탐색으로 건너뛰기

How do I separate Xiaohongshu creator quotes, estimated CPM and actual campaign costs?

To filter Xiaohongshu creators by cost for a photo-post campaign, call data:xiaohongshu.pgy.get_blogger_list with coop.pic_price for the creator’s quote and coop.pic_cpm for the estimated cost per thousand impressions. For a video campaign, use the matching video fields in the table below. Keep quotes, platform estimates and final campaign costs in separate columns, each with the format, source and observation time.

Suppose two creators quote within budget. Compare the two creators’ estimated CPM and review how well each creator’s work fits. For the final cost per engagement, check the actual spend and which interactions the campaign report counts. Confirm both the cost and the interaction count before comparing platforms.

Choose the condition for the format

data:xiaohongshu.pgy.get_blogger_list accepts a separate coop range for each constraint:

ConstraintImage-textVideo
Quotecoop.pic_pricecoop.video_price
CPMcoop.pic_cpmcoop.video_cpm
Engagement costcoop.pic_engage_costcoop.video_engage_cost
Reading costcoop.pic_read_pricecoop.video_read_price

Each coop range is an object containing min / max. Confirm the unit of each metric before setting a range; one budget number does not define every metric.

The example request below filters image-note CPM and sorts by image-note quoted price, ascending. The threshold 100 shows the structure; 100 is not a recommended benchmark or a currency unit.

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_list","params":{"coop":{"pic_cpm":{"max":100}},"column":"picturePrice","sort":"asc","page_num":1,"page_size":20}}' \
-o creator-cost-candidates.json

Hand candidates to the collaboration team

For each candidate, keep the account ID, content direction, returned quote, the selected estimated metrics with their definitions, observation time, relevant work and open questions. The exporter in the candidate guide keeps the documented picturePrice, videoPrice, lowerPrice and estimatePictureCpm fields and preserves unavailable values.

Give an assistant that has the BeatAPI tools configured a task like this one:

Organize candidates using my confirmed image-post quote range. Keep actual quotes and estimated metrics separate with units and sources. Review recent content and audience, then give selection reasons and open checks. Do not present an estimate as my agreed campaign cost.

A complete per-creator image and video cost matrix is a separate requirement: directory filters do not supply every cost field, and unverified cost subkeys remain open. The sections below separate sorting, history and cross-platform campaign comparisons.

Is filtering CPM the same as sorting quoted prices?

No. The example request limits image-note CPM, then sorts image-note quotes ascending with picturePrice. column must use a published enum value; do not invent pic_cpm as a sorting value. videoPrice, pictureReadCost and videoReadCost are also enum values, but not every filter has a same-named sorting option.

Can I calculate a comparable CPM myself?

The general calculation is cost divided by impressions times 1000. Before calculating, verify the cost, currency, impression definition, time period and traffic scope. Reading counts are not automatically impressions, and quotes are not actual spending. The general formula does not show that a platform-provided CPM uses the same numerator and denominator.

The public catalog has no full output schema, but the creator-list response includes picturePrice, videoPrice, lowerPrice, and estimatePictureCpm. The candidate exporter preserves those four fields; the exporter does not establish every display unit, and a quote is still not an actual payment. Verify actual result definitions and quotes before making monetary decisions, and keep missing values missing instead of replacing missing values with zero.

Suppose two creators quote the same price, but one creator has an estimated CPM only for image posts and the other only for video. Put each value in the column for the value’s format. For an image-post campaign, look for a comparable image estimate for the second creator or mark the second creator as pending. After the campaign closes, store the actual spend, impression definition and period separately.

Can I retrieve a daily CPM history directly?

Not directly. The checked filter action, data:xiaohongshu.pgy.get_blogger_list, has no date or history-curve input. The core-data source explicitly returns data.data.dailyData and sumData but does not name the inner date or CPM keys, so dailyData is a trend container, not a ready-to-run daily CPM mapper. If a verified result supplies a defined CPM, you can keep same-scope snapshots with observation timestamps from now on; snapshots cannot recover unsaved historical values. Historical CPM series remain unverified, and follower-growth curves are not cost curves.

Review a creator shortlist · Organic versus all traffic

Sources: free Inspect output for data:xiaohongshu.pgy.get_blogger_list. The page contains no paid selection or measured cost comparison.

The candidate library and performance evidence guide exports the documented quotes and attaches each selected creator’s impMedian, readMedian, interactionMedian, and impMedianBeyondRate. The list source and the core trend source are different endpoints: do not divide a current quote by a median from an unrelated window and label the result as Pugongying’s estimated CPM.

Can I compare a PGY CPE directly with a Xingtu CPE?

No. Match the cost numerator and the interaction definition before comparing costs. A shared label such as CPE does not make two platform metrics the same measurement.

Post-campaign measureCost numeratorInteraction denominator
Xingtu task-report CPETask amountLikes + comments + shares
PGY note-report CPEAdvertising spend generated by using the note as an ad creativeReported total interactions, including likes, saves, comments, shares and follows across the report’s natural and advertising traffic

Xingtu defines the task-report formula in its official marketing-report guide. PGY’s official report-interface metric descriptions distinguish advertising spend, total interactions, and other cost measures. Both are definitions of the platforms’ own reports; neither shows that BeatAPI’s creator-list action returns the same fields.

Worked example: a Xingtu task amount of CNY 1,000 and 100 likes + 50 comments + 50 shares gives 1000 / 200 = 5 CNY per reported interaction. A PGY note with CNY 600 advertising spend and 300 report-defined interactions gives 600 / 300 = 2. The lower PGY figure does not show that PGY delivered the same outcome more cheaply: the PGY CPE numerator excludes the creator fee, and the PGY interaction set includes additional actions.

Is a forecast or creator quote a campaign result?

No. PGY’s official creator-screening metric descriptions describe estimated image/video engagement prices using recent cooperation-note price and interaction data. A quote, an estimated interaction cost and your own completed-order result each have a different scope. A current creator quote divided by an interaction median from a different period is a ratio you constructed, not the platform’s reported CPE.

In BeatAPI, use the documented coop.pic_engage_cost / coop.video_engage_cost filters to narrow a creator shortlist, and keep the picture and video scopes separate. The list and export workflow does not supply your task amount or every advertising-spend component. Take those amounts and the report-defined denominator from your own order or report before calculating a campaign result. Do not fill a missing follow or spend field with zero to make the calculation run.

What should the cross-platform comparison sheet contain?

Record these columns before ranking: platform, order or note identifier, observed_at, currency, cost_amount, cost_scope, engagement_count, engagement_definition, time_window, traffic_scope, source_report and ratio_status. Keep the original platform CPE in its own column. For a custom comparable ratio, use one cost scope and one interaction set that both reports contain, and label the ratio as your own calculation, not as official PGY or Xingtu CPE.

For the custom ratio, put the verified amount in A2 and the matching interaction count in B2:

=IF(AND(ISNUMBER(A2),ISNUMBER(B2),A2>=0,B2>0),A2/B2,"")

When interactions are missing or zero, the formula leaves the custom ratio blank; use ratio_status to distinguish missing_data from zero_denominator. Xingtu’s official report instead shows a CPE of 0 when the interaction count is 0. Keep Xingtu’s displayed 0 when exporting the platform report, but do not read the displayed 0 as free or exceptionally efficient engagement, and do not mix the display rule with mathematical ratios. A genuine zero cost with a positive denominator is a valid zero ratio and a different case.

This page compares the two platform definitions and provides a local spreadsheet formula; the workflow does not fetch order reports, verify actual spending, attribute campaign results, or rank real creators.

How should I review spillover store-visit cost?

Check which visits the source report attributes to spillover, the attribution window, and whether the cost is an estimate or a post-delivery report. With a source report, keep the original values, groups and definitions, then compare photo notes and videos separately. The public contract has no verified fixed spillover-cost field. Keep quotes, impressions, likes and total visits in their own columns. Leave the attributed cost open until spend and visit evidence for the attributed visits is available.