Small MOQ Product Testing: How to Launch Without Overbuying Inventory

The Amazon ad budget you will burn on a product that does not sell is much smaller than the inventory write-off when you ordered 5,000 units of a product that does not sell.

We see this every quarter. A new seller finds a "winning product", gets excited, and orders the supplier's standard MOQ, usually 500 to 1,000 units. When it does not sell at their planned ACoS, they are sitting on 18 months of inventory for a product that needs to be liquidated. The alternative is small-MOQ testing. Order 50 to 200 units, test the market, then scale. Less glamorous, much more profitable.

Why low MOQ testing is safer for new sellers

The math is simple. Say your product has $4 landed cost and sells for $19.99 on Amazon. Your contribution margin after Amazon fees, PPC, and returns is around $7 per unit.

A test order of 100 units at $4 is $400, plus about $300 in setup costs (sample, packaging, freight). Worst case: $700 lost. A "real" first order of 1,000 units at $4 is $4,000, plus about $1,500 in setup. Worst case: $5,500 lost.

The 100-unit test gives you the same information about whether the product sells. The 1,000-unit order ties up your cash for 8 to 12 months if it does not.

Buyers who scale do this:

1. Order 100 units, list, run a small ad campaign.
2. Watch your CTR, conversion rate, and 7-day sales velocity.
3. If the numbers say yes, reorder 1,000 to 3,000 units.
4. If the numbers say no, kill it. Lose $700, not $5,500.

The seller who tested 5 products at 100 units each ($3,500 total) and found one winner is much better off than the seller who bought 1,000 units of one product ($5,500) and found a loser.

How to negotiate sample and trial orders

Factories want big orders. You want a small one. There is a middle ground that works for both sides.

• Ask the agent or a middle-tier supplier, not the manufacturer direct. Smaller MOQs are easier to negotiate through someone with relationships.
• Pay slightly more per unit. A 100-unit trial at $4.80 instead of $4 is fine if it means you do not gamble $5,000.
• Promise the path, not the order. "If this hits 4-star reviews and 50 units a week, my next order is 2,000." Factories respond to a credible path, not promises.
• Skip custom packaging for the trial. Use the factory's stock packaging, or plain cartons. Custom adds $2,000 in tooling and 3 weeks of lead time.
• Use Alibaba's sample order feature for very small quantities (often 10 to 50 units). The unit price is higher, but it is a real production sample, not a hand-made one.

A trial run rarely makes the factory rich. It rarely costs you more than $500 to $1,500. That is the deal.

Which product categories are easier for small MOQ

Not all categories work for small-MOQ testing. The ones that do:

• Apparel and accessories, cut-and-sew at 100 to 500 units is normal in Guangzhou.
• Pet products, many manufacturers run small MOQs (100 to 300) for testing.
• Bags and small leather goods, often 100 to 300 MOQ from established makers.
• Beauty and skincare, stock formulas with white-label packaging start at 200 to 500 units.
• Home decor and small kitchen items, often 200 to 500.
• Phone accessories and small electronics, stock items at 100 to 500.

Categories where small MOQ is harder: custom electronics with PCB design (tooling and component minimums push MOQ to 500 plus), injection-moulded plastics with custom moulds (mould costs $2,000 to $15,000, minimum to amortise is 1,000 plus), and custom textiles (dye lots have minimums, often 500m of fabric).

If you are testing a category in the "hard" bucket, find a stock product that is close to what you want first. Test the concept, not the custom version.

When custom packaging is worth it

For the test order: probably not. Custom printed packaging usually has $300 to $800 in plate or setup costs, plus an MOQ of 500 to 1,000 boxes. On a 100-unit test, you would burn money on packaging that may end up in the dumpster.

For the test, use the factory's stock packaging, a generic white box with your printed sticker, or the polybag the factory ships in with a simple barcode label. This is enough for Amazon FBA. It is enough for TikTok Shop. It is enough to test the market.

For the first reorder (1,000 plus units), custom packaging starts to make sense. By that point you have proven the product, you know your reorder cadence, and the packaging investment pays back across multiple cycles. Plate fees become a fraction of a percent per unit.

How to scale from test order to bulk production

You have your 100 units, you have listed them, you are running PPC. Now what?

Watch the numbers for the first 4 weeks:

• Click-through rate on your main image and title. Below 0.3 percent is a title or image problem. Above 0.5 percent means the product is being noticed.
• Conversion rate on the listing. Below 5 percent means something on the page is not working. Above 10 percent is a strong fit.
• Organic vs PPC ratio. If you only sell through ads, you do not have a winner yet. You have a paid demand spike.
• Return rate. Above 5 percent means the product is not matching its promise. Investigate before reordering.

If the test hits at least 5 to 7 organic sales per week with under 4 percent returns, that is a green light to reorder.

The reorder math: take your weekly velocity, multiply by 12 to 16 weeks (your typical reorder cycle), add 20 percent safety stock. That is your order quantity. For a product doing 10 organic sales a week, that is 120 to 192 units of pure demand, plus PPC sales, plus safety stock. Reorder around 250 to 500 units, not 5,000.

Scale up gradually. Each reorder lets you negotiate slightly better terms with the factory, switch to custom packaging when it is justified, and test the next layer (adding variants, larger formats, bundles).

Most successful Amazon and TikTok Shop sellers we work with do not bet big on first orders. They test five products, find one winner, then double down on the winner with 5 to 10 times larger reorders. That approach takes more cycles, and it also means you do not end up with 5,000 units of a product that turns into dead stock.