influencer-finder

Influencer Finder

Discovers Key Opinion Leaders (KOLs) and micro-influencers whose audiences overlap with your ICP. Searches across LinkedIn, Instagram, YouTube, and Twitter/X to find creators with genuine engagement, topical relevance, and audience alignment. Evaluates authenticity signals, content quality, and partnership potential. Outputs a ranked influencer list with collaboration recommendations and outreach templates.

Prerequisites

• agency.config.json at repo root with services, icp, and optionally outreach sections
• WebSearch tool available
• Optional: linkedin-researcher skill for deeper LinkedIn creator analysis
• Optional: person-researcher skill for enriching influencer profiles
• Optional: visual-brand skill for assessing visual alignment with your brand

Phase 0: Intake

1. Read agency.config.json from the project root.
2. Extract:
   • agency.name, agency.domain -- for partnership context
   • services[].name, services[].keywords -- topic alignment criteria
   • icp.segments[].industries, icp.segments[].titles -- audience overlap targets
   • icp.segments[].markets -- geographic relevance
   • icp.primary_keywords, icp.secondary_keywords -- content relevance signals
   • outreach.tone -- voice alignment for outreach templates (if present)
3. Accept parameters:
   • niche -- primary niche/topic to search (default: derive from config service keywords)
   • platforms -- platforms to search: linkedin | instagram | youtube | twitter (default: all)
   • follower_range -- min-max follower count (default: "1000-100000" for micro-influencers)
   • geography -- target regions (default: derive from ICP markets)
   • content_types -- preferred content types: video | carousel | text | podcast | newsletter (default: all)
   • max_results -- max influencers to return (default: 30)
   • partnership_type -- collaboration goals: content | co-marketing | affiliate | guest_post | podcast_guest (default: all)

Phase 1: Define Influencer Criteria

Build a scoring rubric based on config and parameters:

Relevance criteria:

• Topic alignment: Content covers topics in services[].keywords or related adjacent topics
• Audience overlap: Their followers match ICP titles, industries, and seniority levels
• Geographic fit: Creator or audience concentrated in ICP markets
• Platform presence: Active on platforms where ICP audience spends time

Quality criteria:

• Engagement rate: Comments and shares relative to follower count (>2% good, >5% excellent)
• Content consistency: Regular posting cadence (at least weekly)
• Content depth: Substantive posts vs shallow motivational content
• Authenticity signals: Real comments from real people, not bots or engagement pods

Disqualification criteria:

• Follower count outside specified range
• No posts in the past 30 days (inactive)
• Engagement that looks artificial (thousands of likes, zero meaningful comments)
• Content exclusively self-promotional with no educational value
• Direct competitor or their employee

Phase 2: Platform-Specific Discovery

LinkedIn Discovery

Search queries:

• site:linkedin.com/in "{service_keyword}" "followers" "{industry}"
• "top linkedin creator" "{niche}" "{icp_market}"
• "linkedin influencer" "{service_keyword}" "posts about"
• site:linkedin.com/newsletters "{service_keyword}" "{industry}"

Extract for each:

{
  "name": "Full name",
  "profile_url": "LinkedIn URL",
  "headline": "LinkedIn headline",
  "follower_estimate": "Approximate count",
  "content_focus": "Primary topics they cover",
  "posting_frequency": "Daily | 3-4x/week | Weekly | Irregular",
  "newsletter": "Newsletter name if they have one"
}

Instagram Discovery

Search queries:

• site:instagram.com "{service_keyword}" "{industry}" creator
• "instagram influencer" "{niche}" "{icp_market}" followers
• "{service_keyword}" "instagram" "content creator" "{industry}"
• "top instagram accounts" "{niche}" "{year}"

Extract for each:

{
  "handle": "@username",
  "profile_url": "Instagram URL",
  "follower_estimate": "Approximate count",
  "content_style": "Reels | Carousels | Stories | Mixed",
  "visual_quality": "HIGH | MEDIUM | LOW",
  "niche_focus": "Primary topic area"
}

YouTube Discovery

Search queries:

• site:youtube.com "{service_keyword}" channel "{industry}"
• "youtube channel" "{niche}" "{icp_market}" subscribers
• site:youtube.com "{service_keyword}" "tutorial" OR "guide" OR "review"
• "best youtube channels" "{niche}" "{year}"

Extract for each:

{
  "channel_name": "Channel name",
  "channel_url": "YouTube URL",
  "subscriber_estimate": "Approximate count",
  "content_type": "Tutorials | Reviews | Vlogs | Interviews | Mixed",
  "avg_views": "Rough average per video",
  "upload_frequency": "Weekly | Bi-weekly | Monthly | Irregular"
}

Twitter/X Discovery

Search queries:

• site:twitter.com OR site:x.com "{service_keyword}" "{industry}" bio
• "twitter influencer" "{niche}" "{icp_market}"
• site:twitter.com "{service_keyword}" "thread" followers
• "top twitter accounts" "{niche}" "{industry}" "{year}"

Extract for each:

{
  "handle": "@username",
  "profile_url": "Twitter URL",
  "follower_estimate": "Approximate count",
  "content_style": "Threads | Short-form | Curated | Mixed",
  "engagement_pattern": "High replies | High retweets | Community-driven"
}

Phase 3: Profile Enrichment

For the top candidates from each platform (top 50% by initial relevance):

Cross-platform presence check:

• Search for "{influencer_name}" site:linkedin.com if found on other platform first
• Search for "{influencer_name}" site:instagram.com if found on LinkedIn first
• Map multi-platform presence: more platforms = higher reach potential

Engagement quality assessment:

• Search for "{influencer_name}" "{service_keyword}" comments OR replies to gauge discussion quality
• Look for engagement pod signals: same people commenting on every post, generic comments
• Check if they respond to comments (community builder vs broadcaster)

Content quality assessment:

• Search for "{influencer_name}" "{niche}" article OR post OR video to sample content depth
• Categorize content mix: educational (how-to, tutorials) vs opinion (hot takes) vs promotional (sponsored)
• Note unique angles or perspectives they bring

Audience alignment estimation:

• If they have a newsletter, check topic alignment with ICP interests
• If they speak at events, check event types (industry conferences = good audience overlap)
• If they have a podcast, check guest types (ICP-matching guests = aligned audience)

Phase 4: Scoring and Ranking

Score each influencer on a 100-point scale:

Topic relevance (0-30 points):

• Content directly covers your service keywords: 25-30
• Content covers adjacent topics: 15-24
• Content is in the right industry but different angle: 5-14
• Tangential relevance: 0-4

Audience overlap (0-25 points):

• Audience clearly matches ICP titles and industries: 20-25
• Audience partially overlaps: 10-19
• Audience is in the right industry but wrong seniority: 5-9
• Unclear audience composition: 0-4

Engagement quality (0-20 points):

• High engagement rate + quality comments: 16-20
• Good engagement rate + mixed comment quality: 10-15
• Low engagement rate or suspicious patterns: 0-9

Accessibility (0-15 points):

• Micro-influencer, likely to respond: 12-15
• Mid-tier, may respond with good pitch: 7-11
• Large following, hard to reach: 0-6

Authenticity (0-10 points):

• Clear genuine engagement, no pod signals: 8-10
• Mostly authentic with some questionable patterns: 4-7
• Suspected engagement manipulation: 0-3

Tier assignment:

• TIER 1 (80-100): High-priority partnership targets
• TIER 2 (60-79): Strong candidates, worth pursuing
• TIER 3 (40-59): Secondary targets, pursue if bandwidth allows
• BELOW THRESHOLD (<40): Skip

Phase 5: Partnership Recommendations

For each TIER 1 and TIER 2 influencer, generate:

Collaboration format recommendation:

{
  "recommended_formats": [
    {
      "format": "guest_post | co-created_content | product_review | podcast_interview | joint_webinar | affiliate | ambassador | event_collab",
      "fit_reason": "Why this format works for this influencer",
      "estimated_effort": "LOW | MEDIUM | HIGH",
      "estimated_reach": "Number range based on their audience size"
    }
  ],
  "outreach_angle": "Specific reason to reach out that references their content",
  "value_proposition": "What you bring to them (not just what they bring to you)",
  "talking_points": ["Point 1 referencing their recent content", "Point 2 about mutual benefit"],
  "avoid": "Topics or approaches that would not resonate"
}

Outreach template:

Generate a personalized outreach message skeleton for each TIER 1 influencer:

• Reference a specific piece of their content
• Explain mutual benefit clearly
• Propose a specific low-commitment first step
• Keep it under 150 words

Phase 6: Output

Return structured influencer report:

{
  "discovery_summary": {
    "platforms_searched": ["LinkedIn", "Instagram", "YouTube", "Twitter"],
    "total_candidates_found": 85,
    "after_scoring": 30,
    "tier_breakdown": {
      "tier_1": 5,
      "tier_2": 12,
      "tier_3": 13
    },
    "top_niches": ["Shopify development", "D2C marketing", "Ecommerce CRO"]
  },
  "influencers": [
    {
      "name": "Creator Name",
      "platforms": {
        "linkedin": { "url": "...", "followers": "..." },
        "instagram": { "url": "...", "followers": "..." }
      },
      "primary_platform": "LinkedIn",
      "niche": "Ecommerce growth",
      "content_focus": ["Shopify tips", "D2C strategy"],
      "follower_total": "25,000 across platforms",
      "engagement_rate": "4.2%",
      "engagement_quality": "HIGH",
      "authenticity_score": 9,
      "total_score": 87,
      "tier": "TIER_1",
      "audience_profile": "D2C founders and ecommerce managers in India/US",
      "partnership_recommendations": [],
      "outreach_template": "...",
      "recent_content_sample": "Title or topic of recent post"
    }
  ],
  "platform_insights": {
    "linkedin": { "candidates": 20, "avg_score": 68, "top_niche": "..." },
    "instagram": { "candidates": 15, "avg_score": 55, "top_niche": "..." },
    "youtube": { "candidates": 10, "avg_score": 62, "top_niche": "..." },
    "twitter": { "candidates": 8, "avg_score": 52, "top_niche": "..." }
  },
  "recommendations": {
    "priority_outreach": ["Name 1", "Name 2", "Name 3"],
    "quick_wins": "Micro-influencers who are easy to reach and highly relevant",
    "long_term_targets": "Larger creators worth building relationships with over time",
    "partnership_budget_estimate": "Range based on tier and format"
  }
}

Present formatted summary:

INFLUENCER DISCOVERY REPORT
Searched: {platforms} for "{niche}"
Total found: {N} candidates, {scored} scored

TIER 1 -- PRIORITY TARGETS ({count}):
1. {name} -- {primary_platform} -- {followers} followers -- Score: {N}/100
   Niche: {niche}
   Engagement: {rate}% ({quality})
   Best collab format: {format}
   Outreach angle: {angle}

TIER 2 -- STRONG CANDIDATES ({count}):
1. {name} -- {primary_platform} -- {followers} -- Score: {N}/100
   Niche: {niche}
   Best collab format: {format}

PLATFORM BREAKDOWN:
- LinkedIn: {count} candidates, avg score {N}
- Instagram: {count} candidates, avg score {N}
- YouTube: {count} candidates, avg score {N}
- Twitter: {count} candidates, avg score {N}

RECOMMENDED NEXT STEPS:
1. Reach out to TIER 1 targets this week
2. {quick_win recommendation}
3. {long_term recommendation}

Example Usage

Trigger phrases:

• "Find influencers in our niche"
• "Discover micro-influencers for partnership"
• "Who are the top creators in ecommerce?"
• "Find KOLs for Shopify and D2C"
• "Search for influencers we could collaborate with"
• "Find YouTube creators covering Shopify development"

User: Find micro-influencers in the Shopify and D2C space for co-marketing
Assistant: [reads config, searches LinkedIn/Instagram/YouTube/Twitter for Shopify and D2C creators with 1K-100K followers, scores by relevance/engagement/authenticity, generates partnership recommendations and outreach templates]

User: Find LinkedIn influencers in India covering ecommerce, 5K-50K followers
Assistant: [same flow limited to LinkedIn, filtered to India geography and 5K-50K range, deeper enrichment on LinkedIn-specific signals]