Find Customers: Sales Leads and Comment Section Customer Acquisition

What problem it solves

Sales leads often appear in public comments on Douyin, Xiaohongshu, and Kuaishou as purchase questions, competitor complaints, or scenario pain points. Manual review is slow, and plain export tools only move data without ranking who is likely to buy. This skill turns a product description, service brief, content link, or keyword set into a follow-up lead list and an online report.

How it works

It runs through a dynamic MCP workflow using ppxc-leads-mcp:
- Collect context first: obtain productName, productDescription, platform, and a link or keywords; with a productId, it can request committee keywords via useCommitteeKeywords=true.
- Execute asynchronously: use start_search_run for single-batch runs, start_batch_search_run for batch or long-running scans, and analyze_video_comments for specific links, then poll status until complete.
- Return actionable output: higher-intent leads, source comments, positioning fields, and an online report; trial results can later be saved or unlocked into the customer pool.
- Close the feedback loop: user tags such as accurate, inaccurate, or too_broad, plus follow-up status updates, feed later lead ranking and review.

Boundaries and cautions

It identifies public-comment leads and does not guarantee conversion. Scraping uses the user's platform login state, and VERIFICATION_REQUIRED must be completed by a human. Batch jobs run serially in the background rather than as repeated synchronous waits. Full saved lists, customer-pool history, and follow-up state depend on the tool-provided workflow entries; diagnostics can be exported with export_diagnostics for support review.