Statistics index

1688 statistics, measured rather than estimated

Almost every published figure about Chinese wholesale marketplaces is an estimate of size. These are not. Each number below is a count from a sample we read through the platform’s own read-only API, and each one links to the study that produced it, the evidence file behind it and the limits that apply to it.

6 studies · 688 listing-level rows · most recent collection 12 August 2026. Free to cite with attribution.

Liam Cai · Founder, Supplymo

Maintained from Yiwu, China · Updated as each study publishes

The figures, with their denominators attached

A number without its sample size is not a statistic. Every line below carries both, plus the date it was collected.

Where the shipping weight on a 1688 listing comes from

117 of 120 top-ranked 1688 listings source their shipping weight from the merchant's own entry, and only 3 carry a weight measured by a fulfilment centre.

Collected 2026-08-12 · 120 rows · 8 stated limits. This records the provenance a listing labels for its own weight figure. It does not verify whether any weight is accurate, and it is not a claim that any seller misstated anything.

Method and evidence CSV

Where the shipping weight comes from on 120 top-ranked 1688 listings

n = 120 listings. Collected 2026-08-12. Non-random sample: 12 keyword searches, 10 platform-ranked rows each. All 120 listings displayed a shipping block.

Dispatch promises on 1688 listings

55 of 120 top-ranked 1688 listings returned no dispatch-promise field at all.

Collected 2026-08-11 · 120 rows · 10 stated limits. This measures what a listing displayed before payment, not whether a seller would accept a return or dispatch on time.

Method and evidence CSV

What 120 top-ranked 1688 listings showed for dispatch promise

n = 120 listings. Collected 2026-08-11. Non-random sample: first page of 12 keyword searches, 10 platform-ranked rows each.

Certificates displayed on 1688 listings

156 of 160 top-ranked 1688 listings in compliance-review categories displayed no certificate at all.

Collected 2026-08-05 · 160 rows · 11 stated limits. A listing that displays no certificate may still hold one. This measures display, not possession, and it is not a safety assessment.

Method and evidence CSV

Certificates displayed by 160 top-ranked 1688 listings in compliance-review categories

n = 160 listings. Collected 2026-08-05. Non-random sample: 8 categories, 20 platform-ranked rows each.

Invoice types declared on 1688 listings

72 of 120 top-ranked 1688 listings declared no invoice type in the platform's invoice field.

Collected 2026-08-08 · 120 rows · 11 stated limits. A listing that declares nothing may still issue an invoice on request. This measures the declaration, not the capability.

Method and evidence CSV

Invoice types declared by 120 top-ranked 1688 listings

n = 120 listings. Collected 2026-08-08. Non-random sample: 12 categories, 10 platform-ranked rows each.

Headline supplier scores versus their components

65 of 120 top-ranked 1688 listings showed the maximum 5.0 supplier score, and 21 of those carried a component score at or below 2.5.

Collected 2026-08-04 · 120 rows · 11 stated limits. The component scales are narrower than the headline scale, so the two are not directly comparable. This is a reading of published scores, not an audit of supplier quality.

Method and evidence CSV

Headline supplier score across 120 top-ranked 1688 listings

n = 120 listings. Collected 2026-08-04. 21 of the 65 maximum-score listings carried a component score at or below 2.5.

Price spread between visually identical 1688 listings

41 of 48 top-ranked 1688 listings had a cheaper, visually matching listing on the first page of the platform's own image search.

Collected 2026-08-03 · 48 rows · 9 stated limits. A visual match is not proof that two listings are the same product. Listed prices are not quoted prices, and they exclude shipping, tooling and terms.

Method and evidence CSV

Cheaper visual matches for 48 top-ranked 1688 listings

n = 48 seed listings. Collected 2026-08-03. Median seed-to-cheapest listed-price ratio 2.25×. A visual match is not proof of an identical product.

How these numbers are produced

Every study follows the same shape. A keyword or category frame is fixed first, then rows are read from the platform’s own search and detail endpoints in a single session. Whether a field was returned at all is recorded separately from its value, so an absent field is never counted as a zero. Every published figure is then recomputed from the row-level data by a checker that runs before the site can be deployed, and each study ships with the raw file so anyone can repeat the arithmetic.

What none of these studies can tell you is what a supplier will actually do after payment. They measure what a buyer can see before it.

All research, with methods

Questions about using these figures

Where do these 1688 statistics come from?

Each figure comes from a Supplymo sample read through 1688's own read-only search and product-detail API. 6 separate studies, 688 listing-level rows in total, each published with its raw evidence file, its method and its stated limits.

Can I cite these numbers?

Yes, with attribution to Supplymo and a link to the study the figure comes from. The row-level data is downloadable as CSV under CC BY 4.0. If you need a cut we have not published, ask and we will run it.

Are these samples representative of all of 1688?

No. Every sample is drawn from the first page of keyword or category searches, which the platform ranks. They describe what a buyer landing on a highly ranked listing sees, not an average across the whole marketplace. Each study states this limit and the others in its evidence file.

Do these figures say suppliers are doing anything wrong?

No. Every study measures what a listing displays through platform fields before payment. A listing that shows nothing may still ship on time, hold a certificate or issue an invoice. The gap being measured is visibility to the buyer, not supplier behaviour.