SkillSmith AEO/GEO Citation Booste
Search is moving from blue links to AI answers. When your buyers ask ChatGPT, Perplexity, or Google AI a question, those engines quote a handful of sources — you're either one of them, or you're invisible.
What it does
Most AEO/GEO tools are monitoring dashboards: they tell you your "AI visibility score" and who's cited, then leave the fixing to you. This Skill is the opposite — an executable workflow that produces the work you actually ship.
Give it one page (a URL to fetch or pasted content) plus the real questions your audience asks AI, and it returns five ready-to-ship sections delivering six work products: a content diagnosis, an answer-first AEO rewrite, Schema.org guidance, a white-hat community citation plan, a competitor citation-gap matrix, and a 30-day AI-visibility plan. Everything runs on content you provide — no paid SEO databases. Any number that isn't freely obtainable (e.g. exact search volume) is honestly flagged, never fabricated.
How it works
A six-step workflow run in order, each grounded in current GEO research:
| Step | Output |
|---|---|
| 1. Parse & profile | Extract content; identify the primary entity, claims, structure, and what's missing for citability |
| 2. Map questions → answers | A question→answer coverage table; gaps flagged where no self-contained answer exists |
| 3. Rewrite for AEO | Answer-first rewrite, entity clarity, citable definition/statistic sentences, FAQ, Schema.org recommendations |
| 4. Community plan | Specific subreddits / Quora threads / listicles + white-hat, platform-compliant participation scripts |
| 5. Citation-gap analysis | Factor-by-factor gap matrix vs. competitors + a manual prompt-test protocol |
| 6. 30-day plan | Week-by-week tasks + a free, repeatable visibility-measurement protocol with KPIs |
⭐ Featured example
Input: A SaaS homepage for "InboxZero AI" (an AI email assistant, $12/seat/month) plus three target questions buyers ask AI — e.g. "What's the best AI email assistant for small teams?"
Output: a complete deliverable. The diagnosis flags the tagline "Reclaim your inbox with AI." as zero-citability (metaphor, no entity, no definition), then ships a paste-ready answer-first rewrite plus citable definition and statistic sentences:
![Output preview: answer-first AEO rewrite and citable-sentence skeletons for InboxZero AI]
It then turns abstract "AI visibility" into a checkable scorecard — a competitor citation-gap matrix scored 0–2 per factor, with the biggest gaps to close first, feeding into a week-by-week 30-day plan:
![Output preview: competitor citation-gap matrix scored 0-2 with a 30-day plan]
Data sentences are delivered as [INSERT YOUR STAT] skeletons — the workflow never invents a number you haven't measured.
Why it's different
- Executable, not a dashboard. Mainstream AEO tools (Profound ~$499/mo, Peec ~€89/mo, Otterly ~$29/mo) are monitor-only subscriptions that tell you your score and stop. This delivers paste-ready answer-first content, Reddit/UGC engagement scripts, schema recommendations, and a 30-day task list. They sell monitoring; this sells finished work.
- Built-in community citation strategy. LLM answers lean heavily on UGC (Reddit, G2, Quora), yet almost no tool ships compliant, white-hat, thread-by-thread engagement scripts and listicle-outreach targets. This does.
- Citation-gap analysis you can act on. A checkable factor × competitor matrix plus a prompt-testing protocol you run yourself — abstract visibility becomes a prioritized to-do list.
- Honest guardrails. No fabricated citation data, assumptions clearly labeled, and a clean split between what the Skill auto-generates and what you measure by hand.
Disclaimer / Scope
This Skill is a content-production and strategy engine, not a real-time monitoring bot. The AEO rewrites, community playbook, schema guidance, gap matrix, and 30-day plan are fully automated. The citation-measurement step — running your question set through ChatGPT / Perplexity / Google AI / Gemini and logging who each engine cites — must be performed by you: sandbox web fetches do not equal those engines' real answer outputs, and none expose a free, automatable official interface. The Skill ships a follow-the-steps testing protocol, a logging template, and the KPI math to make that manual pass as light as possible; it does not, and does not claim to, run every engine for you. No one controls AI output — these tactics raise the probability of being cited, they do not guarantee it.
