Competitor Viral Analyzer

Problem

• Analyzing competitor accounts often relies on scrolling posts and gut summaries, which misses dimensions and turns one viral post into a false pattern.
• Teams need to understand which title hooks, content structures, emotional curves, pacing choices, and sharing triggers are reusable.
• The output should not just describe success; it should convert patterns into differentiated topic ideas.

How It Works

• Collection: parses a competitor profile, handle, or URL into platforms such as WeChat Official Accounts, Douyin, Bilibili, Xiaohongshu, and Channels, then gathers recent 10-15 high-engagement posts.
• Scoring: uses viral_scorer.py to compare engagement growth against the account average, separating viral, near-viral, and ordinary posts, then focuses on the top 5-8.
• Five-Dimension Breakdown: each post gets a card covering title hooks 25%, structure 20%, emotion 20%, rhythm 15%, and spread factors 20%.
• Topic Synthesis: extracts common title formulas, structure templates, emotional tactics, topic preferences, and persona traits, then generates benchmark, gap-filling, and extension topics with explicit differentiation.
• Reporting: organizes findings into competitor overview, viral ranking, pattern extraction, topic library, and action recommendations.

Boundaries

• Best for asking “why this creator works” and “how to differentiate in the same niche”; not for broad hot-topic discovery or scoring existing candidates.
• It requires real, collectable competitor content; if fewer than 5 posts can be analyzed, it switches to a limited analysis mode and states the constraint.
• The topic library usually contains 8-20 ideas and should cover at least 3 different viral themes to avoid overgeneralizing from one sample.