Read sales and review signals without inventing a common marketplace metric
Record labels, entities and time windows before comparing displayed sales or review figures across Chinese marketplaces.
Two numbers can look comparable while describing different things. Before making a market-research chart, record the definition available beside each displayed figure and keep unresolved meanings visible.
Apply this method to 1688, Alibaba, Taobao, JD, Pinduoduo, Weidian, Goofish and Yiwugo. An older Reddit question asks about high displayed sales and few reviews. The discussion raises uncertainty; it does not prove inflated transactions or explain current platform counting rules.
Keep the label and the entity together
Record the original label, displayed value, source link and observation time. Note whether the figure belongs to a listing, an option, a shop or another entity. Do not silently treat a shop figure as a product figure.
Also keep any visible period, unit or qualifying text. If the source does not identify a time window or whether the count concerns orders, units or people, mark that definition unresolved. Do not supply a platform-wide definition from another marketplace’s label.
Compare only figures whose meanings have been established
Consider an original fictional listing showing a sales figure of 120 and a review count of 30. Dividing one by the other does not establish a review rate unless the underlying population and time window are known to match. The numbers alone do not establish manipulation either.
Keep observed zero, unavailable and not collected separate. If a Goofish record or another source has no agreed sales field, it should remain outside that comparison rather than enter the chart as zero. A proposed export field requires sample confirmation on each included source.
State what the research can support
Describe the sources, listings, selection method and observation period used in the comparison. A selected set of listings is not a census of the platform, and displayed activity is not a verified revenue figure or a forecast of your own sales.
PandaBuyNow’s public sample pack contains no verified sales or review history. Share the exact source labels and intended research question to discuss feasible fields. Use documented definitions and visible gaps in the final analysis, with questions assigned to an appropriate reviewer.