Hardware is the part of a window that gets asked to work. Everything else holds still. Which is why, when a window fails, it is almost never the aluminium that failed — and why the best hardware choice is not the best brand, but the one your country can still buy in year six.
A window with dead hardware is not a window. It is a translucent wall.
It still lets light in — that is the one property it keeps. But it no longer compresses its seals, so it drafts. It no longer holds the sash true, so water finds the gap. It no longer locks, so the security you paid for is a handle you can no longer turn. A wall with a hole in it lets in less than that, and costs less to fix.
In my experience, hardware is the highest-frequency failure item across everything we inspect and everything that goes wrong after handover. Not profiles. Not glass. Hardware. And most buyers, when sourcing from Foshan, spend their attention everywhere except the one line in the quotation that deserves it.
Why hardware fails first — and why it fails quietly
Window hardware is every moving or locking part of the window: the hinges and friction stays, the multi-point lock and its strikers, the handles, and the gear that drives a sliding sash.
Aluminium does not operate. Glass does not operate. Hardware operates — every day, by the same hand, for thousands of cycles, carrying the weight of the sash while it does it.
But the real mechanism is subtler than "moving parts wear out", and it is worth understanding because it changes what you check:
The window's performance is the hardware's job. A multi-point lock is not primarily a security device. It is the thing that pulls the sash against the frame so the seals compress. Air tightness, water tightness, thermal performance — the numbers in your test report — are held by the lock points and the hinges, not by the profile. A hinge that has sagged by a millimetre or two — less than you would notice on site — is enough to stop the seal compressing. The glass is still perfect. The Low-E is still working. The window still "closes". It no longer seals.
The failure is gradual, so there is no event. Nobody tells you the hardware failed. The condensation line gets a bit wider, the handle gets a bit stiff, the draught gets a bit worse. By the time it is obvious, the frame has been carrying load it was never built to carry.
This is the same structural problem as the thermal break — the part that matters is doing the work invisibly, and the catalogue cannot show you it wearing out. (PA66 vs PVC: How to Check makes the parallel case for the strip inside the profile; why a frame leaks heat where nobody checks makes it for the places a sample never gets cut.)
The three shelves you are choosing from
When you buy from China, the hardware line either says a brand or it says nothing — "standard export hardware", "high quality accessories". Saying nothing is the default, and defaulting is how you end up on the wrong shelf.
A shelf, here, means a class of maker defined by who can service the part where the window stands — not by price. There are three:
One: the international brands — Roto, Siegenia and G-U for the mechanisms, Hoppe for handles, and the German-system makers that supply alongside them. This is the safe answer. It is also the answer every salesperson already knows is the safe answer, which is why "we use German hardware" costs nothing to write.
Two: the leading Chinese brands — KinLong, HOPO, CMECH and a short list beside them. These are not consolation prizes. They are the hardware behind the largest volume of windows built anywhere on earth, at price points that make the import arithmetic work. They have been tested by the biggest building programme in the world, under the roughest conditions, and the good lines are good.
Three: unnamed or unfamiliar makers — including, occasionally, genuinely well-made hardware from factories whose brand has no marketing department. It exists. Buying it takes the ability to tell it from junk.
Most advice stops here and tells you to pick shelf one. That is the wrong question.
The question that actually decides it: what can you buy in year six
A window is not installed once. It is serviced for twenty years. Handles snap. Hinges sag. A multipoint gearbox needs swapping. The question is never only "is this brand good" — it is "on the day it fails, who within reach of your site has this part in stock?"
So the order I give buyers is this:
First: a brand that is common and trusted in your own country. If that brand is also an international one, and the Chinese factory can actually supply that brand at a sane price — that is the best case on the board. You have the brand's quality, and you have your own country's spare-parts ecosystem behind it: the merchant who stocks it, the installer who has serviced it a hundred times, the datasheet in your language.
Second: when the best case is not available, the leading Chinese brands. Their quality is proven at scale, the price gap against the German names is real, and — this is the part nobody says out loud — their geometry and mounting patterns are the de facto standard the profiles are built around. When something needs replacing, your agent in Foshan can have the part in your next container. You are trading a local shelf for a supplier you trust to source it. That is a fair trade if — and only if — you chose shelf two deliberately, not by default.
Third: an unknown maker, and only if you can read hardware. You can specify a good unnamed hinge the way you can specify a good unnamed thermal-break strip. But ask yourself honestly whether you can tell it from a bad one at your receiving dock — on an assembled window, through a coating, at four in the afternoon. If you cannot, that shelf is not priced cheaper. It is priced with a hidden surcharge you pay in year six.
The wrong answer is the one most quotations give by default: a brand name nobody in your country stocks and nobody in your country can service. A Roto that cannot be bought locally and a no-name that cannot be identified are, on the day something breaks, the same thing.
And while we are on year six: don't buy tomorrow's window
A small piece of window-buying common sense, free with the article:
Do not choose a newly launched, unusual design — especially one whose opening method is clever and complex.
Novel mechanisms are, by definition, mechanisms your country's installers have never serviced. Even the parts themselves can be air-freighted from China in days — but air freight does not ship the person who can fit them. The first installer who has actually seen your window's weird tilt-slide-pivot may be the factory's own engineer, at your cost and on your schedule. And the repair bill can easily exceed what the window cost.
Boring designs are boring for a reason: twenty years of hands have made them serviceable. An ordinary tilt-and-turn is a window your own market can keep alive. A clever new shape is a window married to one factory's tooling — and when that factory moves on, the marriage has no exit.
The most reliable "new" thing in a window order should be the glass coating. Everything else earns its novelty by existing in your market for a decade.
Five questions to send as they are, in writing, before you pay
Copy these five into your next email, unchanged:
- Which hardware brand, series and model are quoted — written out, not "standard export hardware"? Which of those are the hinges or stays, which is the lock, which are the handles, and which is the gear on the sliding door?
- Is this brand stocked in my country? Name it, then check: find two suppliers or service agents for it on your own side of the water. If you cannot find them, go back and renegotiate the shelf — not the grade.
- What is each component's load rating against my sash weights — the frame hinges for the heaviest sash, the gear for the largest sliding panel? (The mechanism that carries the weight is where "good brand, wrong part" hides.)
- How many lock points and strikers per sash, and on which edge? Wear on a two-point sash is not repaired like a four-point one.
- Add the spares to the order now: a surplus kit — one handle set per series, a pair of hinges, a striker set, screws — packed in the container. Finding matched metric hardware in your local shop at year three is a nightmare; putting it in the container at year zero is pennies.
A supplier who has real answers will send them in writing. One who does not will keep the phrase "high quality hardware", which is not a specification. Either response is information — the same test we apply to everything else in a quotation. (The WeChat Test has the version of this for the company behind the quote.)
Frequently asked questions
Is more expensive hardware automatically the safer choice? No — the safer choice is the hardware you can actually service where the window stands. An international brand with no local stock is, operationally, an unbranded window with a nicer nameplate.
Should I insist on European hardware on a Chinese profile system anyway? Not automatically. Quality mechanisms from the leading Chinese makers are designed around the extrusion geometries the profiles use. A mismatched premium part inside a correct cheap one is the wrong trade. Choose the brand first, then confirm it is genuinely specified at the part level. And keep the CE question where it belongs: CE marking is declared for the finished window, not for the name on the handle — what EN 14351-1 actually requires.
Can't I just air-freight a replacement part when something fails? The part is the easy half of a repair. The person who services your specific system, with your specific geometry, locally — that is the scarce item. Which is why this entire article comes down to the one question: what exists, nearby, in year six.
We do not sell windows, we take no commission or rebate from any factory, and your payment never passes through us. If you want a quotation's hardware line read properly — brand, series, load rating, and whether your own country can service it — send it over. That is our work.