How to Source From 1688: A Practical Guide for TikTok Shop and Shopee Sellers

Most sellers start on 1688 by typing a search keyword, opening dozens of product tabs, then picking based on gut feeling. Some products win. Most don't, and they never understand why.

Sellers who win consistently on TikTok Shop and Shopee take a different approach. They don't hunt for individual products. They filter systematically.

Here's how to do it right.

What 1688 actually is

1688 (1688.com) is Alibaba's domestic wholesale marketplace, built for Chinese businesses buying from factories and distributors within China. Unlike Alibaba International, everything is in Chinese, prices are in CNY, and there are no mandatory minimum order quantities. That is both the opportunity and the barrier.

For cross-border sellers in Southeast Asia, 1688 is the most practical source of intelligence on what exists, what is priced competitively, and what is selling well in China before those products appear in SEA markets 3 to 6 months later. Catching the trend early is the advantage.

Start with categories, not specific products

The most common mistake: going to 1688 to search for a specific product.

Browse by category instead. Export the full list. Filter before you evaluate.

Your edge is not finding a good product. It is finding it before the next seller does. Browsing by category surfaces products you would never think of if you only searched by keyword. That is where asymmetric opportunities live: products with real demand in China, no dominant seller on your target marketplace, and room to build a content advantage.

Export first, evaluate later

1688 lets you export product listings as Excel files. Use this feature every time you source.

Get everything into one place before you start filtering. Browsing and evaluating at the same time is the fastest way to miss products, lose focus, and waste an hour without making a single decision.

Once you have the full list:

1. Remove obviously unqualified products: poor images, no sales history, abnormal pricing
2. Apply basic filters: minimum monthly sales volume, product age, price range
3. What remains is your evaluation batch, the set that actually deserves your attention

This two-step approach, collect then evaluate, is the single biggest productivity improvement most operators can make immediately.

Three metrics that actually matter

Monthly sales volume

Is anyone buying this product? A listing with zero or single-digit monthly orders is a bet on risky inventory. You need evidence of market demand before committing capital.

Reference threshold: 50 orders per month to consider; 500 orders per month to analyze seriously.

Repurchase rate

Repeat purchases signal a genuinely useful product. Products with returning buyers have staying power on Shopee (where the algorithm rewards consistent sellers) and TikTok Shop (where repeat customers contribute to shop trust scores).

High repurchase rate is the strongest signal we have found for identifying products that last, not just spike and die.

Product age

Products with 12+ months of steady sales on 1688 have survived long enough to be credible. New listings with sudden sales spikes are a yellow flag: could be a real trend, or could be a one-month peak before returns flood in.

We treat high-volume new listings as conditional. Keep them in the batch but cross-check on TikTok before placing any orders.

Filter fast, decide slow

The first pass is elimination: remove what clearly does not qualify. This takes seconds per product, not minutes. You are not thinking. You are filtering.

The second pass is evaluation: the remaining batch gets real attention. This is when you look at the numbers, check TikTok, and make decisions.

Merging these two steps is where most wasted time comes from. Do not evaluate products that should have been eliminated in the first pass.

Cross-check before ordering

Before placing any order, search for the product on TikTok and Shopee:

Nobody is selling it. There is usually a reason. No demand, import difficulty, unsustainable margins, or compliance issues. Seller absence is data.

Everyone is selling it. Margins are likely gone. You would enter at the bottom of a price war with nothing to differentiate.

Moderate competition, solid foundation. This is the signal. A few sellers but no dominant player, real sales history, and room to differentiate on content, packaging, or speed.

The TikTok check is especially important. If a product is generating organic views and orders on TikTok Vietnam or TikTok Thailand right now, consumer demand is real and visible. That signal is stronger than any metric on 1688.

Common mistakes

Browsing and evaluating at the same time. You will be slower and less consistent. Always export first.

Ignoring product age. New listings with high sales are a risk, not a green light.

Skipping the TikTok check. 1688 data tells you about supply-side demand. TikTok tells you about consumer demand. You need both before deciding.

Focusing only on price. The cheapest supplier on 1688 is not always the best. Minimum order quantities, lead times, and reorder reliability matter more at scale.

Evaluating products one by one. Effective filtering works in batches. If you are reviewing individual products in isolation, you are losing to operators who move 10x faster.

Not recording decisions. If you cannot remember why you passed on a product 3 months ago, you will evaluate it from scratch. Every pass decision needs a note.

FAQ

Do I need a shipping agent to import from 1688?
Yes. 1688 is a domestic Chinese marketplace. It only ships within China. You need a consolidation warehouse in China and a shipping agent to move goods to Vietnam, Thailand, or wherever you sell. Many services specialize in 1688 to SEA routes and can consolidate multiple shipments from different suppliers.

Does 1688 require minimum order quantities?
Unlike Alibaba International, most 1688 suppliers do not enforce mandatory MOQs. You can test with small quantities, but unit prices will reflect that. Negotiate directly when ordering at scale.

How do I handle the language barrier?
Chrome auto-translate handles most product browsing. For supplier communication, Google Translate combined with WeChat is sufficient at small scale. At higher volumes, a professional Chinese purchasing agent is worth the cost.

How many products should I screen per batch?
Depends on your category and growth stage. Beginners typically screen 100 to 500 products per batch. At full capacity, our operation processes 2,000 to 5,000 products per week with structured tooling. That is not possible with spreadsheets.

The real advantage is process, not product

Operators who win long-term are not the ones who find the best product. They are the ones with the best repeatable system for finding products, week after week, batch after batch, without starting from zero every time.

That means: consistent filters, a record of what was passed and why, and notes on winning products so you learn what you are actually good at sourcing.

We built because we could not find a tool that does this. Spreadsheets do not scale. Manual 1688 browsing does not compound.

If you are screening more than 200 products per week, a structured process is not optional. It is the line between a business that grows and one that stays stuck.