SEO Opportunity Discovery

Data Sources

Discovery Workflows

1. GSC-Driven Discovery (Weekly)

# Collect fresh data
cd ~/skillboss && npx tsx scripts/seo/collect-gsc-data.ts

Read 28-day data
cat reports/seo/data/gsc-28d-latest.json

Find opportunities:

• High impression, low CTR → Title/description optimization candidates
• Position 4-15, high impressions → Striking distance (content refresh to break top 3)
• New queries appearing → Emerging opportunities to capture
• Declining queries → Content freshness needed

2. Competitor Page Gap Analysis

For each competitor (OpenRouter, LiteLLM, Portkey, Together AI):

Crawl their /alternatives/, /compare/, /blog/, /use/ pages
List pages they have that we don't
Cross-reference with search volume (Ahrefs or DataForSEO)
Prioritize by: search volume × business relevance × creation difficulty

3. Fan-Out Query Mapping (Ethan Smith Framework)

AI breaks complex questions into sub-queries. Map the fan-out:

User asks: "What's the best AI API gateway for Claude Code?"
AI searches:
  ├── "best AI API gateway 2026"        → /blog/best-ai-api-gateway-2026 ✅
  ├── "Claude Code API integration"      → /for/claude-code ✅
  ├── "AI API pricing comparison"        → /pricing ✅
  └── "OpenRouter vs SkillBoss"          → /compare/skillboss-vs-openrouter ❌ MISSING

For each target query, map all sub-queries and ensure we have pages for each.

4. Trend Detection

# Check emerging tools via Google Trends
Look for: new AI tools, framework launches, pricing changes

Signals to watch:

• New AI tool launches (Kiro, Antigravity, etc.) → Create alternatives page within 48 hours
• Major pricing changes (OpenAI, Anthropic) → Update pricing comparison pages
• AI model releases (GPT-5.x, Claude 4.x) → Update /use/ pages + blog posts

5. Competitor Paid Search Mining (Ethan Smith Playbook)

What competitors bid on in Google Ads = high-intent queries worth targeting organically.

Use Ahrefs "Paid Keywords" or DataForSEO Ads data
Filter for queries with commercial intent
Cross-reference: do we have a page for this query?
No → add to creation backlog

Output Format

## SEO Opportunity Report — [Date]

Tier 1: Quick Wins (this week)
Keyword
Volume
Current Position
Action
Page
...
...
...
CTR fix / new page / content refresh
...
Tier 2: High Value (this month)
| Keyword | Volume | Difficulty | Action | Estimated Impact |
|---------|--------|-----------|--------|-----------------|

Tier 3: Strategic (this quarter)
| Keyword | Volume | Competitor Rank | Our Gap | Action |
|---------|--------|----------------|---------|--------|

Pages to Create

[priority] /alternatives/[slug] — targeting "[keyword]"
[priority] /compare/[x]-vs-[y] — targeting "[keyword]"
...

6. Sales/Support → Long-Tail Content (Carta Framework, AEO Conf 2026)

AI prompts are ~60 words avg (vs Google ~4 words).
60% of ChatGPT prompts are 10+ words.
The long-tail content mine is in your OWN customer conversations.

Step 1: Ingest support/Discord/sales conversations
Step 2: Extract: Objections, ROI questions, Integration concerns, Risk narratives
Step 3: Translate to search intercept: "How do I reduce API costs?" → "how to reduce ai api costs 2026"
Step 4: Build content with: Direct answer blocks, Fact anchoring, Citation strength
Step 5: Compound: Conversations → Insight → Content → AI citations → New demand

Sources for long-tail discovery:

• Communities (Discord, Reddit, HN)
• Support tickets / help desk
• User interviews / feedback
• Sales call transcripts (if available)

Decision Rules

• Search volume > 100/month + no existing page → Create page
• Position 4-10 + impressions > 200/month → Content refresh + internal linking
• Position 1-3 + CTR below benchmark → Title/description optimization
• Competitor has page, we don't + volume > 50/month → Create page
• New tool trending on Reddit/HN → Create alternatives page within 48 hours