If you sell on Amazon US and you're thinking about sourcing your next product from China — read this before you touch Alibaba. Most first-time attempts fail in one of four predictable ways, and each one costs $8,000-$50,000 to unwind. Here's how to skip all four.
This is the exact process we run for 40+ US Amazon sellers, including Mava Sports (300+ SKUs, 2M+ customers). No theory — just what actually works in 2026.
Why 60% of first-time China sourcing attempts fail
We've talked to 200+ US Amazon sellers over the years. Roughly 6 in 10 have already tried to source direct from China once — and got burned. The failure pattern is almost always one of these four:
1. You sourced from a trader, not a factory
Alibaba is full of "manufacturers" that are actually trading companies. They buy from a real factory and resell to you with a 15-25% markup. You paid trader prices thinking you were direct. You had no way to know.
2. You skipped pre-shipment QC
Sample was perfect. First bulk order had 12% defects. Reviews on your listing dropped from 4.6 to 3.9 in two weeks. No inspection meant no leverage — factory said "sorry, do better next time."
3. The freight forwarder billed you $8,000 in surprises
Quote was $2,400 for sea freight. Then came the extras: destination charges ($350), port fees ($200), ISF filing ($50), customs bond ($100), chassis fee ($240), detention because your 3PL wasn't ready ($900). Total: $10,400 you didn't budget for.
4. Your agent took a kickback
You paid the "factory price" your agent negotiated. The real factory price was 12% lower. Your agent pocketed the difference, on top of the fee you already paid them. This happens with 60-70% of freelance sourcing agents.
The uncomfortable truth
These aren't "bad luck." They're the default outcome for anyone who tries to shortcut the process. Every single one is preventable — if you have the right process.
The 6-step process we run for every Amazon seller
This is the workflow. Follow all six steps, don't skip. Each one exists because we (and our clients) have been burned by skipping it.
Step 1 — Brief (24 hours)
Before anything happens, we need to know exactly what you want. That means:
- Product spec: dimensions, weight, materials, colors (Pantone codes), features
- Reference product: link to a competitor on Amazon, or a photo — something we can point to and say "like this"
- Target volume: initial test order + realistic annual quantity if it works
- Target landed cost: what your Amazon margins can support
Within 24 hours we come back with: is this feasible, what MOQ to expect, ballpark factory pricing (from memory + database), and 2-3 factories we'd approach first.
Step 2 — Sourcing (5-7 days)
Our China team hits our database of 2,500+ vetted factories. We request quotes from 5-10 factories that match your spec. Important: we never tell them who else is quoting. They compete blind, which gets us better prices.
The quotes come back. We rank them on price, MOQ, lead time, factory reputation, and past client experience. Top 2-3 factories move to samples.
Step 3 — Samples (2-3 weeks)
This is where most sourcing goes wrong. Amateur approach: order 1 sample per factory, ship to your US address, look at it, call it good. What we actually do:
- Order 3-5 sample units per factory (not 1)
- Ship them to our Shenzhen office first
- Our engineer inspects each sample against your spec: dimensions, weight, materials, durability
- Only the samples that pass our internal test get forwarded to you
This filter saves you 4-6 weeks. You get 2-3 real options, not 10 samples where 7 are subpar.
Step 4 — Negotiation (1 week)
Once you approve a sample, we negotiate the full deal:
- Unit price at your target volume
- MOQ (we usually get 20-40% below their initial ask)
- Tooling costs (if custom)
- Packaging design + cost
- FBA-ready prep at the factory (polybagging, FNSKU labels, carton labels)
- Payment terms (30% deposit, 70% on pre-shipment QC pass — never 100% upfront)
- Lead time in writing with penalty clause for missed delivery
You see the real factory PI. We charge a flat % on order value, no hidden markups. If a factory offers us a commission, we refuse — our incentive is to negotiate down for you, not up.
Step 5 — Production + QC (4-8 weeks)
Deposit paid, production starts. Two inspection points:
- Mid-production visit (around week 2-3): our engineer visits the factory, inspects the first 500 units off the line, sends photo report. Catches problems while there's still time to fix them.
- Pre-shipment inspection (before container loads): AQL 2.5 sampling, functional testing, packaging check, container loading supervision. Full written report with photos.
You get the report before paying the balance. If defect rate is over 2-3%, you have leverage: withhold the balance and demand rework.
Step 6 — Delivered (4-7 weeks sea, 1-2 weeks air)
Freight booked, customs cleared, routed to your 3PL, warehouse, or straight into Amazon FC if you have direct-injection set up. One USD invoice covers everything — no surprise fees.
Total timeline (realistic)
Off-the-shelf via sea: 9–13 weeks. Off-the-shelf via air: 6–8 weeks. Custom-tooled via sea: 13–18 weeks. Custom-tooled via air: 10–13 weeks. Anyone promising "6 weeks door-to-door" on a custom product is either lying or counting only production time.
How to spot a trader vs a real factory (5 signals)
The single most expensive mistake is buying from a trader who pretends to be a factory. Here's how to tell:
| Signal | Real factory | Trader |
|---|---|---|
| Product range on Alibaba | Narrow, 1 category | Wide (electronics + apparel + home) |
| Business license type | Manufacturing | Trading / import-export |
| Video factory tour on 24h notice | Yes, willing | "Our factory is 2 hours away, next week" |
| Who owns the tooling | "We do, X years" | Vague / "our partner" |
| Wire recipient matches company name | Yes | Different (trading Ltd.) |
The Amazon-specific stuff that saves you money
FBA prep at the factory
Every Amazon seller pays a US 3PL to prep goods before shipping to FBA: polybag, FNSKU label, carton label. Typical cost: $0.30-0.75/unit. On 10,000 units that's $3,000-7,500.
You can get all of it done at the factory for $0.05-0.15/unit. Savings: 70-80% on prep costs. The factory just needs your barcode files and packaging spec.
Chinese New Year planning
CNY 2026 is February 17. Factories close ~Feb 10, reopen ~Mar 2. Realistic impact window: 6 weeks. If you stock out on Amazon in January-February, you lose ranking that takes months to recover.
Rule: plan Q1 inventory arrivals by December 15 at the latest. Every week later = more risk.
Cost engineering
Once you have a working supplier, focus obsessively on shaving cost per unit. Every $0.10 matters when you're competing on Amazon:
- Packaging: switch from custom box to standard mailer, save $0.20-0.50
- Material spec: 80gsm cotton vs 100gsm often unnoticeable to customer, saves 10-15%
- Shipping density: refold/nest products in carton to fit more per container
- Consolidation: multiple SKUs in one container = shared freight cost
Our Mava case study: 18% landed cost drop on grip pads from packaging refactor alone.
What we charge (for reference)
Our sourcing fee is a flat percentage on order value, agreed upfront in writing. Typically 5-8% depending on complexity. That's it. No commissions from factories, no hidden markups.
Sample stage: charged at cost (usually $50-200 per sample round). Everything else — factory audits, mid-production QC, pre-shipment inspection, freight coordination, customs — bundled in.
"We tested three sourcing agencies before landing on ASP. They were the only ones who walked the factory floor on a Saturday because our launch slipped. That's the difference."
Want us to source your next product?
Book a free 20-minute call. We'll tell you if we're the right fit — and give you a realistic timeline + budget on the spot.
Book a free sourcing call →