> 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.

# How do I find a Channels account’s product-launch videos and select examples to study?

> Choose an account and topic, review a bounded search collection and investigate selected videos for messaging patterns and unanswered questions.

To find a Channels account's product-launch videos, search inside the confirmed account with `data:wechat_channels.v2.fetch_search_channel_videos`, passing the account's `v2_…@finder` username and a launch keyword. Save up to two result pages, write a reviewed shortlist with a selection reason per video, and retrieve details only for the shortlisted videos. Account-topic search answers “what has this account covered?” without reviewing the account's entire video directory.

## 1. Define the examples you need

For launch research, start with a product-launch keyword, then sort the matches into announcements, tutorials, promotions and unrelated matches. Decide what the comparison examines, for example each video's opening, main benefit, intended audience and the action offered to viewers.

For an assistant connected to BeatAPI:

> Search this account for product launches and review at most two pages. Select videos that introduce a new product. Retain their identifiers and returned descriptions with selection reasons. Retrieve details after selection, then compare how they explain benefits and use cases.

## 2. Confirm the account

Use the account's `v2_…@finder` username. If you have an `sph…` short ID or a video share link, follow the [identifier workflow](/faq/wechat-channels-identifiers) to obtain and check the account. If a canonical `v2_…@finder` username is already available, reuse the username.

`data:wechat_channels.v2.fetch_search_channel_videos` takes the account and the keyword and searches only that account. For the account's overall publishing activity, use the [video directory](/faq/wechat-channels-search-pagination) instead.

## 3. Prepare and send the first request

Save the following script as `make_request.py`. `make_request.py` writes the account, keyword and output mode into a request file with an initially empty cursor, and refuses to overwrite an existing request file.

```python
import json, pathlib, re, sys
username, keyword, cursor, output = sys.argv[1:5]
keyword = keyword.strip()
if not re.fullmatch(r"v2_[0-9a-fA-F]+@finder", username) or not 10 <= len(username) <= 256:
    raise ValueError("Expected finder username")
if not 1 <= len(keyword) <= 100 or len(cursor) > 4096:
    raise ValueError("Invalid keyword or cursor length")
body = {"action":"wechat_channels.v2.fetch_search_channel_videos",
        "params":{"username":username,"keyword":keyword,"cursor":cursor,"raw":True}}
with pathlib.Path(output).open("x", encoding="utf-8") as f:
    json.dump(body, f, ensure_ascii=False, indent=2)
```

Replace the account and keyword with your actual target:

```bash
python3 make_request.py '<finder_username>' 'new product' '' search-page-1-request.json
curl --fail-with-body -sS https://api.beatapi.io/v1/social-data/call \
  -H "Authorization: Bearer $BEATAPI_API_KEY" \
  -H 'Content-Type: application/json' \
  --data-binary @search-page-1-request.json -o search-page-1-response.json
```

The request uses `raw:true`; keep `raw:true` for the next page. See current [pricing](https://beatapi.io/pricing). Account resolution, details for selected videos and comments are separate retrieval steps, each requested only when needed.

## 4. Review whether the results fit the research

1. Check HTTP success and business success, then save the request, the response, the `request_id` and the collection time.
2. Check that the returned videos belong to the target account. Record each usable video identifier, description and source position in the shortlist.
3. Give each video a selection reason, such as “launch announcement explaining the intended audience,” or an exclusion reason, such as “old-product promotion.” A keyword match makes a video a candidate; the video's context decides whether the video belongs in the research.

The search response has no fixed search-card projection, and no verified parser maps search results to titles and video links. Review the returned objects directly; you or your assistant edits the shortlist item by item.

For example, searching “new product” within one account finds a launch teaser, a product demo and a clearance sale. To study how the account explains product benefits, shortlist the product demo and record the demo's video ID and your reason. Keep the launch teaser as an example of building interest before launch, and record why you excluded the clearance post. Keep each decision beside the search page reference.

## 5. Continue to page two and investigate selected videos

1. When more candidates are needed, pass the returned `next_cursor` unchanged as the third argument of `make_request.py` and save a separate `search-page-2-request.json`. Keep the account, the keyword and `raw:true` unchanged.
2. After two pages, record the batch's actual scope. Decide any further pages as a separate scope.
3. Merge repeated video IDs in the shortlist and keep every source. If only `exportId` is available, save the `exportId` value and pass the value as the detail input `export_id` to locate the video; do not guess a numeric ID from the title. Confirm the video and the account after retrieving details.
4. If `next_cursor` is empty or repeated, or a request fails, record the reason and keep the successful first page. Pages that were not retrieved are outside the completed scope.

## 6. Deliver examples and findings

The shortlist columns are the video identifier, the returned description, the source page, the collection time, the selection reason and the questions that need detail review. Analyze benefits and presentation from the selected videos' details, and retrieve viewer discussion through the [comment workflow](/faq/wechat-channels-comments).

Attach each finding to the video it comes from, and separate what the video states from your interpretation. Report the pages checked, the videos selected and the items that still need detail verification. The saved search pages stay reusable when you try other keywords later.

`make_request.py` only prepares requests. The relevance review and the shortlist are manual edits, and reading saved search pages needs no new retrieval.

[Video directory and search pagination](/faq/wechat-channels-search-pagination)