Ordinex
← All postsOperator

Requesting and Judging 1688 Samples Properly

June 5, 2026

A lot of sellers importing from 1688 skip the sample step entirely, or treat it as optional. Then the bulk shipment arrives and the goods do not match what they saw on the listing page. By then the capital is in the warehouse and there are no good options left.

A sample is not an optional step

When you place your first order with a new supplier, you are betting that the photos and description on 1688 accurately reflect the real product. Sometimes that bet pays off. Sometimes it does not. The problem is you do not find out until the goods land.

A sample is how you test that assumption before the real money is committed. The cost of one sample, including express shipping from China, typically runs from a few hundred thousand to under two million VND depending on the product. Against a batch of hundreds or thousands of defective or wrong units, that is a small number.

But a sample only helps if you know what you are checking and how to read the result. Holding it in your hand and thinking "looks fine" is not enough.

When to order a sample, and when to skip it

Not every import requires a sample. A few situations where you should prioritize one:

  • First order from a new supplier. This is non-negotiable. You have no history with them and no real quality data.
  • Products with specific technical or material requirements. Electronics, anything that contacts skin, children's products, dietary supplements: these need a direct check, not a judgment from photos.
  • High-value SKUs or large batches. When the value of a batch exceeds what you can absorb if it goes wrong, the sample fee is a reasonable insurance cost.
  • Products where dimensions, weight, or actual packaging matter. Specs on 1688 are sometimes wrong, or measured differently from how you calculate freight. A surprise on weight or volume across a full batch adds up fast.

You can reasonably skip a sample when you have ordered dozens of batches from the same supplier with consistent results, or when you are switching to a slightly cheaper source for a SKU you already know well. Even then, if you are ordering a new variant or changed specification, request a sample of that variant specifically.

How to request a sample properly

Sample requests go wrong in three common ways: not specifying whether you are paying for it or expecting it free, leaving out a clear delivery address, and not listing the exact specs you want to verify.

Samples come in two forms. A free sample is something suppliers offer when they expect a large follow-up order and the product value is not too high. A paid sample is normal for most products. You pay the item cost at retail or at a sample price, plus separate shipping. Do not hesitate to pay for it since this is sourcing cost, not wasted money.

When messaging a supplier on 1688 for a sample, include:

  • The product name and a direct link to the listing. Do not describe it in words. Use the URL to avoid SKU confusion.
  • Quantity. One to three units is usually enough to evaluate.
  • The specs that must match. Color, size, material, version if the product has multiple variants.
  • A full delivery address. Either your China consolidation warehouse address or your Vietnam address.
  • An invoice request. If you need documentation for customs later, say so at this stage.

You can use a translation tool to message in Chinese. Keep the message short. Suppliers handle many requests a day and long messages often lose the last few lines.

How to check a sample properly

When the sample arrives, do not just look it over and decide. A short process makes the evaluation actually useful.

Check appearance and structure. Look carefully, handle it, test basic functions. Compare directly against the 1688 photos and description. Does the color match, does the material feel as expected, are the seams or joints solid. Everyone does this step but doing it carefully makes a real difference.

Measure and weigh it. This matters most for products where dimensions affect shipping cost. If the product is slightly heavier or larger than the 1688 spec, that difference multiplied across a full batch can push your landed cost up meaningfully. Actual weight always overrides the number on the page.

Test it under real conditions. Depending on the product, use it the way a buyer would. If you sell kitchen items, run them. If you sell accessories or clothing, try wearing them. A lot of hidden defects only show up in use, not when you first open the box.

Check the packaging. Suppliers sometimes pack samples more carefully than bulk production. Ask directly whether the actual packaging on the large batch will match what you received. If the packaging is thin or poorly cushioned, goods are more likely to arrive damaged. Raise the standard before you place the batch, not after.

Document everything with photos and video. Photograph the sample from multiple angles and record video if useful. This is your reference when the bulk shipment arrives. If the batch differs from the approved sample, you have evidence to negotiate with or file a dispute.

Samples do not automatically represent the full batch

This is the most important point, and the one people learn after losing money.

A supplier can send their single best unit as the sample, or produce the sample manually with more care than their production line allows at scale. When the full batch runs, the line runs faster, quality control is lighter, and actual goods sometimes fall short of the sample.

Ways to reduce this gap:

  • Evaluate the sample fully before placing the batch. Give yourself one to two weeks to test properly. Do not let the supplier rush you into the next step.
  • Keep the first batch smaller than you planned. Even with a good sample, the first batch should be small enough to absorb the risk if quality turns out to be inconsistent. Around 30 to 50 percent of your intended initial volume is a reasonable starting point.
  • Request pre-shipment photos or QC. When the large batch is ready to ship, ask the supplier to photograph the actual packed goods before sealing the boxes. For high-value batches, a third-party inspection service inside China is worth the cost. Some freight agents who handle China routes offer this as an add-on at a reasonable rate.
  • State quality terms before you place the deposit. Tell the supplier clearly that the batch must match the approved sample. If the defect rate exceeds a certain threshold, you expect compensation or replacement. Say this before paying the deposit, not when the goods arrive at your warehouse.

Price the sample into your cost

The sample fee is not a miscellaneous expense. It is part of your sourcing cost. If you test three suppliers and choose one, the sample cost from all three should be allocated against the first batch you place with the winner, not written off as overhead.

In practice, if each sample costs around one to two million VND and your first batch is 100 units, the sample adds roughly 10,000 to 20,000 VND per unit. That compresses quickly on larger batches. But for a small first test batch, include it in the real landed cost so your margin calculation is accurate from the start.

A bad sample that saves you from a wrong batch has already paid for itself many times over.

Bottom line

A sample does not guarantee a perfect batch, but it is the only step that surfaces problems before the large capital is committed. Request it properly, check it systematically, document the result, and price the cost into your landed calculation. Those four habits make every subsequent batch less risky than the one before.