Lead Generation

MCP-powered workflow for finding and qualifying B2B leads from X/Twitter conversations and profiles.

MCP Tools Used

Workflow

1. Define search queries — Build 3-5 keyword queries combining pain points, competitor names, or buying signals (e.g., "looking for {tool}", "anyone recommend {category}", "switching from {competitor}").
2. Search conversations — Call x_search_tweets for each query with limit: 30. Collect unique usernames from results.
3. Qualify profiles — For each unique username, call x_get_profile. Filter by: has bio, follower count > 100, account age > 6 months, posts in English (or target language).
4. Score intent — Assign 1-5 score based on:
   • 5: Explicit buying intent ("need a tool for...", "budget approved")
   • 4: Comparing solutions ("X vs Y", "switching from")
   • 3: Pain point discussion ("struggling with...", "frustrated by")
   • 2: Topic interest (engages with industry content)
   • 1: Tangential mention
5. Gather context — For top-scored leads (4-5), call x_get_tweets with limit: 20 to identify recent interests, tech stack mentions, and engagement patterns.
6. Check network — Call x_get_following for high-value leads to see if they follow competitors or relevant accounts.
7. Export lead list — Format as CSV-ready output using the template below.

Output Template

## Lead List: {search_topic}
Generated: {date} | Total qualified: {count}

| Username | Score | Followers | Bio Summary | Signal | Tweet URL | Notes |
|----------|-------|-----------|-------------|--------|-----------|-------|
| @{user}  | {1-5} | {count}   | {summary}   | {type} | {url}     | {note}|

### High-Priority Leads (Score 4-5)

**@{username}** — Score: {n}/5
- Signal: "{tweet excerpt}"
- Bio: {bio}
- Followers: {count} | Industry: {industry}
- Suggested approach: {personalized outreach note}

Tips

• Run searches at different times to catch varied audiences
• Refresh weekly — buying signals are time-sensitive
• Cross-reference with x_get_followers to find warm intros through mutual connections