ナビゲーションにスキップ

How do I expand from one suitable Xiaohongshu creator to a useful shortlist?

To expand from one suitable creator, pass the creator’s user ID to xiaohongshu.pgy.get_blogger_similar for platform recommendations. To add conditions such as a Shanghai-based creator, an audience size or a photo-post budget, call xiaohongshu.pgy.get_blogger_list with similar_user_id plus your filters. Then review each candidate against content, audience and cooperation needs.

For example, for a Shanghai coffee-machine event, the reference creator explains home-coffee techniques well. Write down why the reference creator fits: clear operating tutorials, an audience interested in home equipment and suitability for image-led content. List firm requirements, such as creator location and image-post quotation range, separately from the content-style and audience questions that still need review.

Which discovery route fits?

Current taskRoute
See which accounts the platform recommends as similarxiaohongshu.pgy.get_blogger_similar with the seed’s user_id
Add follower size, location or cooperation filtersxiaohongshu.pgy.get_blogger_list with similar_user_id and filters
Start with a topic but no seed creatorUse campaign filters, then select a supported reference

Pass the creator’s user ID, not a post’s note_id. Save the seed creator and your selection reason so the team can see where the candidate list came from.

Start with a recommendation page

curl -sS https://api.beatapi.io/v1/social-data/call \
-H "Authorization: Bearer $BEATAPI_API_KEY" \
-H 'Content-Type: application/json' \
-d '{"action":"xiaohongshu.pgy.get_blogger_similar","params":{"user_id":"<seed_user_id>","page_num":1,"page_size":4}}' \
-o similar-creators-page-1.json

xiaohongshu.pgy.get_blogger_similar accepts page_size 1–4 (default 4) and page_num 1–100. A page holds at most four candidates; work with the candidates actually returned. Review the first batch for topic fit, then request further pages only when the shortlist still has a gap.

How do you add campaign conditions?

Call xiaohongshu.pgy.get_blogger_list with similar_user_id plus the supported fans.count, blogger.location and coop conditions. The filter FAQ shows how to nest the conditions. Use creator location for an attendance requirement and audience location for market reach.

similar_word only echoes the seed creator’s nickname and does not affect recommendations. A keyword search for the nickname is a different discovery method.

To hand the task to an assistant with the tools configured, use a prompt such as:

Use this home-coffee tutorial creator as a reference for a Shanghai event shortlist. Retain their user ID and my selection reasons. Start with one recommendation page; use directory similarity inputs when adding location and image-post conditions. Remove duplicates and existing candidates. Explain why each remaining person deserves closer reading and what still needs checking.

How should the candidate table support a decision?

For each candidate, save the user ID, seed, filters, page, collection time and request ID, then add these review columns:

ColumnWhat to record
Content fitRelevant tutorials or content examples, beyond a coffee-related nickname
Audience questionsRequired location, interests or scale, separating known information from unreviewed items
Cooperation conditionsQuotations for the intended format and event requirements, with their sources
Next stepKeep, read more content, check audience or pause follow-up, with a reason

First deduplicate and remove creators already on the shortlist. Similarity recommendations supply discovery leads; content and audience evidence decide the campaign. Skip further profile work for clearly irrelevant creators. For promising candidates, continue with audience review or quotation comparison.

Observed xiaohongshu.pgy.get_blogger_list responses carry candidates at data.data.kols with userId. xiaohongshu.pgy.get_blogger_similar has no recorded fixed output schema, so read the saved recommendation result instead of assuming the same list path or a numeric similarity score. For a complete candidate table, continue with the shortlist guide.

How do you plan this round?

Count recommendation pages, filtered pages and the follow-up lookups selected candidates need, and price them with current pricing. For example: one recommendation page, then audience checks for the two creators you select. The deliverable is a shortlist with reasons, so request further data only where the shortlist needs it.