WeChat Official Account Comment Sentiment & Consumer Insights

Turn WeChat Official Account article comments and reply threads into evidence-linked consumer questions, needs, objections, purchase signals, supplied brand and competitor mentions, usage scenarios, and content opportunities.

Unofficial tool. This Actor is not affiliated with, sponsored by, or endorsed by Tencent, WeChat, Weixin, or any upstream data provider.

This is the analysis layer

Raw WeChat article scrapers return comment rows. This Actor returns:

• one normalized classification per comment or reply
• transparent positive, negative, mixed, question, or neutral labels
• recurring reader questions, needs, concerns, and intent signals
• exact buyer-supplied brand, product, and competitor mentions
• comparison and purchase-intent evidence
• article-by-article discussion differences
• ranked content opportunities backed by comment excerpts
• optional regulated-topic claim-risk screening
• explicit sample and coverage warnings

The classification is deterministic and auditable. The Actor does not claim to use AI and does not automatically invent or discover brand names.

Four input modes

1. Static preview — eight synthetic comments across three articles, without triggering the $4.99 custom report event.
2. Live article comments — one to ten public mp.weixin.qq.com URLs.
3. Apify Datasets — comments already collected by another Actor.
4. Uploaded comments — strings or JSON objects supplied by the customer.

Live mode calls the separate sian.agency/wechat-official-account-data-scraper for comments and optional reply threads. Its charges are billed separately by Apify.

Pricing

One successful live, Dataset, or uploaded-data report costs $4.99 and analyzes up to 100 comments and replies. The built-in static preview does not trigger the custom report event; Apify may still show the small Actor-start or platform usage charge. If no usable comments are collected or matched, no paid report is created.

Example

{
  "mode": "liveArticleComments",
  "articleUrls": [
    {"url": "https://mp.weixin.qq.com/s/EXAMPLE"}
  ],
  "topic": "Low-sugar breakfast reader questions",
  "commentKeywords": ["价格", "成分", "副作用", "哪里买"],
  "knownEntities": ["品牌甲"],
  "competitorNames": ["品牌乙"],
  "maxTotalComments": 100
}

Important limits

• Many articles expose no public comments.
• Results describe the collected sample, not every reader.
• Sentiment and themes use transparent cue dictionaries and may miss context.
• Exact names must be supplied to measure brands and competitors reliably.
• Comments are discussion evidence, not medical, scientific, or legal proof.

Local validation

$python3 -m unittest discover -s tests -v