Douyin Competitor Research

Problem

When brands, MCNs, or creators study competitors on Douyin, they often lack traceable samples: which titles are performing, who owns them, whether duration and publish time fit the strategy, and how video differs from image posts. This skill turns competitor research into repeatable retrieval steps instead of anecdotal comparison.

How It Works

• Input: a keyword or topic direction, such as ‘low-cost fitness meals.’
• Search: calls douyin_search_videos or the corresponding CLI with filters like sort_type, publish_time_range, duration_range, and content_type.
• Sorting and filters: use time_descending for recent posts, like_count_descending for interaction strength; filter by day, week, or half year; duration under 1 minute, 1–5 minutes, or over 5 minutes; limit to video or image.
• Pagination: do not pass page_token on the first page; continue by returning the full next_page_token unchanged, with --pages or --max-items to control sample size.
• Output: preserve title, author, link or content ID, content_type, and metrics for follow-up analysis.

Boundaries

This is a read-only search capability: it does not log in, post, like, comment, or modify accounts. Results depend on the active API key and returned data. If the API returns insufficient_balance, follow the recharge prompt and retry the same command rather than looping retries. Treat images and media_type set to audio according to their actual resource type, not as default video assets.