influencer-discovery

SKILL.md

Influencer Discovery

Given an industry/niche and target audience, systematically discover influencers across 4 platform categories plus cross-platform discovery angles (conference speakers, podcast guests, newsletter curators), then score and rank by Audience Overlap. Outputs a prioritized list of influencers worth partnering with, sponsoring, or learning from.

Usage

Use when finding influencers to partner with for sponsorships, affiliate deals, or co-marketing. Also useful for identifying thought leaders your target audience already follows, building a media list for outreach or PR, or understanding who shapes opinion in your niche before creating content.

Process

Step 1: Gather Inputs

Ask the user for:

  1. Industry/niche — the topic domain (e.g., "B2B SaaS marketing", "personal finance for millennials", "indie game development")
  2. Target audience — who follows these influencers, framed as the buyer (e.g., "SaaS founders", "freelance developers", "e-commerce store owners")
  3. Platform focus (optional) — prioritize specific platforms (YouTube, X/Twitter, Newsletters, Podcasts, Instagram). Default: search all.
  4. Minimum follower count (optional) — exclude influencers below a threshold. Default: no minimum (micro-influencers included).

Extract from inputs:

  • Niche keywords: The 3-5 topic terms that define the industry
  • Audience profile: Who the target audience is (role, stage, problem domain)
  • Platform priorities: Which platforms to weight more heavily

Step 2: Generate Search Queries

Generate 10 search queries to surface influencers across platform types and discovery angles:

"Top influencers" list queries:

  • "top [niche] influencers [current year]"
  • "best [niche] creators to follow"
  • "top [audience] thought leaders"

Platform-specific queries:

  • "top [niche] youtube channels"
  • "[niche] youtube [subscribers OR creators]"
  • "best [niche] newsletter substack"
  • "top [niche] podcast"
  • "[niche] twitter thought leaders"

Cross-platform discovery:

  • "[niche] conference speakers [current year]"
  • "who sponsors [niche] newsletters"
  • "[niche] podcast guest list"

Step 3: Search YouTube

Search YouTube for top channels in the niche:

  • Use queries: "top [niche] YouTube channels", "[audience] YouTube", "[niche keyword] tutorial/advice channel"
  • For each channel found, collect: Channel name, URL, subscriber count, content focus
  • Aim for 20-30 YouTube channels across macro (500k+), mid-tier (50k-500k), and micro (<50k) tiers

Step 4: Search X/Twitter

Search for thought leaders on X/Twitter:

  • Use queries: "[niche] twitter", "best [niche] accounts to follow", "[identity/role] twitter"
  • Check "listicles" — blog posts titled "Best [niche] accounts to follow on Twitter/X" often list 10-20 accounts at once
  • Include both prolific posters (daily content) and respected practitioners (less frequent but high-quality takes)

Step 5: Search Blogs, Newsletters, and Podcasts

Blogs

  • Search: "top [niche] blogs", "best [niche] blog to read", "[niche] writers"
  • Look for independent writers with their own sites

Newsletters

  • Search: "best [niche] newsletter", "[niche] substack", "[niche] email newsletter"
  • Check Substack's discover page for the niche if applicable

Podcasts

  • Search: "top [niche] podcast", "best [niche] podcast", "[audience] podcast"
  • Check Apple Podcasts and Spotify charts if accessible

Step 6: Search Meta (Facebook and Instagram)

  • Search: "[niche] facebook page", "[identity] instagram", "[niche] instagram creator"
  • Focus on public profiles and pages
  • Meta is lower priority for most B2B niches — if the niche skews consumer or visual, prioritize more heavily

Step 7: Mine Cross-Platform Discovery Sources

Conference Speaker Lists

  • Search: "[niche] conference [current year] speakers", "[niche] summit speakers"
  • Conference keynote and session speakers are pre-vetted as recognized voices
  • Extract speaker names → find their primary platform

Podcast Guest Cross-References

  • Search: "[top podcast in niche] guests", "most popular [niche] podcast episodes"
  • Repeat guests on multiple podcasts = high-credibility signal

Newsletter Curator Lists

  • Search: "who sponsors [niche] newsletters", "[niche] newsletter sponsors"
  • Companies sponsoring newsletters indicate which newsletters have a commercially valuable audience

Step 8: Normalize All Results

Compile all discovered influencers into a single list. For each entry, record:

Field Description
Name Person or brand name
Primary Platform The platform where they have the largest/most active presence
X/Twitter URL Profile URL or "—"
YouTube URL Channel URL or "—"
Newsletter/Blog URL URL or "—"
Podcast URL URL or "—"
Instagram/Facebook URL URL or "—"
Primary Follower Count Count on their primary platform
Content Type What they primarily create (tutorials, opinion/takes, case studies, interviews, tools)
Source Where they were discovered

Deduplication: If the same person appears from multiple discovery sources, merge into one entry. Multiple discovery sources = stronger relevance signal.

Step 9: Score Each Influencer

Score every influencer on Relevance (1–5):

Score Signal
5 Content is directly about the target audience's core problems — almost every post/video is on-target
4 Strong match — most content is relevant, occasional tangents
3 Adjacent — their niche overlaps meaningfully but isn't a perfect match
2 Loose match — some relevant content but audience is broader or different
1 Tangential — topic surface-level overlap but different audience

Audience Overlap Rating

Rating Criteria
High Relevance ≥ 4
Medium Relevance = 3
Low Relevance ≤ 2

Step 10: Sort and Finalize

Sort the full list:

  1. Primary: Audience Overlap rating (High → Medium → Low)
  2. Secondary: Primary follower count (largest first within each tier)

Include micro-influencers (1K–50K followers) alongside macro — do not exclude them.

Output Format

# Influencer Discovery: [Industry/Niche]

**Date:** [current date]
**Industry/Niche:** [description]
**Target Audience:** [description]
**Influencers found:** [count] across [X] platforms
**Audience Overlap breakdown:** High: [X] | Medium: [X] | Low: [X]

---

## High Overlap Influencers

| # | Name | Primary Platform | Followers | X/Twitter | YouTube | Newsletter/Blog | Podcast | Instagram/FB | Content Type | Discovery Source |
|---|------|-----------------|-----------|-----------|---------|-----------------|---------|-------------|-------------|-----------------|
| 1 | [name] | [platform] | [count] | [url or —] | [url or —] | [url or —] | [url or —] | [url or —] | [type] | [source] |

---

## Medium Overlap Influencers
[Same table structure]

---

## Low Overlap Influencers
[Same table structure]

---

## Platform Coverage Summary

| Platform | Count | High Overlap | Notes |
|----------|-------|-------------|-------|
| X/Twitter | X | X | [observation] |
| YouTube | X | X | |
| Newsletter/Blog | X | X | |
| Podcast | X | X | |
| Instagram/Facebook | X | X | |

---

## Observations
[2-3 bullet points: where this niche's influencer ecosystem is concentrated, tier distribution, notable patterns]

## Partnership Opportunities
[Top 3-5 influencers worth prioritizing for sponsorship or collaboration, with a 1-sentence reason each]

## Recommended Next Steps
1. [e.g., "Reach out to top 3 High Overlap newsletter authors — newsletter sponsorships often have fastest lead time"]
2. [e.g., "Monitor [top X/Twitter account] for mentions of problems your product solves"]

Rules

  • Aim for 100+ influencers total. For narrow B2B niches, 50 is acceptable — flag it.
  • Never invent influencer accounts or inflate follower counts — only include verified results from search.
  • Never exclude micro-influencers (1K–50K) — they are often the highest-value partnerships for niche B2B audiences.
  • Never focus only on one platform — cross-platform coverage is a core output requirement.
  • Discovery source matters: an influencer found via conference speakers + podcast guest + YouTube search is more credible than one found via a single listicle.
  • Follower counts are best-effort estimates — verify before outreach.
  • If fewer than 30 influencers are found, try broader synonyms for the niche or audience.
Weekly Installs
7
GitHub Stars
11
First Seen
14 days ago
Installed on
opencode6
gemini-cli6
github-copilot6
codex6
amp6
cline6