Polluxa

Why Samples Come Back Wrong. And How to Get Them Right Sooner.

Most wrong samples start with the spec, not the factory. Here are the six common causes, how to tell which one hit you, and the habits that get samples right sooner.

Why Samples Come Back Wrong. And How to Get Them Right Sooner.

Samples come back wrong when the factory works from information that is incomplete, unclear or out of date. In most cases the supplier made what it was told to make; what it was told did not match what the brand meant. The fix is one current spec, measurable instructions and written feedback on every round.

Each wrong sample adds days or weeks to the calendar, so there is less time left for bulk. Most of the causes sit on your side of the courier, where you can fix them.

Key takeaways

  • Most wrong samples trace back to the spec, not the sewing: a missing trim, an old version or a vague note.
  • Updates sent in separate emails or WhatsApp messages are the fastest way for a factory to end up on the wrong file.
  • Tolerances, colour references and marked-up photos turn opinions into instructions a supplier can follow.
  • One approver per stage stops conflicting feedback from reaching the factory.
  • Keeping specs, comments and approvals on one product record, with every version saved, removes most of the guesswork.

What a wrong sample is. And what it is not.

A sample is wrong when it does not match the approved spec closely enough to move to the next stage. Our guide to types of samples in fashion explains proto, fit, salesman, pre-production and top of production samples, and what each one checks.

The usual faults:

  • Wrong fabric, weight or finish.
  • A colour that does not match the approved standard or lab dip (a small swatch dyed to test a colour).
  • Measurements outside tolerance, or the wrong size made.
  • Missing or wrong trims: zips, buttons, labels, thread colour.
  • Changed construction: seam type, stitching, pocket placement.
  • A print or logo in the wrong place, size or colour.
  • Packaging or labelling that does not follow the instructions.

Not every fault is the supplier's. Before you send a sample back, check whether the spec actually said what you expected.

Six things product teams say. And what is really behind them.

“They made it from the old spec.”

The spec was updated, but the change went out in a separate email or WhatsApp message. The pattern maker kept the attachment from the first send and never saw the update.

Put a version number and date on every send, record what changed, and keep one version everyone can see.

“Where did this lining come from?”

The tech pack named the main fabric, but not the lining, thread colour, care label or polybag. When a detail is missing, the factory fills the gap with whatever it normally uses.

Send a complete tech pack that lists every component, including the small ones.

“We said slightly cropped.”

"Navy", "a bit softer" and "slightly cropped" mean different things to different people. Without a tolerance (the amount a measurement can vary and still pass, for example plus or minus 1 cm on chest width), the factory cannot tell whether a measurement is acceptable, and you cannot fairly reject one.

Give measurements with tolerances, colour references and photos with markups instead of adjectives.

“We already told them that.”

Comments were given on a video call, in a chat thread or as a voice note. Some reached the factory, some did not, and none were written back into the spec, so the next sample fixes half the problems and repeats the rest.

Write feedback on the spec itself after each round, then send one consolidated list.

“Which comments were they meant to follow?”

Design, merchandising and quality each sent their own notes, and some contradicted each other. The supplier picked one, and the sample pleased nobody.

Name one approver per stage. Others can comment, but one person signs off and resolves conflicts before anything reaches the factory.

“Nobody told us they changed the fabric.”

The specified fabric or trim was late or out of stock, so the factory used something close. Nobody flagged it, and the substitution only showed up when the sample arrived.

Ask suppliers to confirm they have the latest version and to flag any material change before they cut.

Check the spec first. Then the sample.

Compare the sample against the spec version the supplier had, not the one on your screen today. Then ask:

  1. Which version did the supplier work from, and when did they receive it?
  2. Was the faulty detail in that version at all?
  3. If it was, was it clear and measurable, with a tolerance or reference?
  4. Did the supplier raise a question or flag a substitution that went unanswered?

If the detail was missing or vague, it is a spec problem, and fixing your process will fix the sample. If it was present, clear and current, and the supplier still missed it, it is a quality problem to raise with the factory directly.

Follow one jacket back to its spec

Imagine Sara, a product developer at an outerwear brand. Her fit sample arrives with the wrong zip and a long sleeve. For example:

  • She opens the version the factory received, not today's version.
  • The zip was never in the BOM (bill of materials), so that one is a spec gap. She adds it.
  • The sleeve length was in the spec with a tolerance, and the sample is outside it. That goes back to the factory as a quality point.
  • Comments from design and quality become one list with photos, signed off by her.
  • The next send carries a new version number and a note of what changed.

It separates what the brand missed from what the factory missed.

Right first time. Eight habits.

  1. Lock the design before you request a sample. Changing the silhouette after the tech pack is sent restarts the work for everyone.
  2. Send a complete tech pack. Every trim and packaging item, base-size measurements and construction notes.
  3. Use measurable language. Tolerances, colour references and marked-up photos.
  4. Keep one version, and say which one.
  5. Write feedback on the spec itself. Then send one consolidated set.
  6. Name one approver per stage.
  7. Ask suppliers to confirm and flag substitutions before they cut.
  8. Track every round per style. When each sample was requested, received and reviewed, and the outcome.

For more on cutting the number of rounds, read how to reduce sampling rounds.

Spreadsheets and chat. Where the causes meet.

Spreadsheets, shared drives and messaging apps work for a handful of styles. As the range grows, versions multiply and comments scatter, so the causes above become one: the spec, the feedback and the approvals live in different places.

With a few styles and one supplier, the habits above matter more than any tool. Beyond that, PLM software for consumer brands puts the tech pack, materials list, comments and approvals on the product record itself. Think of it as the one file your team and your factory both read before anything is cut.

How Polluxa helps

Polluxa PLM is product lifecycle management software for fashion, footwear, beauty, consumer goods and manufacturing brands.

  • Factories no longer work from old files. In Product Development, tech packs, BOMs, comments and approvals live on the product itself, with a full history of every change, so everyone sees the latest spec.
  • Feedback no longer gets lost in words. Visual markups and annotations align teams on design and quality expectations, and samples and fit approvals are tracked per style.
  • Tech packs no longer miss pieces. The Tech Pack Generator builds a supplier-ready tech pack from product data, with images, annotations, markups, specifications and comments.
  • Supplier messages no longer sit in five apps. You talk to suppliers by quote, by product or by order, and each supplier sees only their part.
  • Approvals no longer wait in inboxes. Workflows send each step to the right person with a due date, and show where every product is.

Start this week

  1. Take the last three samples that came back wrong and sort each fault into "spec problem" or "quality problem".
  2. Add tolerances to the measurement chart of one live style before its next send.
  3. Name one approver for each sample stage, and tell your suppliers who it is.

When you want to see sampling on one of your own products, book a demo.

Frequently asked questions

Why do samples from the factory have wrong measurements?

Usually the spec had no tolerances, the factory used an older measurement chart, or the base size was unclear. Send points of measure with tolerances and confirm which version the factory has.

How many sample rounds are normal?

It depends on the product. A simple style with a proven fit may be approved quickly; technical products need more rounds. If the same faults repeat, the cause is usually lost feedback, not complexity.

Who is responsible when a sample comes back wrong?

Check the spec first. If the detail was missing or unclear, the brand needs to fix the spec. If it was present, current and clear, the supplier should correct it, and repeated misses are a quality conversation.

What should I write in sample feedback?

Measured results against the spec, the tolerance each one failed, and a marked-up photo for every visual issue. Send one list per round, from one approver, and save it on the product.

Can PLM software reduce sample errors?

Yes. A PLM system keeps one current spec per product, saves every version and records comments and approvals in one place. That removes the two most common causes of wrong samples: old files and lost feedback.

More Blogs