Sourcing from China for Amazon: Where First Orders Fail

Most first Amazon orders fail on one of ten steps that nobody checked, not on the product. The factory quote is the smallest number in the whole job, and it is the only one most sellers negotiate.

Every one of those is cheap to fix before the money moves and expensive to fix after it does.

Most First Orders Do Not Fail on the Product

The product is usually fine, and the order built around it is what breaks. Sellers spend six weeks choosing between two nearly identical items, then five minutes on the packing, the paperwork, and the person who is supposed to be watching the factory.

The pattern is always the same shape: the cheapest step gets skipped, and the most expensive step pays for it. A $300 check refused in month one becomes a $9,000 pallet of unsellable stock in month four. That is not bad luck. That is arithmetic.

Ten steps decide whether a first order works. The rest of this page walks each one, tells you what usually goes wrong, and points you to the detail when you need it.

Step 1: Pick a Product That Can Survive Its Own Fees

A product has to be profitable after Amazon takes its cut, not before. Start with the price you can realistically hold, subtract the fees, subtract fulfillment, subtract advertising, and only then ask what the product and the freight are allowed to cost. Many ideas die at this table, which is the cheapest place they can die.

Size and weight decide more of your profit than the factory quote does. A product that folds, nests, or ships flat can beat a better product that does not, because you pay for its shape on every unit forever. Work through choosing profitable products before you fall in love with anything.

The number that decides everything is not what the product costs you, it is what survives. Add the duty, the inspection, the packing and labels, the Amazon prep, the storage, the returns you will eat, the discount you will eventually run, and the advertising it takes to stay visible. Then run the whole thing again at a lower price with a higher return rate. A product that only works at launch price, with perfect reviews and nothing going wrong, is not a product yet. It is a spreadsheet that has not met the market.

Step 2: Find a Supplier, Then Find Out Who They Really Are

Half the “factories” quoting you are trading companies, and that is not automatically bad. A good trader beats a bad factory. What hurts you is not knowing which one you are talking to, because you cannot fix a problem in a building your supplier does not own.

Check the company before the sample, not after the deposit. A business licence, a real address, an export record, and a matching bank account answer most questions in a day. Start with a proper Chinese supplier search, then run a supplier quality audit on whoever survives the shortlist.

Three quotes almost never describe the same product. One supplier priced the material you asked for. One quietly dropped an accessory. The cheapest one assumed a thinner wall and a plain brown carton, and said none of that out loud. Write one specification and make every supplier quote against it: material, size, finish, accessories, packaging, testing, and shipping term. A low price only means something once you know what was taken out to reach it.

Step 3: Approve a Sample, Not a Photograph

A sample approved over chat is not a specification, it is a hope. Photos hide color, weight, smell, finish, and the thing that will actually get returned. Hold the sample, use it the way your customer will, then write down what “correct” means.

The approved sample is the only thing you can hold production against. Keep one sealed, sign it, and make it the reference for the run. Managing sample orders is where a first order quietly becomes either controlled or improvised.

A sample and a written specification do two different jobs, and you need both. The sample shows what the product should look and feel like. The specification writes down what cannot be argued about later: dimensions, how much they are allowed to vary, material grade, color reference, finish, weight, packing, labels, and how it all gets checked. A factory can always tell you production looks close to the sample. It cannot argue with a number written into the order.

Step 4: Check Compliance Before You Pay the Factory

Compliance is a matching test, not a quality test. Amazon compares the model number, the brand, and the labels on your paperwork against the exact item on your listing. One file describing a cousin of your product stalls the whole listing, even when the product itself is legitimate.

The factory will send you papers without ever mentioning that the papers are not yours. They cover a similar model, made for another brand, in another year. Do the Amazon compliance check while your money is still in your account, because after the deposit your leverage is gone.

Step 5: Know What Your Category Does to You

Every category kills margin its own way, and the plan that saves one will not save another. Sellers copy tactics across categories and wonder why nothing works. The trap is category-specific, so the checklist has to be too.

Beauty and healthcare punish you on formula, labels, and claims. What you say on the listing is as regulated as what is in the bottle. See beauty and healthcare risks.

Electronics punish you on returns and paperwork. Batteries, radios, and the buyer who expected something slightly different. See consumer electronics risks.

Household goods punish you on size, breakage, and dead stock. Cheap to buy, expensive to move, easy to crack. See household goods risks.

Step 6: Watch the Production You Paid For

Nobody at the factory is protecting your money, and that is a job description, not a moral failure. They are paid to finish the order. You are the only person paid to care whether it is right.

The final check that protects your money happens while the goods are still in the building. A pre-shipment inspection before the balance is paid is the last moment when a problem is still the factory’s problem. After the container leaves, every fault becomes yours to eat.

One check at the end cannot undo the mistakes made at the beginning. Look at the first finished units before the run goes any further, because that is when a wrong color, a bad print, or a sloppy joint costs an afternoon instead of a container. On a bigger or unfamiliar order, look again while the line is still running. The final check then confirms quantity, function, appearance, packing, and labels before you pay the balance. An initial production check and a last check protect two different moments, and neither one covers the other.

Step 7: Pack for the Warehouse, Not for the Shelf

Amazon’s warehouse does not care what your box looks like, only whether it can scan it. Wrong label, wrong placement, wrong bag, wrong carton weight, and your shipment gets refused or charged back after it arrives.

The prep rules assume you already know what is in the box. Get the labels, the poly bags, the carton limits, and the placement right before the goods ship, because fixing it at the other end costs more than the whole prep would have. Read shipping to Amazon FBA before your factory starts packing.

Step 8: Ship for the Margin and for the Reorder

The cheapest freight turns into the most expensive freight the day your listing runs out of stock. Sea protects your margin but demands a forecast you may not have yet. Air protects your launch but can eat the entire profit on a bulky product.

Split shipments exist for exactly this problem. Send enough units by air to launch or to cover the gap, and let the rest come by sea. Sea freight against air freight is a margin decision, not a logistics preference.

Build your reorder date from the whole chain, not from the factory’s promise. Production, inspection, freight, customs, and Amazon’s receiving time all sit between your order and a sellable unit, and the last two are the ones sellers forget.

Step 9: Let the First Month Tell You What to Change

Returns are the cheapest research you will ever get, and almost nobody reads them. Your customers are inspecting your product for free and writing down exactly what is wrong with it.

Reading returns before the reorder is what separates a second order from a second mistake. A wrong dimension, a confusing setup, a smell, a missing part: each one is a factory instruction waiting to be written. Work through Amazon return reasons before you repeat the order, not after.

Step 10: Decide Who Is Doing All of This

Every step above is somebody’s job, and if you have not named that person, the answer is nobody. That is how first orders fail: not through one bad decision, but through ten unassigned ones.

Doing it yourself works when you can be awake for a Chinese working day and get on a plane when it matters. When you cannot, the cost of a person on the ground is usually smaller than one bad container. A sourcing agent for Amazon is worth it when the order is big enough that a mistake hurts, and not before.

FAQ

Q1: How long does a first order take, from search to live listing?

Plan on three to five months, and the sourcing part is the smaller half. Samples and revisions eat more calendar than production does, and warehouse check-in adds weeks that nobody puts in the plan.

Q2: Should I start with one product or several?

One, until you have seen a full cycle end to end. Spreading a first budget across three products buys you three sets of problems and no clear read on which step you got wrong.

Q3: The factory’s MOQ is far higher than the quantity I want to test. What now?

Ask what the minimum is protecting, because it is usually a material run or a machine setup, not a rule. A trading company, a stock color, or a shared run can get you a smaller first order, and paying a little more per unit to test the market is cheaper than owning three thousand of the wrong thing.

Q4: The factory now sells my exact product to other Amazon sellers. Can I stop that?

Not if you bought their existing product, because it was never yours. A real design change or your own mold is what makes a product yours, and that has to be decided before the first run.

Q5: How much does Chinese New Year move my dates?

More than the factory will admit. Lines slow for weeks beforehand, close for three to four weeks, and come back with new staff, so anything landing near it needs a month of air in the schedule.

Q6: My order arrived wrong but it is still sellable. Do I take it?

Decide what the defect costs you in reviews before you decide what it costs you in refunds. Selling a flawed first batch to protect cash flow can bury the listing under one-star reviews you cannot delete later.

Q7: How much cash should I keep out of the first order?

Enough to cover a freight surprise, the inspection, rework, prep, ads, returns, and the reorder that has to be placed before the first run sells out. An order that spends every last dollar leaves you no way to fix what went wrong or repeat what worked.

Q8: The product works. Should I move it to a cheaper factory?

Only when the volume is real and the saving is bigger than it looks on paper. Moving resets everything the first factory learned about your product, and the new one gets to make all the same mistakes at your expense.

Conclusion

First orders fail because ten separate steps get treated like one, not because sourcing from China is hard. Handle them in order and the risk drops at every stage, long before Amazon ever sees the product.

If you would rather have those ten steps run by people who do them every week, that is what a full sourcing service is built for.