We have the sleeves. Two people disagree about the same bright band.

Get those two people looking at the same sleeve the same way. One lamp right against the surface and one glance from the aisle can produce a very long argument about very little shared evidence.

Before photos exist, but one sleeve number is missing. I'm leaving that pair out of the comparison.

Keep the unidentified sleeve aside rather than binning it. Someone may recover its identity, but it shouldn't quietly become a known-good starting sample because the next pad trial needs one.

David's question also helps distinguish incoming finish variation from new contact marks. Keep both in the sample selection once quality has identified them.

I wouldn't order a pile of soft pads yet. We tried a more compliant contact on a cosmetic part and liked the first marks less badly, then found it held on during release. It wasn't a useless trial; it told us the easy-looking change moved the problem. Your supplier should show the sleeve leaving the contacts and sitting properly in the receiving tray, not just a nice surface photograph.

Quality accepts the bright band but rejects two new scuffs at the contact edge. Their method is agreed now.

Does the supplier have those actual scuffed sleeves to compare with the contact edge?

Yes. They found a sharp edge on one sample insert, unlike its drawing.

Then identify that insert separately. Stores needs to know whether a revised drawing is coming or an incorrectly made part is being replaced to the existing one.

And repeat the original comparison with the corrected insert before combining it with a different pad material. Otherwise you won't know which change helped.

Agreed on changing one thing here. My soft-contact story isn't a reason to keep a wrongly made edge.

Replacement is to the existing drawing. Supplier has kept the proposed softer pads out of this repeat.

Has the repeat included release into your actual tray?

Yes. No new edge scuffs on the identified sleeves, but one hangs briefly before releasing.

Useful surface result, and a real release problem to investigate. Give the supplier the affected sleeve identity and the complete sequence; don't improve the surface comparison by omitting that attempt.

Kept it in. Softer-pad purchase is on hold; supplier is investigating the delayed release with that sleeve.

I've seen the successful surface photograph become the only picture retained. Keep the release clip attached to the same trial identity, so purchasing doesn't receive half the result.