120of120
Figures disagreed across paired API rows.
120 of 120 paired API rows returned different repurchase-field values
All 120 paired rows returned both search-row repurchaseRate and detail-response sellerDataInfo.repeatPurchasePercent. Their numeric values differed in all 120, with the detail-response value higher in all 120. This establishes a field-value gap, not which formula or value is correct.
Liam Cai · Founder, SupplymoYiwu, ChinaWhat we measured
Agreement between two named repurchase fields across 120 paired API rows
The figures may use different definitions. Their disagreement is not evidence of fraud, deception or seller misconduct.
The study cannot determine which field is correct, more accurate or more useful. It did not audit either formula, denominator, scope or time window.
Study reading
Two official API fields returned different values
Both fields returned for all 120 paired rows. The values differed in all 120, and sellerDataInfo.repeatPurchasePercent was higher in all 120. The median absolute gap was 41.85 percentage points.
The median parsed search-row repurchaseRate was 12.5, while the median detail-response value was 56.05. The maximum row-level gap was 96.7 percentage points. Those values remain attached to their exact field names.
8 pairs returned zero under search-row repurchaseRate and a positive detail value. This does not prove which field better represents repeat behavior because neither definition or denominator was independently verified.
Study reading
The formulas and scopes remain unknown
Search-row values were integers in all 120 rows, while detail values were integers in 16. The detail-to-search ratio ranged from 1.2213 to 17.72, with a median of 4.1347.
That structure shows the values are not a simple rounded copy under one fixed transform. It does not prove independent algorithms or identify product-versus-seller scope, time windows, populations, return treatment or calculation rules.
The official API schemas name repurchaseRate and repeatPurchasePercent but do not provide the formulas needed to choose between them. A buyer should therefore preserve the field name and avoid turning either into a generic supplier repurchase rate.
Study reading
How a buyer can use the result
Record the exact API field and collection date whenever a repurchase percentage enters a supplier comparison. A bare percentage without source context is not comparable evidence.
Compare candidates using the same field from the same API surface. Do not use repurchaseRate for one offer and repeatPurchasePercent for another, and do not automatically choose the larger value.
Keep repurchase fields as screening clues. Before payment, verify the exact product, supplier, quote, order terms and supporting evidence rather than treating an unaudited marketplace percentage as proof of future performance.
Use this data
Everything here is free to reuse with attribution under CC BY 4.0. If you need a cut of the data we have not published, ask and we will run it.
Quote it, or take the whole file
One line, free to quote with attribution
In a Supplymo non-random sample of 120 paired first-page search rows and product-detail API responses across 12 selected English searches on 1688, search-row repurchaseRate and detail-response sellerDataInfo.repeatPurchasePercent returned different numeric values in all 120 pairs, with the detail-response value higher in all 120.
How to cite this study
Liam Cai. "Two API-returned repurchase fields on 1688." Supplymo, 15 September 2026, https://supplymo.com/research/1688-repurchase-two-figures.
Each row pairs two named API fields collected in one session. The retained source contains no offer or seller identity, so public rows cannot be rejoined independently or de-duplicated by shop. 9 stated limits travel inside the file. Free to reuse with attribution under CC BY 4.0.
Method
How the sample was collected
Everything here is reproducible from the evidence file and the sampling script.
Source
com.alibaba.fenxiao.crossborder:product.search.keywordQuery (search surface) + product.search.queryProductDetail (detail surface), read-only
Sampling frame
12 English keyword searches on the cross-border distribution channel, first results page only, 10 platform-ranked rows per keyword. For every listing, the repurchase figure was read from both the search response and the detail response in the same session.
Collection window
Single day, 2026-08-04 in China
Write operations
None. No order, payment, enquiry, messaging or logistics endpoint was called.
Check the figures yourself
Each row pairs two named API fields collected in one session. The retained source contains no offer or seller identity, so public rows cannot be rejoined independently or de-duplicated by shop.
- Rows
- 120
- Collected
- 2026-08-04
- Stated limits
- 9
Readings tested before publication+
Could the gap be a timing artefact, with the two figures read at different moments?
Method. Both figures were captured in the same sampling session, minutes apart at most, and repurchase rates are slow-moving aggregates.
Result. The gap has a consistent direction on 120 of 120 paired rows and a median size of 41.8 percentage points. Both fields were captured during the same collection session, but the study has no independent update log.
Conclusion. The uniform direction makes a simple minute-scale drift explanation unlikely, but this observational check cannot prove the cause of the difference.
Are these two renderings of the same metric, for example one rounded from the other?
Method. Test three signatures: direction consistency, integer structure, and the ratio between the two figures.
Result. Direction is uniform (detail higher on 120 of 120); every search figure is an integer while 16 of 120 detail figures are; and the detail-to-search ratio scatters from 1.2x to 17.7x with a median of 4.1x. A rounding or fixed transformation would produce a tight ratio, not this spread.
Conclusion. The values are not a simple rounded copy under one fixed transform. The study cannot establish the formulas, scope, time windows or causal reason for the difference.
Does a zero on the search surface mean no repurchase data?
Method. Compare the detail figure for every listing whose search figure is exactly zero.
Result. 8 listings show 0 on the search surface while their detail surface shows a positive figure, reaching 96.7%.
Conclusion. A numeric zero under search-row repurchaseRate does not establish that the sampled offer or seller had no repeat buyers; the field's denominator and scope were not audited.
Limits
The 9 limits of this sample
All 9 stated limitations, in the order the evidence file lists them. A number travels further than the caveat attached to it.
- 1This is not a random sample. Rows come from the first page of a keyword search, which the platform ranks.
- 2repurchaseRate and sellerDataInfo.repeatPurchasePercent are two differently named API fields. We did not audit how either is computed and cannot say which, if either, is appropriate for a particular buyer decision.
- 3The fields may differ by scope, time window, denominator, formatting or another rule. The evidence does not establish a formula, product-versus-seller level or causal explanation.
- 4Nothing here says the platform or any seller has misstated anything. Different API fields can validly return different values, and the study does not establish deceptive, fraudulent or misleading conduct.
Show the other 5 limits+
- 5Figures were captured through the cross-border API on a single day. Other channels, or the same channel on other days, may serve different values, and no trend is claimed.
- 6The 12 keywords are ordinary consumer goods in English. Other categories are not represented.
- 7The rows are listings, not distinct suppliers. Seller identity is deliberately not collected, so the sample cannot be de-duplicated by shop and one seller may appear more than once.
- 8The study reads two API responses. It does not establish what any rendered website, app, other API client or supplier conversation displayed or explained.
- 9Nothing here says a platform or seller has done anything wrong. The finding is a same-session value comparison between two named API fields, not a judgement of seller behavior or platform intent.
Research boundary
What this study cannot tell you
- The study cannot determine which field is correct, more accurate or more useful. It did not audit either formula, denominator, scope or time window.
- Different values do not establish fabrication, deception, fraud, misleading conduct or seller misconduct.
- The study compared API responses and cannot establish what a rendered website, app, other client or supplier conversation displayed or explained.
What this means for a buyer
A sample of this kind is a starting point for questions, not a verdict on any one supplier. Check the exact product and supplier you are about to pay.
Supplier verification
A free check before you commit. We run supplier checks from Yiwu, and we do not promise a lowest price.
Check a supplier →Questions a reader should ask about this
What is the one-line finding?
In a Supplymo non-random sample of 120 paired first-page search rows and product-detail API responses across 12 selected English searches on 1688, search-row repurchaseRate and detail-response sellerDataInfo.repeatPurchasePercent returned different numeric values in all 120 pairs, with the detail-response value higher in all 120. The study compares two named API fields and does not establish which value is appropriate for a buyer decision.
Which API field returned the higher value?
sellerDataInfo.repeatPurchasePercent in the detail response was higher in 120 of 120 paired rows. The study does not know why.
Does the difference prove either field is false?
No. The fields may use different scopes, time windows, denominators, formatting or other rules. Their formulas were not audited.
Does search-row zero mean there were no repeat buyers?
No such conclusion is supported. 8 pairs returned zero under search-row repurchaseRate and a positive detail-response value, but neither field's denominator was audited.
Did the study inspect the rendered 1688 website or app?
No. It compared one keyword-search API row with one product-detail API response collected in the same session. Other interfaces may present different values or explanations.