The real question is not whether to order a 1688 sample. It is: what does being wrong cost me on this specific order?
Frame it as risk math, not a habit of caution. You lose in one of two ways. Pay for a sample, it arrives fine, and you realize you did not need it. Damage: roughly $25 and three weeks. Or skip the sample, place bulk, and the goods arrive wrong. Now you have $1,200 in miscolored stock, no recourse, and payment already cleared.
Three variables drive the decision: order value, product complexity, and supplier history. By the end of this post you have a formula and two worked examples to make the call yourself in five minutes.
What Sampling From 1688 Actually Costs
Supplier-side price ranges from free to $10 USD for most categories. Some suppliers charge retail rather than wholesale for samples, so confirm before paying. Forwarding to Vietnam for a package under 1 kg adds $6 to $16, with delivery in 7 to 15 days.
Total per sample round: $12 to $40, plus time to test and decide. If the sample fails, you restart supplier search from zero and lose another two to four weeks. For how sampling fits into your total landed cost, see how 1688 import costs break down for new operators.
The Financial Risk of Skipping the Sample
Most common failure modes on 1688 bulk orders: color deviates from photos, sizing runs 1 to 2 cm smaller than spec, fabric is thinner than the listing suggests, stitching varies across units, printed logos bleed or peel. None of these show up in product photos.
Scenario: $1,200 first-time clothing order. Goods arrive two shades darker than the listing. You already paid. Return policy covers manufacturing defects only. You mark down 40% and absorb the loss.
Dead stock is a double hit: capital locked up plus opportunity cost for better SKUs you could have stocked instead. The compounding damage from carrying the wrong items is detailed in the dead stock cycle in 1688 importing.
4 Clear Signs You Should Sample First
Order value above $200 to $400 with a supplier you have never used. Adjust the threshold by category, but this is a practical starting point for most operators.
Product has sensory or fit-dependent value. Apparel, footwear, soft home goods, baby items. The primary customer return reason is always "not as expected," and no photo resolves that.
Supplier has under 200 reviews or recent negative feedback on quality consistency in the past 60 days. Overall star rating alone is not enough. Read the recent text reviews.
This SKU is your lead product for the launch, not a small volume test. If a bad outcome compromises the entire opening, the $25 insurance cost is obvious.
When You Can Skip the Sample
Commodity items with low variation risk: plain tote bags, basic charging cables, simple stationery. There is not much to cut corners on, and factories producing these at volume tend to ship close to spec.
Supplier has 500 or more reviews, recent buyer-uploaded video that matches the listing description, and an explicit exchange policy in writing.
Order value is small enough that worst-case loss is an acceptable cost of learning.
You have bought from this supplier before and the last order matched. That direct evidence outweighs any review score.
For orders you want to protect without sampling, how to check 1688 product quality before final payment covers your inspection options.
The Breakeven Formula: Five-Minute Decision
Expected loss = Order value x Estimated defect probability. If that number exceeds your sample cost, order the sample.
Example A. Order: $800. New supplier. Apparel. Defect probability: 15%. Expected loss: $120. Sample cost: $25. Verdict: sample.
Example B. Order: $120. Supplier with 800 reviews. Simple product. Defect probability: 5%. Expected loss: $6. Sample cost: $25. Verdict: place the bulk order.
For the defect probability input, use product type, review quality, and prior order history. First-time apparel order with an unknown supplier: 15 to 20% is an honest estimate. You do not need precision here. You need a number you can defend, and then you run the math.
Getting your landed cost right is the other half of this calculation. Calculate your 1688 cost of goods before committing to confirm margins hold even after absorbing a sample round.
FAQ
How much does a 1688 sample actually cost delivered?
Supplier price: free to $10 USD depending on item type and whether they price it at wholesale or retail. Forwarding to Vietnam for under 1 kg adds $6 to $16. Total per round: $12 to $40, plus 7 to 15 days transit. If the sample fails and you need a different supplier, add another two to four weeks before you can reorder.
Do I need to place a new order for bulk, or can I roll the sample into it?
In practice, you place a new order. Most suppliers do not deduct sample cost from subsequent bulk orders unless you negotiated that in writing before placing the sample. If this matters to you, settle it in chat before paying and screenshot the conversation.
Can a supplier send a good sample but ship lower quality in bulk?
Yes, this happens. Some suppliers set aside their best units for sample requests. To reduce the risk, ask for a random pull from existing stock rather than a handpicked unit. If consistency is critical, order two to three samples from different production lots when that option exists.
Should I order one sample or three to five at once?
Depends on variant count. One color tells you about construction quality but not color consistency across the run. For multi-color orders, get at least two to three variants to check dye and print evenness. For a single-variant product, one sample is usually enough.
Is there a way to reduce risk without ordering a sample at all?
Yes. Third-party pre-shipment inspection, buyer-uploaded video rather than seller photos, and buyer Q&A checks all help reduce uncertainty. Checking 1688 quality before final payment walks through each of those steps. Ordinex Scout also surfaces supplier health signals, review patterns, and flagged listings so you can filter before committing to any order. Scout is currently in private beta at ordinex.cc if you want early access.