| name | seo-optimizer |
| description | SEO and AI-Search (GEO) specialist. Use for optimizing traditional search (Google/Bing) and AI
engines (ChatGPT, Perplexity, Google AI Overviews), schema markup for AI citation, Core Web
Vitals (LCP/INP/CLS), citation-worthy/zero-click content structure, E-E-A-T signals, and
adding privacy-aware web analytics (GA4, Plausible, Umami — cookie-consent/GDPR trade-offs).
|
SEO Optimizer
Lean router for search + AI-search optimization. The core model and the few
high-leverage rules are below; all deep detail (full schema JSON, per-platform
source lists, detailed checklists, Finnish-language SEO notes) lives in
references/ai-seo-patterns.md — read that file when implementing.
When to Use
- Optimizing for traditional search (Google, Bing) or AI search (ChatGPT,
Perplexity, Google AI Overviews, Claude)
- Implementing schema markup for rich results / AI citation
- Improving Core Web Vitals (LCP, INP, CLS)
- Structuring content to be cited by AI; zero-click / E-E-A-T work
Why GEO Matters (2025–2026)
- ~61% of searches start on AI platforms by 2026; budget ~50% of SEO effort to GEO
- 58–60% of Google searches are zero-click; CTR drops ~61% on queries with AI Overviews
- Proper schema → ~3.7x more likely to be cited by AI
Core Model: SEO vs GEO
| Factor | Traditional SEO | GEO |
|---|
| Goal | Rankings & clicks | Answer inclusion & citations |
| Primary metric | CTR, position | Citation frequency, inclusion rate |
| Backlinks | Critical | Weak signal for AI |
| Content structure | Helpful for users | +40% more citations |
| Freshness | Moderate | High |
| Keyword density | 1–2% | Less relevant |
| Schema markup | Rich snippets | Major citation boost |
GEO and SEO are complementary — structured, well-attributed, fresh content wins
both. Optimize for answer inclusion, not just position.
The High-Leverage Rules
Apply these first; they drive most of the citation/ranking gain. Largest
citation boosts (full table in the reference):
| Element | Citation boost |
|---|
| Original data tables | ~+410% |
| Complete schema markup | ~+370% |
| FAQ schema / Q&A format | ~+320% |
| Attributed expert quotes | ~+40% |
| Inline citations to authoritative sources | +30–40% |
- BLUF (Bottom Line Up Front) — answer the question in the first
paragraph, ≤50 words, then explain. AI extracts the lead.
- Schema markup — at minimum
Article (with author/sameAs/dates),
FAQPage, and Organization. Full copy-paste blocks in the reference.
- Original data tables — unique, sourced, dated tables are the single
most citation-worthy element.
- Freshness — set and keep
dateModified accurate; AI weights recency.
- Machine-readable E-E-A-T — author credentials in schema, first-person
experience, citations to/from authoritative sources, HTTPS + clear contact.
- Entity clarity — define each entity at first mention, name it
consistently, back it with schema.
Core Web Vitals (single source of truth)
| Metric | Good | Needs work | Poor | First levers |
|---|
| LCP | ≤2.5s | 2.5–4.0s | >4.0s | fetchpriority="high" + preload hero, CDN, WebP/AVIF |
| INP | ≤200ms | 200–500ms | >500ms | break long tasks (>50ms), defer JS, Web Workers, virtualize lists |
| CLS | ≤0.1 | 0.1–0.25 | >0.25 | explicit img/video dimensions, aspect-ratio, reserve embed space |
(INP replaced FID in March 2024 and measures the full interaction lifecycle.)
Detailed techniques + code in references/ai-seo-patterns.md → "Core Web Vitals".
Project Context: Raamattu Nyt
This is a Finnish-language Bible app. SEO/GEO work here is primarily
Finnish: Finnish keyword intent, Finnish Q&A phrasing, and Bible-reference
discoverability. See the "Finnish Language SEO Notes" section in
references/ai-seo-patterns.md before writing copy or schema. Prefer real,
app-specific entities (verses, topics, reading plans) over generic examples.
Checklist (pre-publish)
Content: BLUF answer first · primary keyword in title (<60 chars) + H1 ·
meta description 150–160 chars · FAQ section + schema · original data/table
where relevant · author attribution w/ credentials · 3–5 contextual internal
links · canonical set.
Technical: schema implemented & validated · images optimized + alt text ·
LCP ≤2.5s / INP ≤200ms / CLS ≤0.1 · mobile responsive · dateModified fresh.
GEO: direct answer in first paragraph · inline citations to authoritative
sources · entity defined at first mention · Article + FAQPage schema.
Full expanded checklists: references/ai-seo-patterns.md → "Implementation
Checklists".
Monitoring
- SEO: Google Search Console, PageSpeed Insights, Ahrefs/SEMrush
- GEO: citation velocity, answer-inclusion rate, brand-mention frequency,
source attribution (test representative queries in ChatGPT/Perplexity)
Analytics & Measurement
No web analytics is installed on Raamattu Nyt yet. There is no single "best"
tool — the choice hinges on the cookie-consent trade-off: GA4 sets cookies
→ needs a consent banner + Consent Mode v2 (EU/GDPR), while cookieless tools
(Plausible, Umami) need no banner, are ~1 KB (CWV-friendly), and fit RN's
EU/trust context. Add Google Search Console regardless (free, no cookies).
For the full decision matrix, GDPR trade-offs, and copy-paste integration (where
the snippet goes in index.html, the VITE_* ID in config.ts, and the SPA
route-change pageview hook), see
references/analytics-integration.md.
Reference
references/ai-seo-patterns.md — platform source preferences, full schema
code (Article/FAQPage/HowTo/SoftwareApplication), detailed citation-boost
stats, implementation + CWV checklists, content-formatting patterns, and
Finnish language SEO notes.references/analytics-integration.md — web analytics decision matrix
(GA4 vs Plausible vs Umami vs PostHog), GDPR/consent trade-offs, and
Raamattu Nyt integration how-to (SPA pageview tracking, env config, privacy).
