About this skill

Problem

Product claims are often buried under marketing language, while the real buying signals appear in reviews, follow-up comments, warning posts, and competitor comparisons. Manually searching multiple platforms, normalizing evidence, and turning keywords into a usable report is easy to do inconsistently.

How It Works

selling-point-miner starts from a product name and follows a structured workflow:

Identity check: use web_search to confirm brand, company, specs, and primary sales channels before comparing data.
Multi-platform collection: prioritize Douyin, Xiaohongshu, Taobao/Tmall, JD, search engines, and industry media, capturing quotes, source URLs, and data dimensions.
Cross-validation: classify feedback as positive, negative, neutral, or scenario-bound, and count only real keyword occurrences with sentiment.
Selling-point extraction: derive points from repeated praise, unexpected delight, repurchase signals, competitor comparisons, and inferred strengths from complaints.
Report output: generate a Markdown report and summarize the core findings.

Boundaries

The skill depends on public search results and does not access protected backend data. When pages are inaccessible, samples are thin, or reviews conflict, it should preserve evidence, mark sample limits, and avoid presenting inference as fact.

Screenshots

Use Cases

Before a new product launch, extract real positive selling points from Douyin, Xiaohongshu, and Taobao reviews.
Before competitor comparison, identify competitor names and differentiators mentioned by users to complete a benchmark matrix.
During negative feedback review, classify complaints and warning posts by scenario, batch, and channel to spot contested points.
When preparing ad copy, convert real repurchase signals and unexpected delight moments into usable selling-point material.

Best For

E-commerce operators: need to turn real review signals into launch selling points and ad material.
Product managers: need to extract high-frequency positive and negative feedback plus usage scenarios when evaluating product direction.
Brand marketers: need to identify differentiators and unmet scenarios from reviews during competitor comparisons.
Market researchers: need to organize cross-platform sentiment into a deliverable Markdown report.