Skip to navigation

Where to get Xiaohongshu notes, comments and account data

Where you get Xiaohongshu data depends on whose notes you study. For reads and impressions on your own notes, start in Creator Center. For a paid collaboration, use the Pugongying or campaign report your team can access. To see what readers ask under other creators’ notes, open a few known note links first. For a recurring sample, search with BeatAPI’s xiaohongshu.app_v2.search_notes, then fetch comments for the selected notes.

Where should you start for your own notes or campaigns?

QuestionStart hereRecord
How did our notes perform?Sign in to the app’s Creator Center, open recent data and choose a note. Store operators can use the notes center or note analytics in Qianfan.Note identifier, metric label, report window and export time
What did the campaign achieve?Use the Pugongying or advertising report available to your team, and match the report to the project and post.Campaign, dates, definitions of impressions and reads, attribution window
What user data can our app request after login?Check the account platform’s scopes and your app’s approval.Approved scope, authorized account and returned fields

The Xiaohongshu account platform lists basic_info as available and read_notes as planned. The scope list covers the account integration only; the marketing platform and mini-apps have their own capabilities and requirements. Verify access with your app’s approved permissions and the actual response. Account scopes · Marketing platform · Official note analytics example

What should the first public research batch include?

For two or three known posts, open the original posts and record what matters by hand. For repeat discovery, send a POST to https://api.beatapi.io/v1/social-data/call with Authorization: Bearer <BEATAPI_API_KEY> and Content-Type: application/json. Start with one topic and one page:

{"action":"xiaohongshu.app_v2.search_notes","params":{"keyword":"咖啡机清洁","page":1}}

Check that both the HTTP request and the action succeeded, then review candidate notes under data.data.items[].note. Save the query, note ID, title, observed interactions and raw response. Use search excerpts to select notes. For user complaints, read a chosen comment sample and keep the original wording and context. Report how many distinct notes support each finding. Mark missing reads, impressions and attribution fields as missing. Search export · Reads and impressions

Pick an export based on what you have

Starting pointWhat to hand overWhere to start
A keyword or saved search pagesCandidate notes, excerpts and selection reasonsSearch CSV export
A verified creator profileOne inventory row per noteCreator note inventory
Links or IDs for selected notesEngagement snapshots or a table of sourced commentsMetric snapshots · Comment research
Creator criteria or selected productsPartnership candidates or reviews grouped by SKUCreator shortlist · Product reviews

For example, pick two relevant notes from the keyword CSV, then retrieve comments for those two notes only. Join the candidate sheet and the comment sheet on the note ID, and keep the scope of each lookup.

How should the monthly report guide your next topic?

Suppose you need to decide whether to make a descaling tutorial. In Creator Center, choose one reporting month and find two of your own cleaning notes. Record reads and impressions with the dashboard’s labels and date range. From the public sample, save two comments about lingering odors from two different notes. Keep each original post, comment text and observation time in a separate topic-research sheet.

The two-comment sample supports “Some users asked about odors after descaling. Include a troubleshooting segment.” To assess how widespread the odor issue is, apply the same selection rule to more notes, and report how many notes contained the issue out of the notes you actually reviewed.

If you still need to know when the odor occurs, focus the next round on comments that mention the machine or cleaner. In the monthly report, put your own tutorial’s impression changes and the new questions from public samples on separate rows, with the dashboard period and the original notes. The team can then decide whether to review an existing tutorial or write new instructions.

Compare the effort of a manual, existing or pay-per-request workflow

Ask each workflow for the same deliverable, such as ten sourced candidates and three verifiable user questions. Measure usable results, missing entries, review time and repeated work. For a tool you maintain, replay a saved response through the export, and include operating, maintenance and review time.

For ten single-page searches and three single-page comment lookups each week, four weeks means 40 search requests and 12 comment requests. Price the search and comment actions separately, and add any detail lookups or extra pages. Compare the cost of delivering one usable batch. BeatAPI pricing