> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.beatapi.io/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.beatapi.io/_mcp/server.

# Can I retrieve another Channels creator’s play count, and how do I select strong-engagement videos?

> Select research examples using defined like and share counters, keeping observation times and missing values without relabeling read_count as plays.

Choose the metric by whose videos you are reviewing. **For your own account, use the play and completion reports you can access in Channels Assistant. For another creator's videos, shortlist examples with the returned `like_count`, `forward_count`, `comment_count` and `fav_count`.** Keep `read_count` under the original name `read_count`: the current field description of `read_count` does not support treating `read_count` as a play count or as the denominator of a completion rate.

## Shortlist examples using public engagement

1. Use the [account-video directory workflow](/faq/wechat-channels-search-pagination) to collect a set of videos with descriptions, IDs and publication times.
2. Rank the saved observations separately by `like_count` and by `forward_count`. The existing exporter already produces `metric-rankings.csv` and a companion JSON file, so ranking needs no additional detail retrieval.
3. Select relevant videos from both ranked lists and review each video's topic, publication time and presentation. A high like count is a reason to look at what makes the content appealing; a high share count is a reason to investigate why people passed the video along. Both rankings give research leads, not conclusions.
4. When you need feedback context, read the [comments and replies](/faq/wechat-channels-comments) of the selected videos.

To hand the shortlist to an assistant connected to BeatAPI, use an instruction like this:

> Within the target account's collected videos, list the top 10 separately by likes and by shares, with description, publication time, video ID and observation time. Select three product-launch examples and explain each example's content angle. List videos with unavailable counters separately.

The two ranked lists give a readable shortlist for studying the product, audience needs and presentation, instead of collapsing every video into one score.

## What does each field contribute?

| Field           | Use in this analysis                                                                                    |
| --------------- | ------------------------------------------------------------------------------------------------------- |
| `like_count`    | Compare reported likes and investigate content appeal                                                   |
| `forward_count` | Compare reported shares and select candidates with more sharing activity                                |
| `comment_count` | Select discussions worth reading; keep the count separate from comments actually retrieved              |
| `fav_count`     | Record favorites and investigate reasons people might save the content                                  |
| `read_count`    | Keep the original name and value; do not label `read_count` as plays, unique viewers or completion data |

Separate like and share rankings surface different examples, and adding the two counts together hides the difference. Keep both ranked lists unless your task defines an explicit scoring rule.

## Check time and unavailable values before ranking

Older videos have had longer to accumulate interactions, so put `create_time` and `collected_at` next to each count. For campaign research, select the publication window first, then rank the collected sample.

A reported zero can be ranked. Missing and `null` values stay unranked. Keep lower-bound labels such as “100k+” as returned; the existing helper does not turn lower-bound labels into exact ranking values. Equal counts may tie. If fewer than 10 records have usable counts, list the records that have counts and state the gap.

For example, write: “This collection selected 18 videos; 16 have exact like counts. The table shows the top 10 of those 16.” The numbers only illustrate the wording; take the actual counts from `coverage.json`. Report the collection scope and observation time with the findings.

## Does ranking need another retrieval?

No. Rank the saved pages locally. To investigate later changes, collect a new observation batch and compare the new batch with the old one. New list or detail requests are billed at current [pricing](https://beatapi.io/pricing); filtering and ranking saved data happen locally.

[Video directory and account search](/faq/wechat-channels-search-pagination) · [Channels questions](/faq/wechat-channels)