described here reflects typical practice for this category rather than a fixed commercial offer.

This page is written for procurement and operations teams qualifying custom corporate merchandise suppliers in Armenia. Corp-Merch is an independent B2B merchandise supplier. The capability blocks, scoring scale, thresholds and exclusion conditions described here are a reasoned framework for supplier qualification; they are not legal, tax or financial advice, and an organisation should adapt them to its own procurement policy and internal controls. Any delivery term, payment term or lead time referenced describes typical practice in this category rather than a fixed quotation. Each order is priced individually against a written specification, an approved proof and a signed agreement.