After the Filament Claim: Root Cause and the Next Order
A credit note closes the invoice, not the cause. This guide covers the half of a quality claim nobody teaches: asking the factory for a named root cause and a change, telling real answers from polite ones, and verifying the fix on the next shipment.

A quality claim usually ends one of two ways. The supplier sends a credit note and everyone moves on. Or the supplier sends a credit note, and the same failure comes back two orders later. The difference is not the money. It is whether anyone ever answered the question the claim was really asking: what went wrong, and what changed so it does not happen again?
Most advice stops at collecting evidence. This guide covers the other half: how to ask for a root-cause answer, how to tell a real one from a polite one, and how to verify the fix on the next order.
Table of Contents
What Does a Good Claim Actually Ask For?
Say a customer returns a spool that snapped mid-print, and the lot matches your records. You send the supplier photos, the lot number and the failure description. That opens the claim.
Now add one sentence: “Please also tell us what you find caused this, and what you are changing, so we can check it on the next shipment.” That sentence is the difference between a refund and a fix. Most suppliers will answer it, because factories run their own investigations; you are only asking to see the output.
The evidence side of this — photos of the label, the batch reference, printer settings — is covered in detail elsewhere, including our sample approval checklist. Treat that as the opening move. What follows here is the middle and the end of the game.
How Do You Ask for the Root Cause?
Be specific, or you will get a shrug in writing. Ask for three things:
- The finding: what did they actually check? The lot record, the line it ran on, the material lot that fed it, the winding or drying step.
- The cause they can name: not “quality issue”, but which step let what through. A named cause can be checked. A vague one cannot.
- The change: what will be different on the next run. A setting, a check, a filter, a gauge, a procedure.
Keep it short and keep it in the same thread as the claim, so the answer lives with the evidence. A supplier that answers in three sentences has done this before. A supplier that answers with a long apology and no content is telling you something too.
One question per claim, and always the same question. If you ask about root cause on some claims and not others, the answer becomes a favour the factory does occasionally. If you ask every time, politely and identically, it becomes part of how you two do business — and suppliers organise themselves around customers whose questions never vary. It also keeps the paper trail short enough that anyone can read it a year later.

How Do You Tell a Real Answer From a Polite One?
You will know a real root-cause answer the way you know real work anywhere: it is specific, it is slightly boring, and it admits a mechanism.
| Polite answer | Real answer |
|---|---|
| “We have strengthened our quality control.” | “The drying step for that lot ran short; the operator log now requires a recorded drying time before winding.” |
| “It was an unusual batch.” | “The masterbatch for that colour changed supplier in that month; we have moved it back and added a first-spool print check for colour changes.” |
| “We will pay more attention.” | “Our gauge flagged drift after that run; it is recalibrated and the alarm limits are tighter.” |
Notice what the real answers have in common. Each one names a step, a change, and something you could look for on the next delivery. That last part is where the real check happens, and it is the step almost nobody takes.
How Do You Verify the Fix on the Next Order?
Verification is not an audit. It is one deliberate look.
When the next shipment arrives, print the same test that failed before, from the same position in the order if you can. If the fix was real, the failure mode is gone. If it recurs, you now have two documented events with lot references, which changes the conversation from “please look into it” to “here is the pattern”.
For failures that carried mechanical weight — brittle parts, layer splits — you can ask that the corrective check be run to a published method, tensile testing to ISO 527-1 among them, so the before and after numbers can actually be compared. The standard-language approach is the same one our quality control process uses for incoming material. Numbers with a named method travel; adjectives do not.
Write the result next to the claim in your records. One line: claim, cause stated, fix verified or not. Over a year, that file becomes the most honest supplier review you own.
And if the next order is better but not clean? Say so, with both results attached. Partial fixes are common — a drying problem solved while a winding issue remains — and they are only visible to a buyer who checks the same test each time. Suppliers respect the second data point more than the first, because it proves somebody is watching.
When Does a Claim Pattern Become a Sourcing Decision?
One claim with one good fix is a supplier working properly. Nobody’s line is perfect; the question is whether the loop closes.
The pattern to watch is repeats of the same failure class, or causes that keep being “unusual batches”. At that point the claim file stops being customer service and becomes procurement data. It belongs in the conversation you have before the next commitment, alongside the other checks described in How to Buy 3D Filament in Bulk. A supplier who fixes things once each is worth keeping. A supplier whose fixes do not hold is a price you are paying in a currency you did not agree to.
So keep the record in a shape you can read later. One line per event, five columns: date, lot, failure, cause stated, fix verified or not. A table like that takes thirty seconds to fill after each claim, and once a year it answers the only question that matters at renewal time: is this supplier getting better, standing still, or drifting?
True or False: Claim Beliefs
True or False: The goal of a quality claim is compensation.
False. Compensation closes the invoice. The goal of a claim is the cause and the fix, so the same failure does not return with the next order.
True or False: Suppliers never tell buyers the real root cause.
False. Most factories investigate internally anyway. Buyers who ask a specific three-part question (finding, cause, change) usually get a real answer, because the answer already exists.
True or False: A fix should be trusted once the supplier confirms it.
False. A fix is trusted when the next shipment passes the same test that failed before. One deliberate look beats any reassurance.
FAQ
What should I send with a filament quality claim?
Photos of the spool label and lot reference, the failure described in plain terms, your printer and settings for that print, and how many spools are affected. That package lets the factory find the lot and start a real investigation.
How long should a root-cause answer take?
A first answer with a named cause and change is reasonable within days; deeper checks can take longer. What matters is that the answer names a step and a change, and that you verify it on the next shipment either way.
What if the supplier only offers a credit note?
Take it, and keep asking the three-part question in writing. If the answer never improves, the pattern itself is your data for the next sourcing decision.
Do I need to re-test every shipment after a fix?
No. One deliberate check on the next delivery, using the test that failed before, is enough. After that, normal incoming checks carry the weight.
Who should own this loop on our side?
Whoever answers the customer complaint — usually the same person who filed the claim. The loop only has three moves: ask the three-part question, check the next delivery, write one line. It does not need a quality department. It needs someone who remembers to close the file.
Conclusion
A claim is a question, and a credit note is not an answer. Ask for the finding, the named cause and the change. Tell the polite non-answers from the real ones by whether they name a step you can check. Then check, once, on the next shipment, and write the result down. Buyers who close this loop stop paying for the same failure twice, and their supplier lists quietly sort themselves.
Tell us the material, volume, packaging and destination — we reply within 12 hours. Reference MOQ from 100+ spools (≈100 kg) for standard colors.
Discuss Your Project