Entry Door Lock & Deadbolt Inspection China: ANSI Grade, Latch Function & Finish QC for FBA Importers
The Lock That Wouldn’t Latch A seller’s best-selling entry set — a keyed deadbolt plus knob combo at a very competitive price — started generating returns nobody could explain.
The Lock That Wouldn’t Latch
A seller’s best-selling entry set — a keyed deadbolt plus knob combo at a very competitive price — started generating returns nobody could explain. Buyers said the door “wouldn’t stay closed” and “opens with a push.” The returned locks worked fine on the seller’s bench. The problem was invisible in the US: the latch spring was just weak enough that the deadlatch didn’t hold, and it only showed on doors with a slightly tight strike fit. 🔐
Door locks are a trust product — buyers do not forgive a lock that fails, and the reviews are brutal. A China inspection service can’t test every lock, but it can test the right sample hard enough to find the batch problem before 8,000 sets sail. Here is what we check on entry door sets and deadbolts.
Why Cheap Entry Sets Fail
Most China-made entry sets use zinc-alloy (Zamak) bodies — fine for residential use when cast properly, a problem when the alloy is thin or porous. The real failures cluster in four places: the latch assembly (weak springs, thin bolts), the deadbolt throw (short, misaligned, stiff), the strike plate (soft metal, short screws), and the finish (peeling within months). Add a fifth: the ANSI grade stamped on the box often doesn’t match the hardware inside. ❌
| Claimed grade | Typical use | Cycle expectation | What we actually see |
|---|---|---|---|
| Grade 3 | Standard residential | ~25,000 cycles | Most common import grade; check latch steel and springs |
| Grade 2 | Heavy residential / light commercial | ~100,000 cycles | Frequently overclaimed — verify cylinder and bolt construction |
| Grade 1 | Commercial | 400,000+ cycles | Rare on FBA; requires real lab certification, not a stamp |
Grade is a design and durability claim, not a visible feature — which is exactly why the stamp gets borrowed. We treat it as something to verify, never to assume. ⚠️
The 8-Point Entry Set Checklist
| # | Check | What we verify on site |
|---|---|---|
| 1 | Body material & weight | Zinc vs brass vs steel per spec; weight and casting quality; no porous or thin-wall bodies |
| 2 | Latch & deadlatch function | Bolt throw, spring tension, deadlatch holds the latch in — 100+ open/close cycles |
| 3 | Deadbolt throw | Full 1-inch throw, smooth rotation, no binding at 200 cycles |
| 4 | Key & cylinder | 50 insert/rotate cycles each side; pin count matches marketing; keys turn smoothly |
| 5 | Strike plate & screws | Plate gauge, screw length (3-inch screws for the studs), finish match |
| 6 | Handing & backset | Left/right handing per SKU; 2-3/8 vs 2-3/4 backset adjustable as listed |
| 7 | Finish | Salt-spray check, rub test, edge coverage — no peeling, fading or bare spots |
| 8 | Kit completeness | Keys, screws, latch, strike, template, English instructions — counted per set |
Function defects are the dangerous class here. On the AQL scale, a latch that won’t hold is Critical — it is a security failure, not a cosmetic one — so it sits at AQL 1.0, while finish issues run at Major AQL 2.5 and Minor at 4.0. You can run the sample math in our AQL calculator before booking. 🔑
The Case: 8,000 Entry Sets
The seller above had ordered 8,000 sets — $37,600 — from a Zhejiang factory that had delivered three good orders. On the fourth, at the pre-shipment inspection with an AQL 2.5 sample of 200 sets, we found 6% with weak latch springs (the deadlatch didn’t hold — the exact return reason he was seeing), 4% with deadbolts that bound after 50 cycles (misaligned bolt guides), 3% missing the screw pack, and 2% with finish peeling on the rose edges.
The factory reworked under supervision: springs swapped, bolt guides aligned, screw packs completed, 200 sets re-cycle-tested. The rework ran $2,400. The alternative — 8,000 sets with a 6% latch failure — was a return wave, a security complaint (the liability kind), and a dead listing. The seller estimates $11,000+ avoided, and the corrected order passed at 1.1% majors. 📉
How to Spec a Lock Order So It Can Pass
Half the lock failures we catch are spec failures — the factory delivered exactly what the PO asked for, and the PO didn’t ask for enough. Four lines in your spec prevent most of them.
Declare the grade honestly. If you sell to homeowners, a Grade 3 with verified construction is perfectly fine — writing “Grade 2” on the box when the hardware inside is Grade 3 is how listings and trust die together. State the grade you will market, and let inspection verify the hardware matches it.
Define function, not just looks. Specify a minimum deadbolt throw (1 inch), latch spring behavior and a cycle expectation — “500 open/close cycles without binding” gives the factory and the inspector the same target.
Name the finish, not the color. “Oil-rubbed bronze” and “satin nickel” describe a specific coated surface; “dark brown” describes nothing. Put the finish name from your listing on the PO so the golden sample becomes the referee when a batch comes out shinier or darker.
List the kit contents. Keys per lock, screw lengths, latch, strike, template, manual — every item with a quantity. An entry set that arrives with three screws instead of six fails before it is installed, and the count is only enforceable if the PO said six. 📝
Packing & FBA Readiness
Locks ship with a lot of loose hardware, and FBA has three rules that bite: keys and screws must be secured inside the product box (buyers hate rattling bags of loose parts), every polybag over five inches needs a suffocation warning, and cartons must stay under 22.7 kg. We also drop-test sampled cartons at 60 cm — a dropped corner chips finishes and bends strikes, and a bent strike plate becomes a “won’t latch” return before the buyer ever installs it. ✅
The Bottom Line
Entry sets are a high-trust, high-return-risk category where a $169 inspection visit protects a $37,000 order — and a listing that took a year to build. We inspect locks and building hardware daily: 2,000+ inspections a year, reports within ~24 hours, pricing from $169 per man-day.
If your lock order is in production or about to be, contact us with your PO and we will run the 8-point checklist — latch cycles, bolt throw, keys, strike and finish — before it sails. 🔒
FAQs
Are Chinese factories actually ANSI/BHMA certified?
A few are. Most stamp “Grade 2” or “Grade 3” on the box based on the design they copied, not on lab testing. On inspection we verify the stamp exists, compare the construction (latch steel, springs, cylinder) to what the grade implies, and run our own cycle test on samples. If a seller needs a real certified grade for commercial buyers, we check the lab certificate against the model number.
Mechanical or smart lock — do the checks change?
The mechanical core stays the same: latch, deadbolt, strike, handing, finish. Smart locks add electronics, so the checklist grows — FCC/UL documentation, app pairing, keypad durability, battery life, and the manual override that has to work when the electronics die. This article covers the mechanical entry set; smart locks get the electronics module added on top.
How many keys should an entry set include?
Two or three keys per lock is the US standard, and the bitting (the cut pattern) should be even, not a shallow cut that wears out. We insert and rotate every sampled key 50 times to catch sticking. If you sell keyed-alike pairs — front door and back door using one key — we verify the pairs actually match.
Do security pins matter for a residential lock?
They help against picking, and they are a cheap differentiator on Amazon listings — “5-pin anti-pick cylinder” sells. What we verify is honest marketing: open the cylinder, count the pins, confirm the security pins are actually there. We have caught “anti-pick” cylinders with plain driver pins more than once.
What if my listing sells both left-handed and right-handed doors?
Handing mistakes are a top return reason for entry sets. Most modern deadbolts are reversible in seconds, but knob and lever sets are handed. On inspection we verify the handing configuration matches each SKU’s listing, and we flag any set where the latch is not reversible when the listing promises universal fit.
Frequently asked questions
Are Chinese factories actually ANSI/BHMA certified?
A few are. Most stamp “Grade 2” or “Grade 3” on the box based on the design they copied, not on lab testing. On inspection we verify the stamp exists, compare the construction (latch steel, springs, cylinder) to what the grade implies, and run our own cycle test on samples. If a seller needs a real certified grade for commercial buyers, we check the lab certificate against the model number.
Mechanical or smart lock — do the checks change?
The mechanical core stays the same: latch, deadbolt, strike, handing, finish. Smart locks add electronics, so the checklist grows — FCC/UL documentation, app pairing, keypad durability, battery life, and the manual override that has to work when the electronics die. This article covers the mechanical entry set; smart locks get the electronics module added on top.
How many keys should an entry set include?
Two or three keys per lock is the US standard, and the bitting (the cut pattern) should be even, not a shallow cut that wears out. We insert and rotate every sampled key 50 times to catch sticking. If you sell keyed-alike pairs — front door and back door using one key — we verify the pairs actually match.
Do security pins matter for a residential lock?
They help against picking, and they are a cheap differentiator on Amazon listings — “5-pin anti-pick cylinder” sells. What we verify is honest marketing: open the cylinder, count the pins, confirm the security pins are actually there. We have caught “anti-pick” cylinders with plain driver pins more than once.
What if my listing sells both left-handed and right-handed doors?
Handing mistakes are a top return reason for entry sets. Most modern deadbolts are reversible in seconds, but knob and lever sets are handed. On inspection we verify the handing configuration matches each SKU’s listing, and we flag any set where the latch is not reversible when the listing promises universal fit.