How to approve a custom merch sample — and what a sample can't prove

A sample is not a luxury — it’s the last cheap moment to catch a mistake. But most sample checks are done backwards: the order arrives, people stare at it on the desk, and approve the colour they like. A sample approves a spec, not your mood.

The short answer: decide the acceptance criteria before the sample exists, check them in a fixed order, and write the verdict down. The sample proves what that one piece is — it does not prove the batch, the next batch, or the journey in the box.

1. Write the criteria before the piece arrives

The sample can only be approved against something. Take the spec sheet you asked for and turn each line into a pass/fail question — with the answer you’re willing to accept, chosen in advance, not after the fact:

Line on your spec	Check	Acceptable range (example)
Size	Ruler/caliper vs. stated dimension	Within ~1 mm on a 50 mm piece (what you set, not what they guess)
Material & thickness	Measure; feel; edge inspection	State the tolerance before it arrives
Finish (flat / dome / effect)	Look at 45° under steady light	The effect reads as ordered; no stripes or soft spots
Colour	Print vs. your reference, in daylight	Accept the system shift, reject unwanted shifts (see note below)
Hardware / back	Physically attach and remove it	It fastens as intended, and survives being used once
Artwork detail	At use distance, not in zoom	Smallest text and thinnest line still read

The “example” column matters: if you don’t decide what’s acceptable, every deviation becomes a negotiation, and the sample approves nothing.

2. Check in this order — and why

Order is the point: cheap, structural checks first; judgement checks last. Measuring size before looking at colour keeps the decisions from blending.

1. Geometry against spec. Size, thickness, cut accuracy. This is measurement, not taste — and it’s the easiest thing for a supplier to fix.
2. Construction. Edges, dome coverage, print-to-edge fit, hardware seat, back mechanism. This is where the piece tells you it was made and not just printed.
3. Print, under proper light. Steady daylight or a neutral lamp, straight-on and at 45°. Check colour against your reference and the smallest detail you asked for. Neon shifts are real; if your design lives in saturated colour, this is the check that matters (more on RGB vs CMYK).
4. Transparency and underbase. Hold clear acrylic to the light. Where colour must be opaque, it should be; where it must be clear, it should be (white ink on clear bases).
5. The use test — the step most checklists skip. The sample should work in the exact situation your product lives in: put the badge on the bag it will ride on, the magnet on the fridge door at the weight it must hold (what a magnet must hold), the keychain on the actual keyring, the sticker on the surface it will be stuck to. A sample that looks right on a desk and fails in use is a failed sample.
6. Write the verdict. Book or email, one line: “approved as received,” “approved with change — dome deeper,” or “rejected — size off by 2 mm.” Passing a specified line — the size, the finish — protects you at the bulk stage; “looks good” protects nobody, including the supplier.

3. What a sample cannot prove

None of the following is answered by one approved piece, and no supplier’s sample should have to answer for them:

• The batch. A sample is one piece, usually the best one in the room. What you’re buying is pieces #40 and #400 — consistency across the run is a run question, answered at arrival with a first-run check and, if the supplier is new, a real one-SKU test. The honest expectation: the sample proves a representative piece, barely.
• The next batch. Months later, ink, calibration and skill can drift. That’s why the approved sample gets kept as the reference and the reorder gets checked against it — the reorder guide is about exactly this.
• Shipping. A sample survives a padded envelope; the run ships in bulk, and packaging is part of the agreement, not the sample. Agree packaging in writing and inspect on arrival.
• Long-term behaviour. A new dome can look perfect and still yellow over years; a magnet’s hold can change with the surface it’s stuck to. Choose materials for these reasons (ask which resin, confirm what it must hold) — no sample previews the future.
• Colour in every light. The sample proves the colour under your light. Expect the system shift, flag the rest.

In short: the sample is the strongest honest evidence before production — for the geometry, construction, print and use of one piece. Everything after that is a process question, and the process is what a written spec sheet and an arrival check are for.

Set your criteria before you ask for the sample: the spec sheet is what the sample gets checked against, and the quote form returns one in writing for every product.