We rejected three batches of semi rimless glasses last quarter. Not because we were being picky. Because every batch had passed the factory's own QC, been signed off, and then arrived at our warehouse with loose nylon cords. Groove depths that didn't match the spec sheet. And on half the pairs, the left temple was visibly misaligned.
I handle quality and brand compliance for an optical wholesaler. Roughly 200+ SKUs a year—maybe 220, I'd have to check the system. In 2024, our first-delivery rejection rate sat around 17%. Almost all of it was half-rim.
If you're doing semi rimless glasses sourcing right now, or you're about to place an OEM order for half rim glasses, you've probably hit some version of this wall. This post isn't about "how to pick a supplier." That part is easy. I want to talk about why semi-rimless is structurally harder than full-rim—and what we changed to bring our rejection rate under 6%.
The Surface Problem Looks Like a Factory Issue
The first instinct is to blame the factory. Everything I'd read about procurement said to get multiple quotes, compare pricing, screen vendors. Sounds reasonable—most sourcing advice says the same thing. In practice, with half rim glasses, the opposite was true: the more quotes we collected, the more the quality dropped. Because no two factories meant the same thing by "half rim."
That's the outer layer. It looks like "this factory is bad." It's actually an RFQ problem.
The Deeper Problem: Semi-Rimless Isn't Full-Rim Minus a Wire
Most buyers treat semi-rimless glasses as full-rim frames minus one strip of metal. Same material, same lens, just less. So the quote should be lower, the process should be shorter, the risk should be smaller.
It's backwards. Semi-rimless adds two manufacturing steps that full-rim doesn't need: lens grooving and nylon cord threading. A full frame wraps the lens completely—the edge is essentially hidden. A half frame exposes the lens edge to the end customer. Groove position, depth, and cord tension are all visible on the shelf.
Here's something vendors won't put in the quote: the quality of those two steps has less to do with a factory's equipment and more to do with how worn the grooving wheel is that day, and how much tension the threading operator applies by feel. Neither shows up in a price sheet. Neither shows up in a sample order either—samples get made by the senior hand. Bulk production runs on a different shift.
We also hit a communication failure early on. First RFQ round, I wrote "semi rimless eyeglasses." They quoted back "half frame." When the bulk order landed, I realized their "half frame" meant a convertible structure—wear it full, snap off the bottom wire. One term, two meanings. We were using the same words but meaning different things. Discovered it when the lenses arrived and nothing matched our drawings, two weeks too late.
So the real variable with semi rimless glasses isn't "factory capability." It's spec control. Whoever writes the spec tight wins.
What It Costs When You Don't Fix It
Rejecting a half-rim batch has a visible cost (rework) and three hidden ones.
First: air freight for reorders. Ocean is roughly 30 days, air is 5–7. The gap runs $4,000–$8,000 per shipment depending on frame count and pack volume (based on our forwarder quotes, Q4 2024; verify current rates with your carrier).
Second: shelf dates. Our customers are optical retailers and distributors. Retail slots get booked in advance. Miss by two weeks and the slot is gone.
Third: return rate. This is the one that hurts quietly. After semi rimless glasses reach end consumers, snapped nylon cords become the single most common warranty claim. On a 2,000-pair run, expect 3% to 5% back by month six. Honestly, that third number is the one that kills a brand. The first two are money. The third is reputation.
What Took Me Three Years to Figure Out
It took about 400 half-rim orders and three years to understand that the root cause isn't "factories can't do it." It's that we weren't saying what we meant.
Semi-rimless doesn't tolerate ambiguity the way full-rim does. A full frame with a slightly off groove still has wire wrapping the lens—it gets caught. A half frame has no backstop.
The turn for us wasn't changing suppliers. It was turning the spec into something you can verify against. After that, half rim glasses stopped being a coin flip.
The Four Things We Changed
Simple to write down. Every one of them came from a bad batch.
- Three numbers locked into the RFQ. Groove angle (typically 110°–120°), groove depth (calc'd off lens edge thickness, usually 0.60–0.80 mm), and cord spec (most commonly 0.80–1.00 mm). No "industry standard" language. On semi-rimless there is no universal standard.
- Destructive testing on the first 50 pairs. Pull the cord to failure, log the breaking force. Anything under 15 N gets returned. (We started at 20 pairs—sample size was too thin, bit us twice before we raised it.)
- Same supplier for two consecutive quarters. Half-rim process stability accumulates. Switching vendors resets wheel wear and threading habits to zero.
- Third-party inspection pre-shipment, not on arrival. Rejecting at port means the air freight is already spent. Catching it before it ships costs you a week.
One more thing worth flagging: if your brand describes product as "half rim" or "rimless" in marketing, that's a product specification claim. Per FTC advertising guidance (ftc.gov), claims have to be truthful and substantiated. So writing the spec tight isn't just a procurement habit—it's a compliance habit.
After we put those four controls in place, our first-delivery rejection rate dropped from 17% to roughly 5–6%. The rest is color matching and temple assembly, which is a different problem entirely.
Bottom Line
Semi-rimless looks like a niche category. It is. But plenty of people are in it. The real differentiator isn't who finds a cheaper factory—it's who writes a spec the factory can't hand-wave past. Semi rimless glasses for men, women's frames, unisex styles—the SKU doesn't matter much. What matters is whether you put groove angle, cord tension, and pre-shipment acceptance criteria on paper before the PO goes out.
Do that, and the rest is just normal production variance.