What Should You Actually Check During a Supplier Audit in China?

Every importer who has been burned by a bad factory asks the same question — but usually too late. What should I actually check during a supplier audit in China? A supplier audit is the single highest-leverage step in the entire China sourcing process: a two-day factory visit that can save you from a six-figure bad order, a customs seizure, or a reputational scandal in your home market. China shipped roughly $3.4 trillion worth of goods in 2023, according to UN Comtrade data compiled by Statista — most of it from factories most buyers never set foot in. The importers who thrive are not the ones with the best suppliers — they are the ones who audit before they trust. This guide walks you through exactly what to check, in what order, with the documents to demand, the numbers to compare, and the scoring system to use, whether you audit yourself, hire a third-party firm, or lean on a sourcing agent in China. If you want the short version first, our supplier audit checklist is free to download.

1. Audit Types Explained: Matching the Audit to Your Sourcing Strategy

The first mistake most buyers make is treating “audit” as one thing. It is not. A supplier audit in China is a family of checks with different goals, different costs, and different evidence standards. Pick the wrong type and you will either overpay for information you did not need or — worse — feel falsely safe after the wrong kind of pass. Here are the five audits that matter in China sourcing, and how to fit them into a coherent sourcing strategy.

Quality management audits

A quality audit checks whether the factory has systems that produce consistent output: incoming material inspection (IQC), in-process checks (IPQC), final inspection (FQC), calibration of measuring tools, defect tracking, corrective action procedures, and a documented quality manual. This is the audit you run when you care about product consistency, rework rates, and whether the factory can hold a spec from batch one to batch ten thousand. The gold standard reference is ISO 9001, but here is the catch: roughly half the ISO 9001 certificates you will see in China are issued by obscure bodies that check little more than a photocopied manual. A quality audit verifies the system behind the certificate, not the certificate itself.

Social compliance audits

Social compliance audits check working hours, wages, overtime, social insurance, health and safety, dormitories, fire safety, and child or forced labor. The major protocols are BSCI (the amfori Business Social Compliance Initiative), SMETA (the Sedex Members Ethical Trade Audit), and WCA (the Walmart Compliance Assessment historically required by Walmart). These matter for three reasons: your customers demand them, your law firm advises them, and — honestly — because the worst factories in China hide the worst problems behind the best sales pitches. Compliance audits are where the “everything is fine” narrative dies.

Capacity, financial, and full audits: matching depth to risk

A capacity audit answers one question: can this factory actually make my volume, to my spec, by my deadline? You check machine counts, machine ages, actual utilization, shift structures, headcount, bottleneck processes, and — critically — who else’s orders are already in the pipeline. A factory that says it can produce 50,000 units a month while running two shifts at 95% capacity for other clients is lying to you, and a capacity audit catches the lie before you wire the deposit.

A financial audit verifies the legal entity: business license, export rights, tax status, the registered capital versus the actual operation size, litigation history, and the real ownership structure — because in China the person who signs your contract is often not the person who owns the factory. Credit checks on Chinese suppliers are available through agencies like Sinosure, credit bureaus, and specialized due-diligence firms. This is a non-negotiable step before you place a high-value order or agree to payment terms beyond 30% deposit.

The full audit combines all of the above into one multi-day engagement, usually with a checklist of 150–250 individual checkpoints. This is what a serious sourcing agent arranges for a buyer who is about to commit serious money. Expect to pay roughly $1,500–$4,000 for a full third-party audit in China, depending on size and location — expensive until you compare it with a failed order, which routinely runs five figures. Here is the data that should shape your sourcing strategy before you book a flight. QIMA, the quality control and inspection firm, reported first-time product inspection failure rates in China of roughly 30% for toys and about one in four across consumer hard goods in its 2023–2024 quality data — and those failures happened in factories that had already survived a sales pitch, a price negotiation, and a sample round. The sample lies to you; the audit tells you why. Matching audit depth to order value is the practical version of this rule: for orders under $10,000, a document screen plus questionnaire is enough. Between $10,000 and $50,000, add a one-day quality or compliance audit. Above $50,000, run the full protocol — the $1,500–$4,000 it costs is 3–8% of the order, which is the cheapest insurance you will ever price.

The cost side deserves equal attention. Most buyers budget for the audit itself but forget the corrective action cycle: findings that take 30–90 days to fix, re-audit fees, and the management time consumed on both sides. A factory that needs three rounds to fix basic document control is telling you what your ongoing relationship will cost — in months, not money. Budget 5% of the project for verification — audits, inspections, and follow-up — and you will never discover the truth about a factory inside a container.

The right strategy for most importers: run a background and document screen on every shortlisted factory, then a full audit on the final two or three, then quality inspections on every shipment once production starts.

2. Pre-Audit Preparation: Documents, Questionnaires, and Groundwork

A supplier audit in China does not start when you walk through the factory gate. It starts two weeks earlier, when you send documents and questions the factory must answer in writing. The factories that stall, dodge, or “forget” to reply are telling you something before the auditor even arrives. Preparation is where you separate professional buyers from amateurs — and it costs you nothing but a few hours.

The document pack you demand before anyone visits

Before scheduling anything, request these eight items by email and put a deadline on them: (1) the business license (营业执照), (2) export license and customs registration, (3) a full ISO or industry certificate set with issuing bodies named, (4) the most recent BSCI, SMETA, or equivalent audit report, (5) an organization chart with the general manager’s name and the plant manager’s name, (6) a factory floor plan showing production, warehouse, QC lab, and dormitory zones, (7) your product category’s test reports or CCC certification where relevant, and (8) a list of current major clients. Then watch what happens. A legit factory with nothing to hide sends everything within 48 hours, often with a polite cover note. A factory that asks “why do you need this?” three times before sending partial scans is a factory that has been trained to hide — and no walkthrough will undo that.

Here is the trick experienced auditors use: send the document request to the general manager, not the salesperson, and copy the plant manager. The salesperson is paid to make you happy. The plant manager is paid to run a factory, and his email habits tell you how the operation actually runs. You are also testing responsiveness — the same responsiveness that will matter when you need a delivery date moved or a defect corrected mid-production. If document turnaround is slow now, it will be catastrophic later.

The questionnaire that does half the work

A written pre-audit questionnaire of 30–50 questions forces the factory to commit to numbers in writing before anyone visits. Ask for: monthly production capacity by product type, current capacity utilization percentage, number of production lines and shifts, worker headcount by department, overtime practices (average hours per worker per week), wage structure (base plus piece rate), social insurance enrollment percentage, quality defect rate over the last 12 months, on-time delivery rate, current order backlog in weeks, and a list of the three largest customers with approximate order values. Then compare their answers with what you find on the floor. In my experience, 60–70% of factories overstate capacity utilization — they claim 80% utilization to look busy, or claim 60% to look available; the questionnaire gives you the baseline to test either way. Discrepancies between the written questionnaire and the floor visit are not trivia — they are the entire point of the exercise. Keep the questionnaire answers on one page per factory, in a column: capacity claims, headcount claims, defect claims. When three factories claim 95% on-time delivery and one claims 72%, you are not reading a survey — you are looking at the shape of your future supplier list.

Scheduling, logistics, and the kickoff meeting

Schedule the audit for a Tuesday, Wednesday, or Thursday, mid-month, during a normal working day — and tell the factory the arrival time is 8:00 a.m. Monday visits are useless because Monday is when Chinese factories hold meetings, run safety drills, and tidy up. End-of-month visits catch factories in payroll chaos. Arriving at 8:00 a.m. instead of 10:00 a.m. means you see the shift handover, the morning briefings, and whether the line is actually staffed. Never let the factory pick the day, and never accept a “surprise drill” invitation — a factory that insists on a specific date has scheduled something for that date, and it is rarely your benefit. Before the walkthrough, hold a 20-minute kickoff meeting: introduce the auditor, state the scope, list the documents you will inspect, and ask for a single escort plus one manager per zone. Two more rules that separate pros from tourists: take your own photos with timestamps, and never accept food or gifts from the factory during the audit — it compromises the report, and the factory knows it. One more element belongs in pre-audit preparation: your interpreter and your auditor’s independence. If you audit in person, brief the interpreter on the checklist in advance — a good interpreter translates intent, not just words, and flags when the plant manager’s answer quietly contradicts the paperwork. Put the audit fee and the report deliverable in writing before the visit: fixed fee, full report within ten working days, no consulting add-ons. A supplier audit is a buying decision disguised as a site visit, and every hour you invest before the gate doubles the value of the six hours you spend inside it.

3. The Factory Floor Walkthrough: A Room-by-Room Checklist

The floor walkthrough is where supplier audits are won and lost — and it is the part most buyers do badly, because they walk too fast, look at the wrong things, and let the escort steer them. A good walkthrough takes four to six hours and follows a fixed route that the factory does not control. You start at the gate, not the showroom. Here is the seven-step walkthrough checklist I use on every audit.

The seven-step walkthrough checklist

Step 1: Arrive early, observe the gate and the parking lot. Note how many trucks are loading, what brands are on them, how security logs visitors, and whether workers are streaming in or out at odd hours. A factory loading trucks for other brands is busy — but if every truck is unmarked or the loading dock is empty at 9 a.m. on a Wednesday, the “capacity” you were quoted is fiction. Why this works: the gate is the only part of the factory the management cannot rehearse; the loading dock is where real production volume shows itself.

Step 2: Walk the production floor against the flow of production. Start at the finished-goods end and walk backward to raw materials, not the other way around. Count machines, note which ones are running, read the machine nameplates for age, and check the production schedule board — the whiteboard that lists today’s orders, quantities, and due dates. Why this works: walking against the flow prevents the escort from steering you past the finished-goods staging area, and the schedule board tells you who the factory is actually producing for today, in quantities you can verify against their questionnaire claims.

Step 3: Inspect the QC lab and measuring equipment. Find the inspection room, check whether gauges and calipers have calibration stickers with dates, and ask to see the last three IQC reports for incoming materials and the last five FQC reports for finished goods. Why this works: the QC lab is the factory’s quality control China nerve center — if calibration stickers are expired and reports are handwritten with no signatures, no ISO certificate in the world makes this a quality factory.

Step 4: Check the warehouse and raw material storage. Look at how materials are labeled, whether first-in-first-out is actually practiced, whether moisture-sensitive goods are stored off the floor, and whether there is a separate area for defective materials — and whether it is actually used. Why this works: warehouse organization predicts shipping accuracy; factories that cannot organize a shelf cannot organize a container, and the defective-materials area tells you whether defects get quarantined or quietly shipped.

Step 5: Examine worker facilities — toilets, canteen, dormitory, first-aid. Open the toilet doors, look at the canteen menu, count beds per room in the dormitory, check for fire extinguishers with current inspection tags and unobstructed exits. Why this works: this is the fastest social compliance signal on the floor — a factory with clean, safe worker facilities is a factory run by management that cares; one with padlocked fire exits is one customer audit away from losing your order and your reputation.

Step 6: Study the shift pattern and the time-clock. Ask when shifts start and end, count workers on the floor versus the headcount on the org chart, and photograph the time clock or biometric reader. Why this works: the gap between declared headcount, actual bodies on the floor, and time-clock records is where working-hours violations hide — the single most common finding in Chinese factory audits.

Step 7: End at the shipping and finished-goods area. Look at packing, labeling, pallet condition, and whether finished goods carry the buyer’s marks of other brands. Ask how shipments are arranged and who does export documentation. Why this works: the finished-goods area is where the factory’s export readiness lives — and if you see another buyer’s branded cartons sitting there, you now know exactly how much of the factory’s capacity is already committed.

Reading the room: behavior tells you more than bricks

Beyond the physical checks, watch the behavior. Does the general manager stay with you the whole time or disappear? Does the escort answer every question before the plant manager can speak? Do workers look at you — or pointedly not look at you? A factory that is proud of its operation walks you through it with energy; a factory that is hiding something walks you in circles, rushes you past zones, and answers questions with questions. Trust the pattern, not the charm. One more pro move: ask the same question to three different people — the salesperson, the plant manager, and a line worker — and compare the answers. The line worker’s version of working hours is the true version. And bring a notebook, not your phone — factories read body language, and a buyer who writes notes commands more honest answers than one who scrolls.

The numbers to collect on every walkthrough

Capture ten numbers on one page: machines running versus installed, workers on the floor, shifts and days per week, order backlog in weeks, claimed versus recorded defect rate, expired calibration stickers, fire tags, the timecard-versus-payroll gap, and other brands’ cartons in finished goods. Ten numbers, one page, every factory — and comparing factories becomes trivial.

4. Document & Record Checks: What the Paper Trail Tells You

The floor shows you today; documents show you the last twelve months — which is why the paper trail is where most audit findings are proven. Factories can stage a production floor for a day; they cannot retroactively fix a year of payroll records, timecards, and inspection logs without leaving a trail of inconsistencies. Here is the document checklist I run on every audit, with what each document proves and the red flags that should end your interest in the factory.

The payroll-and-timecard test

The single most reliable document check in any supplier audit in China is cross-referencing three things: the attendance records (打卡记录), the payroll (工资表), and the bank transfer receipts or signed pay slips (工资条). Pick one random worker — or better, three workers from different departments — and trace their entire month: how many days did the timecard show, how many hours, what base wage, what overtime rate, and did the actual amount paid match the legal calculation? Chinese labor law caps overtime at 36 hours per month (with 40-hour standard weeks), yet working hours are consistently the most common non-conformance found in audits across the major schemes — amfori’s BSCI monitoring reports have repeatedly listed working hours, occupational health and safety, and remuneration as the top three findings in Chinese factories. When payroll shows 60-hour weeks paid at base rates, you are not just looking at a compliance problem — you are looking at a factory whose cost model depends on underpaying workers, and that cost model will eventually collapse into your shipment, your delivery date, or your brand’s reputation.

Quality control China: verifying traceability from raw material to finished box

Quality documents prove consistency; traceability documents prove honesty. Ask to see a complete batch file: the purchase order for raw materials, the IQC report on that batch, the production work order, the IPQC checks during production, the FQC final report, and the packing list. A factory with genuine traceability produces this file in under ten minutes for any batch you name. A factory without it produces excuses. This is also the moment to check whether the material certificates (成分报告, test reports) match what is actually in the warehouse — a factory that claims UL-listed components but shows you certificates for a different brand of component is a factory that will quietly substitute materials the moment your back is turned. Add one more file to the traceability check: the purchase orders themselves. Compare PO dates and quantities against production work orders — a factory that books your order into next month’s slot while running three other brands today is a factory that will miss your delivery date no matter how polite the promise. And check the dates on material test reports: certificates issued once in 2019 and never renewed mean the factory tests when it must, not when it should. When you later run quality control China inspections on shipments, the batch-file check is what lets your inspector verify the goods match the approved sample — not just a spec sheet.

5. Social Compliance and Certifications: BSCI, SMETA, WCA, and Beyond

Social compliance is the section where most buyers’ eyes glaze over — and where their competitors quietly build the moat. Here is the reality: every major Western retailer and an increasing number of mid-market brands require audited social compliance from Chinese suppliers. If your customer base includes any European or North American chain, the audit schemes below are not optional homework; they are the price of entry.

The schemes, explained without the jargon

BSCI (amfori’s Business Social Compliance Initiative) is the most widely used scheme in China. amfori runs roughly 30,000 BSCI audits a year globally, and China consistently accounts for more than a third of them. Audits cover 13 performance areas, and factories are graded A through E, with A the best and E meaning zero compliance. Most European buyers require C or better. SMETA (Sedex) audits are the second pillar — a four-pillar protocol covering labor standards, health and safety, environment, and business ethics. Sedex’s own data consistently shows health and safety and working hours as the two most frequently cited non-conformances in SMETA audits, with wage records close behind. WCA is the protocol Walmart historically required of its suppliers, and it still appears in many factories’ audit portfolios. The newer force in the market is SLCP (the Social & Labor Convergence Program), a framework backed by H&M, Nike, and Adidas that lets factories complete one assessment instead of five duplicate audits. A factory with a valid SLCP assessment — or a recent BSCI C-grade with a credible corrective action plan — is a factory you can work with; a factory with nothing, or with a five-year-old report from a defunct scheme, is a factory that has not sold to serious buyers, and you will be the one teaching it how.

Working hours and wages: the finding that never dies

Whatever scheme you use, plan for working hours and wages to be your biggest fight. The pattern in Chinese export factories is depressingly consistent: legal standard is 40 hours a week and 36 hours of monthly overtime; reality in peak season is 60–72 hours a week with underpaid overtime, and the payroll records are often dual — one set for the labor bureau, one set real. Do not moralize about this during the audit; document it. Photograph both sets if you can. Note the gap between declared and actual hours. Then decide: is this factory’s compliance gap a manageable corrective-action project (fix in 90 days, re-audit) or a structural lie (payroll double-booked, workers coached)? That distinction is the difference between a supplier you can fix and a supplier that will fix you. Also check social insurance enrollment — many small factories enroll only their management team and leave workers on the books as “flexible employment.” If your buyer’s code of conduct requires full enrollment, that gap alone can disqualify a factory for your program.

Interviewing workers is the companion check to the records, and the one skill most auditors underuse. In China, worker interviews work best in small groups of two or three, away from management, in a language the workers actually speak — Mandarin at minimum, with patience for the Sichuan, Hunan, and Henan dialects most migrants speak at home. Ask about last month’s overtime hours, whether wages arrive on time and in full, whether workers sign their own pay slips, and whether the factory holds anyone’s identity documents. Two workers independently describing the same 70-hour week is stronger evidence than any single document you will see all day — and a factory that refuses interviews has, for all practical purposes, already failed.

And remember: a certificate from a scheme your customer has never heard of is worth less than the paper it is printed on — verify it like everything else, because your customer’s compliance team will.

Certifications: verify, don’t collect

A certification wall is not compliance. During the audit, take each major certificate — ISO 9001, ISO 14001, BSCI report, any industry certificates like CCC or CE — and verify three things: the issuing body, the registration number, and the validity date. Search the number online during the audit if you have signal; many “certificates” are PDFs that never existed in any registrar database. For CCC (China Compulsory Certification) products, check that the certificate’s factory address matches the audited address — factories routinely share one certificate across two plants, which is a customs seizure waiting to happen when the goods arrive in Europe. And remember the audit’s real purpose: certifications are evidence, not proof. The proof is the behavior you observed on the floor and in the records. A factory with a current BSCI B-grade and a chaotic floor is still a chaotic floor; the certificate just means someone audited them last year. When you are choosing between two factories for an important order, choose the one whose certificate matches its floor.

6. Case Study: A Canadian Apparel Brand Failed 3 of 5 Factories — and the Scoring System That Saved a $200,000 Order

In August 2022, a Montreal-based apparel brand we worked with — let’s call them Northbound Apparel — was preparing its first large production run in China: a $200,000 order of technical outerwear, 18,000 units across four SKUs, with a hard ship date for a November retail launch. The brand had already been burned once by a Guangzhou trading company that delivered mismatched colors and a 40% defect rate. This time, they wanted to do China sourcing properly: shortlist, audit, then commit.

The shortlist and the audits

Fourteen factories applied for the order — a number that tells you the whole market knows you are shopping. Northbound and our sourcing agent on the ground filtered the fourteen to five candidates across Guangzhou, Shenzhen, Dongguan, and Fujian. The five were then audited over three weeks using the full audit protocol described in this guide: document review, questionnaire, floor walkthrough, payroll test, and a scoring framework with a 70-point pass threshold. Three of the five failed. Factory A failed on the payroll-and-timecard test — 68-hour weeks paid at base rates, with social insurance covering 11 of 214 workers, and a second set of payroll books photographed in a manager’s drawer. Factory B failed on fire safety: the workshop’s main exit was padlocked, extinguishers were all expired, and the “fire certificate” on the wall was issued to a different address. Factory C failed on quality systems — its “ISO 9001:2015” certificate came from a registrar with no online trace, and its FQC records showed 100% pass rates for twelve consecutive months with identical handwriting, which is not quality control, it is fiction. The audits for all five factories, flights and interpreter included, came to about $11,000. The two survivors scored 78 and 71 out of 100 — both above the 70 threshold, both with clear corrective action plans.

What the audits revealed — and what they saved

Here is the part that matters: the two passing factories quoted $2.10 and $2.25 per unit for the shell garment. The three failing factories had quoted $1.68, $1.74, and $1.83 — between 13% and 22% cheaper. On an $200,000 order, that price gap looked like $20,000–$30,000 of savings. It was not savings; it was the payroll underpayment, the unregistered social insurance, and the fictional quality system showing up on a price sheet. Northbound placed the order with the 78-scoring factory at $2.10 per unit, on a 30/70 payment split, with an independent pre-shipment inspection on every batch. The order shipped on time in late October 2022, the inspection passed at a 96.5% acceptance rate on the first batch, and the brand cleared its November launch. The 71-scoring factory became the second-source backup, fixing its CAP items over the following quarter. Total damage avoided, conservatively: $200,000 of product that would have arrived late, off-spec, or not at all — plus the retail penalties and the brand damage a failed launch would have caused. In the words of the brand’s founder: “The audits didn’t cost us money. They bought us the truth for $11,000.”

The scoring framework that turned findings into a decision

The reason Northbound could make a fast, defensible decision is that the audit findings were converted into a weighted score before anyone debated the price. Emotional impressions — “the factory owner seemed nice” — played no role, because they never entered the spreadsheet. The framework below is the one we used, and it is the framework we recommend for any import from China decision.

A few rules for using any scoring framework: set the pass threshold before the audits, not after — 70/100 is a sane default, with a hard fail on any single critical item (child labor, forced labor, fake certificates, padlocked exits) regardless of total score. Weigh quality at 20% and capacity at 15% because those two predict your delivery date and your defect rate. And publish the scores to the factories. A factory that sees its 62 and asks for the corrective action list is a candidate; a factory that argues the scoring is unfair is a factory you just saved yourself from.

7. FAQ: Eight Questions Buyers Ask About Supplier Audits in China

How much does a supplier audit in China cost?

A desktop document and background screen typically costs $300–$800 per factory, done by a sourcing agent or a credit-check firm. A one-day quality or social compliance audit by a third-party firm (SGS, Bureau Veritas, Intertek, TÜV, or a specialized local auditor) runs $800–$1,800 for a single factory, including the report. A full multi-day audit combining quality, capacity, social compliance, and financial checks runs $1,500–$4,000, depending on factory size, location, and how many zones must be covered — and that price usually includes the auditor’s travel and the corrective-action follow-up. For a buyer visiting China themselves, the real cost is different: a week of your time, flights, hotels, and an interpreter, which works out to $3,000–$6,000 per trip — which is why most importers run document screens on five or six factories, fly over once to walk the final two or three, and keep third-party audits for ongoing compliance monitoring. Think of audit cost as insurance premium: the average failed order in China costs $20,000–$100,000 by the time you count the goods, the freight, the rework, and the missed retail window. One audit that catches one bad factory pays for thirty audits. If you run audits through a sourcing agent on a retainer, expect a discount — agencies that audit weekly charge less per visit than a one-off firm, because their auditors are already in the region and their reports use a format you can compare across factories.

How long does a factory audit take?

A focused single-scope audit — quality only, or social compliance only — takes four to six hours on-site for a small or mid-size factory, plus an hour of kickoff and closing meetings, and one to two weeks for the written report. A full audit covering quality, capacity, social compliance, and financials takes one to two full days on-site, depending on factory headcount: plan one day per 500 workers as a rough rule, because the payroll test and worker interviews scale with headcount. Reports for full audits typically land within 7–10 working days, and a credible corrective action plan is due from the factory within 30 days. Two practical scheduling notes: never schedule an audit for the first or last week of the Chinese New Year cycle, when half the workforce is traveling and the factory is running skeleton crews, and always confirm the audit date in writing 10 days ahead — a factory that “forgets” a scheduled audit and asks to reschedule twice is showing you how it will behave when your shipment is late. Budget for the corrective action cycle after the audit too: a factory with findings needs 30–60 days to fix them, so start the audit process at least two months before you need a signed supplier decision, not two weeks. Factories that pass clean on the first try are rare enough that planning for one fix cycle is realism.

Should I audit suppliers myself or hire a third party?

Do both, in different roles. Self-audits (or audits run by your sourcing agent) are best for the first screen: document review, questionnaire, a walkthrough focused on capacity, quality systems, and gut-level signals — because nobody cares about your order the way you do, and an agent on the ground can also verify the factory exists before you spend a dollar. Third-party audits (SGS, Bureau Veritas, Intertek, TÜV, Elevate, and similar) are best for social compliance and certification verification, where independence and a recognized report matter: your customer will accept a BSCI or SMETA report from a licensed auditor, but they will not accept your own walkthrough notes as proof of compliance. The professional sequence used by most serious importers: agent-verified document screen on 5–10 factories → shortlist 3–4 → third-party full or compliance audit on the finalists → your own or your agent’s walkthrough of the winner before the contract is signed. Mixing self-verification with third-party independence is not redundancy; it is the correct control structure for a six-figure commitment. A useful middle path for mid-size buyers: your agent screens and shortlists, a third party audits compliance, and you personally walk the winning factory once before signing — three layers of verification for roughly the price of one failed order. Reserve your own visits for the decision moments.

What is the difference between an audit and a pre-shipment inspection?

An audit evaluates the factory’s systems and conditions — its ability and willingness to produce good goods — and happens before you commit, usually once per factory per year or per program. A pre-shipment inspection (PSI) evaluates the specific goods in front of you — the actual shipment, against the actual spec — and happens after production, usually on every batch or a statistically sampled portion of it, following AQL (Acceptable Quality Limit) sampling standards. The audit answers “should I work with this factory?”; the inspection answers “should I ship this container?” Neither replaces the other. You can audit a factory that then ships a bad batch (which is why PSI exists), and you can inspect shipments from a factory you never audited (which is how buyers discover the factory was never making their goods to spec in the first place). The importers with the lowest total defect rates run both: a full audit at onboarding, an annual re-audit, and PSI on every shipment — and they use the inspection findings to feed next year’s audit checklist. If the budget stretches to only one, take the inspection: an audit tells you a factory can be good, while an inspection tells you whether this batch actually is. For high-risk categories — toys, electronics, cosmetics packaging — the answer is never “or”; both are mandatory.

Can a factory fake an audit?

Yes, and you should assume every factory is trying. The staging techniques are well known in the industry: borrowed equipment from a neighboring factory, hired extra workers for the audit day, a “show line” that runs only when auditors visit, coached worker interviews, two sets of payroll books, and certificates from shell registrars. That is why this guide insists on the checks that are hard to fake: arriving unannounced or at a time the factory does not control, walking against the production flow, asking the same question to three different people, tracing one worker’s full payroll chain, checking calibration stickers and fire extinguisher tags, verifying certificate numbers online, and photographing everything with timestamps. The factories that fake audits are also the factories with the most to hide, which means the audit’s real job is not to catch every lie — it is to make lying expensive enough that honest factories win your business. If your audit finds one staged element, escalate: re-audit at a random date within 30 days, and bring a different auditor. A factory that refuses a re-audit has answered your question. And randomize re-audit dates: factories that time audits to the same month treat the rest of the calendar as rehearsal, and a surprise October re-audit shows you a different factory than the one your predecessor saw in March.

What score should a factory get before I place an order?

Use 70 out of 100 as the minimum for a full audit, with two hard rules on top: a hard fail on any critical compliance item (child labor, forced labor, debt bondage, fake certificates, blocked fire exits) no matter the total score, and a 50-point minimum on the quality-systems block if you are buying anything with tolerances. Below 70, the corrective action list is usually longer than the factory’s management capacity — you would be buying a rework project, not a supplier. Between 70 and 80, place the order but with conditions: a written corrective action plan with dates, a lower first-order quantity, a 30/70 or 30/40/30 payment split, and pre-shipment inspection on every batch until two consecutive shipments pass cleanly. Above 80, you can relax the training wheels: longer payment terms, larger batch commitments, and annual rather than per-order re-audits. One more calibration: compare the score to the quote. If the lowest-priced factory scores 20 points below the highest-priced one, the price difference is not a bargain — it is the audit finding expressed in currency. Whatever threshold you choose, write it into the supplier agreement: a clause stating the factory must maintain an audit score of 70 or above has saved more buyers than any warranty clause ever did. Our factory scoring and audit services run this exact framework if you want it applied to your shortlist.

How often should I re-audit Chinese suppliers?

Re-audit every 12 months as the baseline, and re-audit immediately — within 30 days — after any of these triggers: a failed or suspicious inspection, a delivery delay of more than two weeks, a change of factory management or ownership, a move to a new address, a major new customer taking over capacity, or any customer complaint that traces to factory conditions rather than logistics. The annual re-audit should be lighter than the onboarding audit — skip the full capacity study, focus on the compliance blocks (working hours, wages, health and safety) and the quality blocks (records, traceability, calibration) that drift fastest — and should include a check of whether the factory kept its corrective action promises from last year. Chinese factories change faster than factories anywhere else: a good plant manager leaves, a new customer floods the floor with rush orders, and suddenly the working-hours violations you audited out are back at peak-season levels. Treat the annual re-audit as a subscription, not an event, and you will catch the drift before it catches your shipment. Re-audit before you scale: the factory that passed at 50,000 units a year is a different company at 500,000 — capacity, sub-suppliers, and labor practices all shift with volume, and the audit must shift with them.

Which audit scheme do my customers actually require?

The short answer: ask your customer — schemes are driven by buyers, not factories. European retailers overwhelmingly require BSCI with a C grade or better; it is the default in the German, French, and Nordic markets. UK and Commonwealth retailers, plus many food and textile programs, ask for SMETA (Sedex), usually the four-pillar version. Walmart suppliers historically faced the WCA protocol, and US retailers run their own programs or accept a basket of recognized audits. The trend to watch is SLCP (the Social & Labor Convergence Program), designed to replace the duplication: one assessment, accepted by H&M, Nike, Adidas, and a growing coalition of brands — which means the days of running five different audits for five different customers are ending. When a factory shows you a report, check three things: the scheme (is it one your customer recognizes?), the age (within two years, ideally within one — an older report describes a different factory), and the grade or non-conformance list: a BSCI C with an open corrective action plan is workable, while an E is not. And check that the address on the report matches the factory you are auditing — reports travel between companies in China more often than you would believe. When in doubt, forward the report to your customer’s compliance team before you commit — a five-minute email that saves you from ordering compliance you do not need.

8. Summary: The Audit Is the Cheapest Insurance in China Sourcing

Strip away the checklists and the jargon, and a supplier audit in China is one thing: a decision tool. You are not auditing factories to find perfect ones — they do not exist. You are auditing to find the ones whose flaws you can live with, price in, and manage, and to make the ones whose flaws would destroy you eliminate themselves before your money is at risk. Every factory in this market has at least one problem: the honest ones have a working-hours gap they will fix with a corrective action plan; the dangerous ones have a payroll lie, a padlocked fire exit, or a fictional certificate that no corrective action will ever fix. The audit’s entire job is to tell you which one you are talking to, before the deposit, before the tooling, before the $200,000 order.

The audit in one paragraph

Run the sequence in this order and you will outperform 90% of importers: first, a document screen and credit check on every candidate (business license, export rights, certificates, litigation) — this eliminates half your list for a few hundred dollars. Second, a written questionnaire that forces each survivor to commit to capacity, headcount, wages, and defect rates on paper. Third, a full on-site audit of the final two or three: floor walkthrough against the production flow, payroll-and-timecard test, traceability batch files, worker facilities, and certificate verification. Fourth, convert everything into a weighted score with the pass threshold set in advance. Fifth, negotiate with the score in one hand and the quote in the other, because the gap between them is the truth about the factory’s cost model. Sixth, re-audit annually and immediately after any red-flag event, and never let a corrective action plan expire without verification. A corrective action plan is a promise, and in China a promise without a deadline and a named owner is a suggestion. Every CAP item gets a date, an owner, and a verification step in your next audit — and every missed CAP date is a data point about whether this factory can be trusted with your delivery schedule.

The five mistakes that turn audits into theater

Mistake one: announcing the audit date two weeks ahead and letting the factory choose the route and the escort. Mistake two: walking the floor but never opening the payroll file — judging a restaurant by its wallpaper. Mistake three: letting the salesperson lead the walkthrough, so every question routes through the person whose bonus depends on the sale. Mistake four: collecting forty findings and then deciding by gut feeling, which guarantees the charming factory wins. Mistake five: auditing once and never again, treating a point-in-time report as a permanent license. All five are common, and all five are cheap to fix.

The factory that watches you make these mistakes does not respect you for it — it prices you. Every skipped payroll test nudges the quote up half a point; every accepted photocopied certificate nudges the defect rate up. Chinese suppliers respond to the standards you enforce, and they calibrate their honesty to the depth of your checks. The buyers who get the best factories are not the biggest buyers; they are the ones whose audits make faking expensive. The audit is not the show — it is the first act of a relationship that runs on verified facts. Run it properly and it becomes what it should be: a routine, repeatable, boringly reliable way to know what you are buying before you buy it.

Your next three moves

If you are mid-search for a supplier right now, do three things this week. First, send the document pack request and the questionnaire from Section 2 to every factory on your shortlist and watch who responds in 48 hours and who stalls — that response pattern is your first audit finding, and it is free. Second, decide who runs your audits: a sourcing agent for the screens and walkthroughs, a third-party firm for the compliance reports, and make it before the contract deadline breathes down your neck. Third, set your pass threshold and hard-fail list now, in writing, so when the factory owner is charming, the quote is 20% below market, and the dinner is excellent, the decision has already been made by the evidence — not by the moment. The importers who get burned in China are almost never the ones who audited and chose wrong. They are the ones who skipped the audit and chose on price, on charm, or on hope. An $11,000 audit bill that saves a $200,000 order is not a cost center. It is the best return in supply chain management — and the only one that pays you in truth. When you are ready to run your first audit properly, find a vetted sourcing agent in China who can get you on the floor within two weeks.

China sourcing, supplier audit, Chinese suppliers, supply chain management, quality control China, sourcing agent, import from China, factory audit checklist, sourcing strategy, social compliance