How Do You Switch from Trading Companies to Chinese Factories Without Regretting It?

You hear the same number at every import meetup: cut out the middleman, buy straight from Chinese factories, and you’ll save 30 percent. It’s the most repeated promise in China sourcing, and it’s also the most dangerous. Trading companies do add 15 to 30 percent onto factory prices on average — sometimes more. But the markup is only half the equation. Direct sourcing from China can slash your landed costs, or it can burn your deposit, delay your season, and poison a product line you spent a year building. The outcome depends entirely on how you switch. This guide walks you through the full migration: what the markup really looks like, how to verify a factory is actually real, and how to phase the transition so you keep selling while you test.

The 30% Markup Myth: What Your Trader Really Costs You

Let’s start by dissecting the number that starts every sourcing conversation. The “30 percent” figure isn’t fabricated — it’s just incomplete. Industry sourcing benchmarks compiled across platforms like Alibaba, Global Sources, and Made-in-China consistently put trading company markups in a 15 to 30 percent band for consumer goods, with some categories pushing past 40 percent. But that range hides more than it reveals.

Where the 30% Number Comes From

A trading company in Yiwu or Guangzhou buys from the same factory you could buy from. Their cost is the factory’s FOB price. They then add their margin, which covers their rent, their buyer-facing staff, their English-speaking salespeople, their sample management, and their profit. For a standard SKU — a stainless steel water bottle, a silicone spatula, a basic LED lamp — that margin lands at 15 to 25 percent. For anything requiring coordination — custom packaging, multi-SKU consolidation, private label — it climbs to 30 percent or beyond.

Here’s what most importers miss: the trader’s markup compounds. If the factory’s FOB price for your water bottle is $2.00, a trader selling at $2.60 adds 30 percent to the factory price. But the $0.60 is not the real gap. You also pay freight on the higher value, you pay duties on the higher value, and your insurance, customs broker fees, and financing costs all scale with it. At the end of the landed-cost chain, a 30 percent FOB markup can become a 34 to 38 percent landed-cost difference. That’s the markup myth in its purest form: buyers think they’re comparing a $0.60 gap when they’re actually comparing a much larger one.

Category matters as much as markup. Simple commodity goods — steel wire baskets, acrylic boxes, basic plastic housewares — carry the smallest trader margins, because the factories themselves compete on price and the trading layer adds little value. Engineered goods with certification requirements — electronics, toys, anything with UL, CE, or FCC paperwork — carry the biggest margins, because the trader’s real service is managing compliance documentation. This asymmetry is your friend when you pick which SKUs to migrate: the categories where traders add the least value are exactly the ones where factory-direct sourcing is safest and most profitable. Conversely, migrating a certified electronics SKU before you understand the compliance workflow is how brands burn their first year of savings on failed certifications. Match the SKU type to the sourcing strategy, and the markup math finally starts working in your favor.

What the Markup Actually Buys You

Before you demonize that markup, look at what it purchases. A good trading company provides:

• Curated factory access. They’ve already vetted the factories they represent. You’re not gambling on an unknown workshop.
• Language and logistics bridge. They handle the WeChat negotiations, the freight booking, the export documentation, the HS code classification.
• Small-order absorption. Many traders will happily pack 300 units of a product that a factory would refuse to touch at that volume.
• Problem absorption. When a batch arrives late or defective, the trader mediates. You have someone to shout at who isn’t the factory owner’s brother-in-law.

None of that is worthless. The question is whether it’s worth 30 percent of your product cost forever. For a brand doing $50,000 a year in orders, the trader’s fee is a rounding error compared to the risk of a botched switch. For a brand doing $500,000 a year, the same fee is a $150,000 line item that could fund your entire QC program, your own bilingual sourcing hire, and a factory audit budget — with money left over.

Case Study: Brightpath Home & Garden

Brightpath Home & Garden, a Columbus, Ohio retailer of kitchen and garden accessories, spent 2019 to 2021 buying through a single Shenzhen trading company. Their average order was $18,000, their markup across 22 SKUs averaged 28 percent, and they never questioned it. In late 2020 they ran a line-by-line cost audit and found that three of their highest-volume SKUs — a bamboo cutting board set, a herb scissors bundle, and a cast-iron trivet — carried markups of 31, 34, and 37 percent respectively.

In March 2021 they moved those three SKUs to factory-direct sourcing: two factories in Yongkang, one in Jieyang. Same specifications, same packaging, same freight forwarder. Their landed cost dropped 26 percent on the moved SKUs, which translated to $42,600 in savings in the first twelve months — against a switching cost of roughly $9,000 in samples, travel, and third-party inspection fees. They kept the trader for their remaining 19 low-volume SKUs, where the trader’s consolidation service still made economic sense.

The lesson: the markup myth is real in magnitude but not in direction. You don’t save 30 percent across the board. You save 25 to 35 percent on the right SKUs — the ones with volume, stable specs, and simple packaging — and you lose money switching the wrong ones. That distinction is the entire strategy of this article.

Why Trading Companies Still Win the First Three Orders

If factory-direct pricing is so much better, why does anyone still use a trader? Because the first three orders through a new factory can cost you more than a decade of markups. The switch fails most often in the first 90 days, and it fails for predictable reasons.

The Three Things Traders Do That Factories Won’t

1. They absorb your small orders. Minimum order quantity (MOQ) is the silent killer of the factory switch. Public data from Made-in-China.com and Alibaba’s platform listings shows the median MOQ for consumer goods sits around 500 to 1,000 pieces, with many electronics and molded-plastic items requiring 2,000 to 5,000. A trading company will consolidate your 300-piece order with other buyers’ orders and ship it happily. A factory quoting you 300 pieces will either refuse, demand a 300 percent price premium, or quietly slot you behind their larger customers.

2. They buffer your amateur hour. Your first factory negotiation will include questions you didn’t know were wrong, specs you didn’t know were ambiguous, and a quote that assumes tolerances you didn’t know existed. Traders absorb that chaos. They’ve explained “PP material, food grade, BPA free” to a hundred buyers before you. A factory salesperson has explained it to a dozen, and they’re tired.

3. They give you someone to blame. This sounds like a joke, but it’s a real risk-management function. When your shipment arrives with the wrong handle color, you need a counterparty whose business model depends on keeping you happy. A trader loses your account if they mishandle it. A factory has a hundred other accounts; your $15,000 order is Tuesday afternoon.

When the Trader Is the Right Call

There are three situations where the trader genuinely beats direct sourcing, and you should stop feeling guilty about it:

• Your order is under the factory’s MOQ. If you can’t reach 500 pieces per SKU, the math rarely works in your favor. The trader’s consolidation effectively buys you volume you don’t have.
• You need mixed containers. A 20-foot container holding 14 different products from 6 factories is a logistics problem. Traders exist to solve exactly that. Direct sourcing it means 6 supplier relationships, 6 QC visits, and 6 sets of export documents.
• You’re testing a market. New category, unproven demand, uncertain spec stability — that’s not the moment to commit to tooling, molds, or a dedicated supplier relationship.

The Real Price of the Switch

The markup you pay the trader shows up on every invoice. The cost of switching is spread across a dozen places where you’ll never see a line item. Add them up before you commit: your own hours — expect 4 to 8 hours a week of supplier management during the first year, more during the test-order phase; cash tied up in deposits and molds — a 30 percent deposit on a $40,000 order is $12,000 parked in a supplier’s account for 60 to 90 days; the learning curve — your first factory negotiation takes twice as long and produces a worse deal than your tenth; and the quiet cost of attention — every hour spent managing a new supplier is an hour not spent on product, marketing, or customers. None of these show up in the price comparison, and all of them are real.

Phased migration works precisely because it spreads these costs across a year instead of concentrating them in one quarter. The buyers who regret the switch are rarely the ones who paid too much — they’re the ones who paid with attention they couldn’t spare. If you’re a two-person team with a 90-day design cycle, the switch may genuinely not be worth it yet. That’s not failure; that’s arithmetic. Revisit the math every six months — the SKU that made no sense last year can become your best candidate the moment your volume grows.

Case Study: Nordic Trail Co.

Nordic Trail Co., a Stockholm-based camping accessories brand, learned this the hard way. In 2022 they moved their flagship camping lantern to a factory-direct arrangement with a Dongguan manufacturer. The price was excellent — 24 percent below their trader — but the MOQ was 3,000 units per colorway, and their quarterly demand was 1,200. They took the 3,000-unit bet, hit 1,100 units of unsold inventory, and ate $14,000 in warehousing and markdowns. Meanwhile their smaller SKUs, still on the trader, sold through cleanly.

Their fix took 18 months: they renegotiated with the factory to a 1,200-unit MOQ in exchange for a 4 percent price increase, split colorways across two seasons, and permanently reserved the trader for their four low-volume SKUs. By mid-2024 their landed cost on the lantern was still 19 percent below the trader baseline, and their inventory turns were back to healthy. The case study here is not “traders are bad” — it’s that volume gates everything. If you can’t meet the factory’s MOQ with realistic sell-through, the trader isn’t a luxury. It’s your bridge.

Strategy: The Phased Migration Plan

The importers who switch successfully don’t flip a switch. They run a phased migration over 12 to 24 months, where the factory and the trader coexist, compete, and gradually swap roles. Here’s the framework that consistently works.

Phase 0: The Cost Baseline (Months 0–2)

Before you approach a single factory, quantify what your trader actually costs you. Pull your last 12 months of orders and compute, per SKU: trader price, landed cost per unit, annual volume, and annual landed cost. Then estimate the factory-direct price from public listings on Alibaba and Made-in-China for the same specification — treat those as floor prices, not quotes. Flag the SKUs where the gap is at least 20 percent AND annual volume exceeds 2,000 units. Those are your candidates. Everything else stays with the trader for now.

Phase 1: The Test Order (Months 3–6)

Pick one SKU — ideally your highest-volume, simplest, most stable product. Source it from 2 to 3 candidate factories simultaneously. Run each through the verification process in the next section of this guide. Order samples from all of them, then place a test order (150 to 300 units, or whatever the factory will accept) with the best one. The test order should be small enough that a total failure costs you less than one month of trader markup on that SKU. You’re not buying product here; you’re buying information about lead times, communication quality, defect rates, and honesty.

Phase 2: The Parallel Run (Months 7–12)

Now run the factory and the trader on the same SKU simultaneously for two or three order cycles. This is the part most buyers skip, and it’s the part that prevents regret. Compare them on five axes: landed cost, on-time delivery rate, defect rate at incoming inspection, communication responsiveness, and problem resolution speed. The trader will still win some of these axes. That’s fine. You’re collecting data, not declaring a winner.

Keep score like a procurement department, not a buyer: landed cost, on-time delivery measured to the day, defect rate at inspection, response time in hours, resolution time in days. Write the scores down. Writing them down is what stops the parallel run from becoming an infinite loop — a surprising number of buyers run factory and trader side by side for years without deciding, because the comparison is never quite finished. Set the decision date in Phase 0 and keep it. If the factory wins three consecutive cycles on landed cost with comparable quality, the migration proceeds. If the trader wins, keep the SKU with the trader and test a different one — not every SKU needs to migrate for the strategy to work.

Phase 3: The Scale Migration (Months 13–20)

Move SKUs to the factory in tranches of two or three, always starting with the ones that passed the parallel run. Keep the trader as a fallback supplier for at least one order cycle per migrated SKU — a hedge against factory capacity surprises during your peak season. This is also when you negotiate annual pricing, volume rebates, and mold ownership terms with the factory.

Phase 4: The Relationship (Months 21+)

By this point, the factory should account for 70 to 80 percent of your spend, with the trader kept for long-tail SKUs and emergency capacity. Book a face-to-face visit at least once a year. Chinese factory relationships — like most long-term supplier relationships — run on trust built in person. One physical visit a year does more for your pricing and priority treatment than a dozen WeChat negotiation marathons.

The 20/80 Rule of SKU Migration

A useful heuristic: migrate the 20 percent of your SKUs that generate 80 percent of your volume, and leave the rest alone. For most importers, that means 3 to 6 SKUs move factory-direct, and the other 10 to 30 stay with the trader. This rule has a second benefit nobody mentions: it keeps the trader profitable enough to care about your small orders, so the relationship survives the migration. You want the trader alive and cooperative — they’re your safety net.

Case Study: Harbor Supply Group

Harbor Supply Group, a New Jersey distributor of commercial janitorial supplies, ran this exact playbook from January 2023 to March 2025. They identified 14 of their 61 SKUs as migration candidates — a mop bucket line, two sprayer SKUs, and a chemical bottle line — representing 76 percent of their China-sourced volume. They tested two factories per SKU group, ran a six-month parallel period, and migrated in three tranches. Their trader’s role shrank from 100 percent of China sourcing to 31 percent, but the trader retained the low-volume chemical refills that Harbor didn’t want to manage. Net result: 23 percent lower landed cost on migrated SKUs, zero stockouts during the transition, and one factory relationship (the mop bucket manufacturer in Taizhou) that now produces 41 percent of Harbor’s total imported volume.

Execution: Verify the Factory Is Actually Real

Here’s the part where the switch gets genuinely scary: a significant share of “factories” you’ll meet online are not factories. They’re trading companies posing as factories, or worse — shell operations that don’t exist outside a WeChat account. The verification process is not optional, and it’s not that hard. It has two pillars: license checks and the on-site audit, known in Chinese as 验厂 (yàn chǎng).

License Checks: The Public Records You Can Pull Yourself

Every legitimate Chinese factory is registered with the State Administration for Market Regulation, and that registration is publicly searchable through the National Enterprise Credit Information Publicity System (国家企业信用信息公示系统). You can access it at gsxt.gov.cn — it’s free, it’s in Chinese, and it’s the single most powerful due-diligence tool most Western buyers never use.

What to check:

• Business scope (经营范围). The registration must explicitly include manufacturing (生产, 制造, 加工) of your product category. A company registered only for “trading” (贸易, 批发) is a trader, full stop. That’s not automatically disqualifying, but it means you’re paying trader pricing — negotiate accordingly.
• Registration date. A factory claiming 15 years of export experience should have a registration date that matches. Newly registered companies aren’t automatically fraudulent, but they demand extra scrutiny.
• Registered capital (注册资本). Not a reliability guarantee — Chinese registered capital is declared, not always paid in — but a claim of “$5 million in assets” against a 500,000 RMB registration is a red flag.
• Business status. Check for 吊销 (revoked) or 注销 (cancelled) status. You’d be surprised how often buyers negotiate with companies whose licenses were cancelled years ago.
• Cross-reference the license name with your contract and bank account. Payment should go to the registered entity named in the contract. If the factory asks you to pay a different entity — especially a personal account — stop and ask why.

For a step-by-step walkthrough of the license verification workflow — the exact fields to check and what the registration screens should look like — see Caijing188’s factory verification guide. The same logic applies to export credentials. Legitimate exporters hold an Import/Export License (进出口权) and a Customs Registration Code — ask for scans plus their latest export customs declaration form. A real factory that exports regularly has these ready in a folder; a middleman will stall.

The 验厂 (Factory Audit): What to Check When You Visit

A license confirms a company exists. The on-site audit confirms it can actually make your product. You have three options: visit yourself, hire a third-party audit firm (SGS, Bureau Veritas, Intertek, TÜV), or both. For your first relationship, do both — one visit yourself, one third-party audit.

On the visit, check these twelve points:

1. Production lines running. A factory that’s “too busy to show you the floor” during a scheduled visit is managing your perception. Walk the floor unannounced if you can.
2. The equipment matches the product. A stainless steel factory should have presses, welders, and polishing stations. A “steel factory” that shows you an assembly room and an office is a trading operation.
3. Worker density. A factory claiming 300 workers should have a floor that looks like it. Count heads in the workshop, not the brochure.
4. Raw material stock. Real factories hold raw material inventory. No material on the floor usually means no production.
5. QC stations. Look for in-line inspection points, measuring tools, and rejected parts bins. Factories that make quality products have visible quality infrastructure.
6. Export packing area. Check whether they actually pack export cartons — export packaging differs from domestic packaging in real, observable ways.
7. The office. Is there a real office with real staff, or a showroom with a receptionist? A showroom-only operation is a trader with a lease.
8. Ask for production records. Last month’s production orders, machine logs, or shipping records. Real factories can produce them in minutes.
9. Meet the quality manager. Not the salesperson. The QC manager should discuss your product’s tolerances, failure modes, and test methods in detail.
10. Check certifications on site. Certificates photocopied in a PDF are cheap. Certificates with a live auditor verifying them on the floor are real.
11. Assess their other customers’ products. Real factories have other buyers’ goods in production. Seeing recognizable third-party brands on the floor is strong evidence of actual manufacturing.
12. Eat lunch with the staff. It sounds absurd, but it’s a real signal: a factory that treats its workers decently is one that will survive long enough to matter. High-turnover sweatshop operations threaten your supply continuity.

Case Study: Solstice Lighting

Solstice Lighting, a Los Angeles decorative lighting importer, almost lost $60,000 to a fake factory in 2023. They found a “manufacturer” on Alibaba offering brass table lamps at 31 percent below their incumbent supplier. The company had a polished website, a catalog, and a WeChat account with two years of activity. The license check saved them: the National Enterprise Credit Information Publicity System showed the company registered for “household goods trading” — not manufacturing — with a registered capital of 300,000 RMB and a 2021 registration date that contradicted their “15 years of export experience” claim. A reverse image search on their “factory floor” photos revealed they were stock images.

Solstice switched to a verified manufacturer in Zhongshan, flew their operations manager out for a 验厂 visit in March 2023, and confirmed production capacity with a live line walk. The verified factory’s price was 22 percent below the original trader, and Solstice’s first two production runs (4,000 units total) shipped defect-free. The entire investigation cost them one day of work and zero dollars. The fake would have cost them $60,000 and their holiday season.

Execution: Contracts, QC, and Payment That Can’t Backfire

A verified factory can still ruin your quarter if your contract, quality control, and payment terms are sloppy. This section covers the three execution details that separate a smooth switch from a regrettable one.

Contract Clauses That Matter More Than Price

Your purchase contract with a Chinese factory is a real legal instrument, even if it’s a simple one-pager. Five clauses deserve your attention:

• Specification annex. Attach the full product spec as an annex: materials, dimensions with tolerances, colors with Pantone references, packaging artwork, carton dimensions, and weight. Chinese commercial law and arbitration practice treat the spec annex as binding. Vague specs are where defect disputes are born.
• Inspection and acceptance terms. Specify who inspects, when, and what happens on failure: pre-shipment inspection by a named third party, a defect threshold (commonly 2.5 percent AQL), and the remedy — rework at factory cost, replacement, or refund.
• Delivery and liquidated damages. A fixed shipment date, a grace period, and a per-day penalty. Chinese factories routinely prioritize customers who have penalties in their contracts over customers who don’t.
• Mold ownership. If you paid for tooling, the contract must state that the mold is your property and must be returned or destroyed at contract end. Negotiate this before you pay for the mold, not after.
• Dispute resolution. Standard terms specify CIETAC (China International Economic and Trade Arbitration Commission) arbitration or Hong Kong arbitration. This is not a trap — it’s the practical venue where a Chinese counterparty can actually be held to account. A contract with no dispute clause leaves you with no practical remedy.

Quality Control: Inspect Before It Ships, Not After It Arrives

The single biggest QC mistake in the factory switch is inspecting at your warehouse. At that point, the product is on your dock, your money is spent, and your leverage is gone. The correct sequence:

1. Pre-production sample approval. The factory sends you an approved sample before bulk production. You approve it in writing. This sample becomes the reference standard for the whole order.
2. In-line inspection at 30 to 50 percent of production, if the order is large enough to justify it. This catches problems while rework is cheap.
3. Pre-shipment inspection (PSI) at 100 percent completion, by a third-party firm, against your AQL level. The PSI report is your gate: no PSI pass, no payment release, no shipment.

Third-party PSI costs roughly $250 to $500 per visit depending on location and firm. Compared to the alternative — receiving 8,000 defective units — it is the best-value insurance in importing.

One more QC nuance: align your defect thresholds with your channel. A 2.5 percent AQL is standard for general consumer goods, but if you sell through Amazon or big-box retail, your practical tolerance is closer to 1 percent — returns, reviews, and chargebacks make retail defects far more expensive than wholesale ones. Consider a stricter AQL for retail-bound SKUs, and add packaging checks to the PSI scope: carton crush tests, drop tests, and label placement. Each one costs a few dollars and catches failures that are ten times more expensive to fix after arrival.

Payment Terms That Protect You

The standard China payment structure for a first-time direct relationship is 30 percent deposit, 70 percent against the pre-shipment inspection report and copy of the bill of lading. That structure is worth defending. Red flags to avoid: requests for 100 percent payment upfront (except for very small orders), payments to personal accounts, payments to an entity different from the contract signer, and demands to pay before samples are approved.

As trust builds across order cycles, you can negotiate toward 20/80, 10/90, or even open account for repeat suppliers. But the first year should be boring and conventional. Boring payment terms are how you keep your leverage.

Case Study: VoltEdge Electronics

VoltEdge Electronics, an Austin-based accessories brand, learned the value of contract discipline in 2023. Their first factory-direct order — 8,000 units of a USB-C hub, sourced from a verified Shenzhen manufacturer — arrived with an 11 percent defect rate: cold solder joints and misaligned housings. Their contract specified a 2.5 percent AQL, third-party pre-shipment inspection, and rework-at-factory-cost remedies. Because they had followed their own process, the PSI had flagged the problem before shipment; the factory was contractually obligated to rework the batch at its own cost, and VoltEdge’s deposit structure meant the final 30 percent payment was withheld until the rework passed a second inspection. Total cost of the incident: about $4,000 in delayed-shipment freight. Total cost without the contract: $31,000 in defective goods and a ruined product launch. The factory rework took 19 days, the second PSI passed, and VoltEdge shipped one month late but with a 1.8 percent defect rate. That relationship is now in its third year, and the factory has become their highest-volume supplier.

The Data: Markups, MOQs, and China Sourcing Trends

Numbers anchor this decision better than opinions. Here’s what the public data says, with source context, followed by the comparison table you’ll want to show your business partner.

The Markup Data

China’s General Administration of Customs reported total exports of roughly $3.58 trillion for 2024, up about 5.9 percent year over year after a dip to $3.38 trillion in 2023 — the world’s largest export machine, and getting bigger. Within that flow, the share moving through intermediaries is significant but shrinking. Industry sourcing benchmarks from B2B platforms and sourcing agent surveys consistently estimate that trading companies and sourcing agents add 15 to 30 percent on top of factory FOB prices for consumer goods, with complex or small-quantity orders at the top of that band. A 2023 survey of US importers published by the American Apparel & Footwear Association (AAFA) member trade press found that buyers who migrated at least half their volume to direct factory relationships reported average landed-cost reductions of 18 to 25 percent within 18 months — numbers consistent with the markup band above, minus switching costs.

The MOQ Data

MOQ is where the markup myth dies. Platform data from Alibaba and Made-in-China.com shows median MOQs for consumer goods of roughly 500 to 1,000 pieces per SKU, with molded plastics, electronics, and apparel frequently requiring 2,000 to 5,000 pieces. Only about 15 to 20 percent of listed consumer-goods suppliers offer MOQs under 300 pieces — and those are disproportionately trading companies or hybrid operations. The practical translation: your ability to go factory-direct is gated by volume, not desire. If your order profile can’t clear the MOQ bar, the trader markup is the price of doing business.

The Direct Sourcing Trend

US Census data shows US imports from China holding around $430 to $440 billion per year through 2023–2024 — down from the 2021–2022 peak but still the dominant sourcing lane. The structural trend underneath those flat numbers matters more: the share of imports moving through dedicated sourcing agents and direct factory relationships has grown steadily over the past decade, while traditional import/export trading houses have consolidated. Trade press coverage of China sourcing — including the sourcing benchmarks and factory guides published at Caijing188 — consistently describes the same shift: buyers going direct for high-volume SKUs and retaining intermediaries for the long tail. The data doesn’t say “abandon traders.” It says “the middle is being hollowed out”: the SKUs with volume go direct, the small stuff stays intermediated.

Trader vs. Factory: The Comparison Table

Read the table with your own order book in hand. The honest takeaway for most importers is a portfolio answer, not an either-or: 70 to 80 percent of your volume belongs factory-direct, and 20 to 30 percent belongs with an intermediary — the split most successful migrators report after two years. Middlemen who consolidate, manage compliance, or handle your long tail are doing real work at a defensible price; the table is a decision tool, not a verdict on middlemen.

Case Study: Atlas Outdoor

Atlas Outdoor, a Denver camping gear brand, is a clean dataset for this whole section. In 2022 they sourced 12 SKUs through a Guangzhou trading house at an average 26 percent markup, with MOQs of 200 pieces. In 2023 they migrated 4 high-volume SKUs to two verified factories — a tent manufacturer in Xiamen and a cookware factory in Yongkang — accepting MOQs of 1,000 and 1,500 pieces respectively. Their annual volume justified the jump. The result, measured over 12 months: 24 percent lower landed cost on migrated SKUs, 8 percent higher on-time delivery rate, and one painful quarter of 9 percent defect rate on a tent batch that their new PSI protocol caught before shipment. Their 8 remaining low-volume SKUs stayed on the trader, where they still belong. Atlas’s blended sourcing cost dropped 13 percent across the portfolio — the real-world number, not the myth.

FAQ: Eight Questions Buyers Ask Before They Switch

1. How do I know if my order volume justifies going factory-direct?

Run the math on one SKU. Your annual volume × the price gap between trader and estimated factory price must exceed your switching costs: samples ($50–$300 each), a verification visit ($2,000–$4,000 including travel, or $300–$600 for a third-party audit), a test order’s risk, and your time. As a rule of thumb, a SKU needs at least 2,000 units of annual volume and a 20 percent price gap to justify the switch in year one. Below that, the trader is usually cheaper once you count your own hours. The readiness scoring table in the next section formalizes this: score 3 or higher on volume and price gap before you invest in samples. Also factor in the softer costs: your own time for supplier management, the learning curve of negotiating with factories, and the emotional cost of a failed first order. Two practical tests settle most cases. First, the 2×2 rule: if you can’t order at least twice per year at the factory’s MOQ, the relationship never stabilizes. Second, the line-item test: list every service your trader performs — consolidation, documentation, dispute mediation, language support — and price what it would cost to replicate each yourself. If the price gap covers replication and still leaves 10 percent, switch. If not, the trader is doing real work you’d be paying for anyway.

2. What’s the fastest a safe switch can happen?

Four to six months for a single SKU, realistically — and that’s fast. The compressed path: week 1–2 license verification of 3 candidates, week 3–6 samples and quotes, week 7–10 test order (150–300 units), weeks 11–16 parallel run of two order cycles, month 5 scale. Anything faster usually means you skipped a verification or QC step, and the regret rate on skipped steps is brutal. If a supplier pushes you to skip sampling or inspection “to save time,” treat that as a warning, not a convenience. Timing matters as much as speed. Chinese factory capacity runs on its own rhythm: peak season runs roughly August to November, when production slots are tight and quality control stretches thin. Schedule your test order in the slow season — roughly February to June — for faster samples, better prices, and more attentive service. Add four weeks of buffer to every lead time the factory quotes during the first year; factories learn to over-deliver on schedule only after they trust you. And remember Chinese New Year: everything shuts down for two to four weeks, so a January order quoting February delivery is either a lie or a miss in the making. The buyers who regret the switch almost always regret its speed, never its destination. Move at the pace of verification, not the pace of enthusiasm.

3. Should I visit China before switching?

Yes, at least once, for the factory you intend to bet on. A physical 验厂 visit catches what documents can’t: whether the production line is real, whether the workers are competent, whether the owner’s promises match the floor’s reality. Budget $3,000–$5,000 for a 4–5 day trip covering two or three factories in one region (Guangzhou–Shenzhen–Dongguan, or the Yangtze Delta around Ningbo–Shanghai–Hangzhou). The visit pays for itself if it prevents one bad order. And schedule it during a production week, not a Chinese holiday week — visiting during Chinese New Year is a classic rookie error. Plan the itinerary like a procurement professional, not a tourist: one region, two to three factories, and a spare day for unannounced visits — the most valuable meeting is often the one you didn’t schedule. Budget for a translator who knows manufacturing vocabulary, because a mistranslated tolerance is exactly the kind of ambiguity that sinks orders. If a visit is genuinely impossible, hire a third-party auditor and ask for live video evidence: a dated WeChat video walkthrough of the floor with the production manager on camera. It’s not as good as being there, but it filters out the photo-only fake factories. And treat the visit as relationship capital, not a cost — the factory owner who has shaken your hand will quote you differently for the next three years.

4. What if the factory’s first order is defective?

You handle it with the contract, not with anger. The sequence: document the defect with photos against the approved sample, notify the factory within the contract’s claim window (typically 7–14 days of receipt), invoke the inspection clause, and withhold the outstanding payment if the remedy isn’t agreed. If you did pre-shipment inspection, the defect should have been caught before shipment — that’s the point of the PSI gate. If it slipped through anyway, the AQL clause defines what counts as a defect and what remedy applies. Realistically, the first order through a new factory has a 5–15 percent chance of a significant quality problem; that’s why the test order is small and the QC budget is not. The deeper lesson: defect disputes are decided by paperwork, not persuasion. Photograph everything against the approved sample, ruler and date visible. If the factory claims the defect is “acceptable industry tolerance,” your AQL clause is the tiebreaker. Expect the negotiation to take two to four weeks — Chinese factories respond to written evidence and payment leverage, not threats. If they stonewall, escalate: sales manager, general manager, then arbitration. Know that CIETAC arbitration is a last resort — $5,000 to $15,000 and six to twelve months; in practice, a documented claim against a factory that wants your repeat business resolves at the first escalation. That’s why verification and contracts matter more than any inspection gadget.

5. How do I negotiate MOQ down without wrecking the price?

Ask for a volume-breakdown structure instead of a flat MOQ. Factories will often accept a lower initial order (say 500 instead of 1,000) if you commit to a 12-month forecast of 3,000, or if you accept a 3–5 percent price premium on the small first run that’s refunded as a credit on the second. Another lever: share the risk by offering to pay a slightly higher deposit (40/60 instead of 30/70) in exchange for MOQ flexibility. The worst way to negotiate MOQ is to pretend you have volume you don’t have — factories can smell a bluff, and it damages your credibility for every future negotiation. A third lever: split the MOQ across your own sales channels — Amazon, your site, and retail can consume the same 1,000-piece MOQ through three forecasts. A fourth: ask for the MOQ in two runs — 500 now, 500 confirmed within 60 days, with materials already ordered — factories accept this when it smooths their own production planning. Factories set pricing based on what they believe about your future volume; buyers who deliver on forecasts get better treatment than those who inflate them. Finally, remember that MOQ is negotiable per SKU, not just per factory: the MOQ on a factory’s core product is often firm, while the MOQ on a product similar to one they already make is soft. Lead with that question.

6. Is it cheaper to hire a sourcing agent than to go fully direct?

Sometimes, and it’s a legitimate middle path. Sourcing agents typically charge 5–10 percent of order value (or a flat monthly retainer of $1,000–$3,000), which is far below the trader’s 15–30 percent markup but still gives you a boots-on-the-ground partner who verifies factories, attends inspections, and handles disputes. An agent is the right answer when you want direct pricing without building your own China operation — say, under $300,000 annual spend. Above that, a dedicated in-house sourcing hire or your own audit program usually pays for itself. Many buyers use an agent for the first year of migration, then drop to 2–3 percent for monitoring once relationships are stable. One more structural warning: agent conflicts of interest are common — many agents take factory commissions on top of their fee, quietly defeating the purpose of going direct. Ask for commission disclosure in writing, and cross-check factory quotes against the factory’s own public listings. A clean structure is a flat fee plus fully disclosed commissions, with auditable factory prices. The hybrid model also works: use an agent for verification and inspection only, at $300 to $600 per audit visit, while you negotiate and manage the relationship yourself. That gives you direct pricing with professional guardrails. Whatever you choose, keep the structure simple enough that you know where every dollar goes — opacity in sourcing costs is how the 30 percent markup myth survives.

7. How do I protect my designs and molds?

Three layers. First, mold ownership in the contract, as discussed — your mold, your property, return-or-destroy clause. Second, non-disclosure agreements with the factory and any subcontractors (Chinese courts and arbitration panels do enforce NDAs, contrary to popular belief, when they’re specific and reasonable). Third, supply-chain hygiene: don’t share your full retail strategy, don’t let the factory see your retail prices, and watermark your design files so a leak is traceable. Accept the residual risk honestly: if your product is a huge commercial success, someone in China will copy it. That’s true with traders too — the difference is you’ll know exactly who your counterparty is. One more layer: control your sample flow. Factories share samples with vendors and, occasionally, with competitors who ask nicely — label every sample with your brand, date, and a control number, and track where each one goes. For exclusive products, consider paying a 2 to 5 percent exclusivity premium for a contractual agreement not to sell your design to other buyers in your market; it’s common in furniture, lighting, and housewares. And audit your own chain: freight forwarders, agents, and even designers can leak designs. The most practical protection is speed — a product on the market and rebuying at scale is far harder to copy profitably than one in development. Copycats follow sales.

8. What’s the biggest reason switches fail?

Overreach. Buyers migrate too many SKUs at once, skip verification, accept MOQs they can’t sell through, or cut the trader off before the factory relationship is proven. The failure signature is almost always visible in month two: a 6-SKU migration, one unvisited factory, no third-party inspection, and a payment structure that front-loads risk. The fix is boring: one SKU, verified factory, inspected orders, parallel run, gradual tranches. Every importer who has made the switch successfully — Brightpath’s $42,600 in year-one savings, Harbor Supply’s 23 percent cost cut, Atlas Outdoor’s 13 percent blended reduction — followed a version of that boring path. Regret is not the price of switching; it’s the price of skipping steps. A second failure mode is under-management after the switch. Buyers assume a verified factory runs itself; then lead times stretch, defect rates creep, and communication slows — and they only notice at the container. The factories that serve you best are the ones you measure quarterly against the same five axes from the parallel run. A third mode is abandoning the trader too early: buyers cut them to save the markup, discover the factory can’t cover an emergency peak-season order, and re-hire the trader at a premium. Keep the trader on a standing small-order arrangement — it’s cheap insurance. The subtlest failure is expectations: the switch saves 20 to 25 percent on migrated SKUs, not 30 percent on everything. Buyers who expect the myth abandon a strategy that was working.

Summary: The 8-Step Switch Checklist

Your China Sourcing Readiness Score

Before you send a single inquiry, score your readiness. The table below gives you a quick self-assessment. Score each line 1 to 5 (1 = not ready, 5 = fully in place). Total: 35–40 means you’re ready to start; 25–34 means fix the weak lines first; under 25 means stay on the trader for now and build volume.

Readiness Assessment Table

The 8-Step Migration Checklist

Step 1: Run the per-SKU cost baseline. Compute landed cost per SKU through your trader and flag everything with a 20%+ gap and 2,000+ annual units. Why this works: it converts a vague “maybe we should go direct” into a ranked list of exactly what’s worth switching.

Step 2: Verify candidate factories by license. Pull registration records from the National Enterprise Credit Information Publicity System, confirm manufacturing scope, cross-check names against contracts and payment accounts. Why this works: this free step eliminates the fake-factory category — the highest-consequence, lowest-cost risk in the switch.

Step 3: Visit or audit your top candidate. Fly in, or hire SGS/Bureau Veritas/Intertek, or both. Walk the floor, check the 12-point audit list, meet the QC manager. Why this works: factories behave differently when a real human has seen their floor; your leverage and your pricing both improve afterward.

Step 4: Order samples from 2–3 factories and test hard. Run the samples through your actual quality checks — don’t just eyeball them. Why this works: sample quality is the single best predictor of bulk quality, and testing against your real spec catches miscommunication while it’s still free.

Step 5: Place a small test order and inspect it pre-shipment. 150–300 units, third-party PSI, no exceptions. Why this works: a small order caps your downside at one month of trader markup while teaching you the factory’s real lead times, communication, and honesty.

Step 6: Run the parallel period. Keep the trader on the same SKU for 2–3 order cycles and compare landed cost, on-time rate, and defect rate head-to-head. Why this works: the comparison converts opinion into data, and it keeps your fallback warm.

Step 7: Migrate in tranches with annual contracts. Move 2–3 SKUs at a time, negotiate annual volume pricing and mold ownership, and keep the trader for the long tail. Why this works: tranching prevents the one-disaster-wipes-everything scenario and gives each new factory relationship time to mature.

Step 8: Institutionalize the relationship. Book annual visits, schedule quarterly business reviews, and audit the factory’s performance on the same five axes you used in the parallel run. Why this works: supplier performance decays without measurement — the factories that stay excellent are the ones whose customers hold them to a published scorecard.

Run this checklist once per SKU, not once per factory. A factory can be excellent for your mop buckets and wrong for your sprayers — capacity, tooling, and experience vary by product line. Most migrators run the checklist three to five times before they find the mix that sticks, and that’s normal — Harbor Supply Group’s 14-SKU, 26-month migration above is the model to copy.

The Bottom Line

The 30 percent markup myth is real in magnitude and misleading in application. Yes, trading companies add 15 to 30 percent to factory prices, and yes, direct sourcing from China can cut your landed costs by 20 to 25 percent on the right SKUs. But the switch only works when it’s phased, verified, and measured — one SKU at a time, through license checks, on-site audits, third-party inspection, and a parallel run that keeps your trader as a safety net. Do it that way and the switch stops being a gamble and becomes an engineering project with a predictable payout. Skip the steps and you’ll join the long line of importers who tried the shortcut and spent a season explaining to their bank why the container was full of defects. The factories are real, the savings are real, and the process is proven — you just have to do the boring work in the right order. For updated cost benchmarks, contract templates, and audit checklists, Caijing188 keeps a running library for importers doing exactly this switch.

Tags: china sourcing, direct sourcing from china, factory sourcing, trading company markup, verify chinese factory, factory audit, 验厂, minimum order quantity, china import guide, sourcing checklist