탐색으로 건너뛰기

Which reads and impressions can I compare across Xiaohongshu campaign posts?

To look up reads and impressions for another creator’s Xiaohongshu note, call xiaohongshu.pgy.get_note_detail with the verified note ID. When present, record data.data.readNum as reads and data.data.impNum as impressions. The public detail actions xiaohongshu.app_v2.get_image_note_detail and xiaohongshu.app_v2.get_video_note_detail can return view_count; report view_count under its original name.

If note A has a read count and note B only has a save count, mark B’s reads as pending. Compare metrics that both notes report on the same basis.

What are you trying to verify?

TaskStart hereHow to use the result
Current public engagement on campaign postsSaved likes, saves, comments, shares and actual collection timeKeep known engagement even when reads are missing.
A client’s explicit impressions/read requirementThe separate PGY subset or an authorized dashboard sourceMatch the note, definition and observation window; do not replace impressions with a public field.
Content performance across creatorsCheck objects, post ages, times and comparable measures firstKeep posts with missing data visible instead of selecting only measurable successes or ranking missing values last.
Acquisition or sales returnOrder costs and attribution evidence from the relevant sourceReads, impressions and interactions are not sales.

What do missing, zero and failed mean?

  • Missing or null in a successful response: this observation did not provide that measure. Retain other known fields without inserting zero.
  • An actual numeric zero: keep the zero with its source and time. A zero does not prove that nobody encountered the content.
  • A failed request: mark the record as not retrieved and inspect the request journal. A failed request is different from a successful response that lacks an optional metric.
  • Values with different sources, times or definitions: mark them for alignment before ranking or calculating one shared ratio.

For a hypothetical note A with reads and note B with only saves, you cannot conclude that A had more reads. Compare the compatible measures both notes have, and record B’s reads as outstanding. Let the reporting question decide whether another collection is worth it; do not retry paid actions just to fill cells.

What should the comparison table contain?

For each post, keep the note ID, the metric source, the actual observation time, the values and one status: returned, not provided, failed or needs alignment. The status columns are suggestions for a human working table; the exporter does not create every judgment column. Keep every post on the roster, and state what is comparable, what is missing and who verifies each gap next.

To compare saved snapshots, use the saved-snapshot comparison workflow; editing the table needs no new query. For fuller client reporting, use the closeout evidence checklist. The field paths below are for running and checking the requests.

When you have a public note snapshot and a Pugongying report, give public views, report reads and report impressions separate columns. When you have only the public snapshot, leave reads and impressions empty. Before reporting campaign reads, confirm that the Pugongying report refers to the same note as the original link.

Where is the field?

The paths below refer to the raw /v1/social-data/call response, not to the items wrapper of a tool preview.

Detail actionNote location
xiaohongshu.app_v2.get_image_note_detaildata.data[].note_list[]
xiaohongshu.app_v2.get_video_note_detaildata.data[]

Read view_count on the matching note object. The same object can also contain liked_count, collected_count, comments_count and shared_count; these four fields are separate engagement metrics and do not substitute for impressions. A xiaohongshu.app_v2.get_video_note_detail response can contain several notes, so match the requested id instead of taking the first object.

Can I replace a missing view_count with zero or estimate views from likes?

No. A missing field means the result did not provide the metric. Record zero only when the response returns a numeric zero. The contract gives no fixed ratio between likes, saves, and views, and no formula for estimating impressions.

The two App detail schemas on this page do not promise an impression field. This scope applies only to xiaohongshu.app_v2.get_image_note_detail and xiaohongshu.app_v2.get_video_note_detail, not to every Xiaohongshu capability. For campaign reporting, check the relevant reporting action and the eligible notes separately.

How should a comparison table record view_count and missing values?

Store the note ID, the action, the collection time, and the actual metrics, plus a column that says whether view_count was present. Leave missing values blank so spreadsheet formulas do not turn blanks into zero. Counts collected at different times are not one simultaneous snapshot. Keep failed queries apart from successful notes that lack optional fields.

Choose a detail action from a share link · Search and select notes

Contract basis: free Inspect results for data:xiaohongshu.app_v2.get_image_note_detail and data:xiaohongshu.app_v2.get_video_note_detail.

Which action has a documented exposure field?

xiaohongshu.pgy.get_note_detail documents data.data.impNum for exposures and data.data.readNum for reads. xiaohongshu.pgy.get_note_detail is a separate request from the App detail actions and uses different paths. The campaign subset exporter validates note identity and the nested business result, then keeps missing values and failure coverage. The documentation defines the fields but does not guarantee a value for every note, and null is not zero. Attribution fields that are not yet verified are listed in the campaign report boundaries.