AI 提示詞： SEO 優化

I want to rank for topics related to [main topic/niche].

My website: [URL or description]
Current authority level: [new site / some rankings / established]
Target audience: [who searches for this]
Business model: [how you make money from traffic]
Top competitor URLs: [paste 3-5 competitor pages ranking for your target terms]

Build keyword clusters:
1. 5 pillar topic clusters with 8-10 supporting keywords each
2. For each keyword: search intent (informational / commercial / transactional / navigational)
3. AI Overview risk assessment: flag keywords where Google's AI already gives a direct answer — label each as Critical / High / Moderate / Low risk
4. Difficulty tier: quick wins (low competition) vs. long-term authority plays
5. Content format recommendation for each keyword (blog post, comparison guide, landing page, interactive tool, data study)
6. Internal linking map showing how clusters connect to each other
7. 5 'zero competition' long-tail keywords most sites overlook
8. For Critical/High AI-risk keywords: suggest a defensible angle (original data, first-hand experience, proprietary framework) that AI Overviews can't replicate

Score this page and tell me exactly what to fix.

Target keyword: [primary keyword]
Secondary keywords: [2-3 related terms]
My page URL: [URL]
My page content: [paste full text, or first 1000 words for long pages]
Top-ranking competitor: [paste URL or content of #1 result]

Score the page on 4 dimensions (each out of 25, total out of 100):

1. KEYWORD OPTIMIZATION (title tag, meta description, H1, H2s, keyword placement, density, semantic variations)
2. CONTENT QUALITY (depth vs. competitors, unique value-add, E-E-A-T signals, readability)
3. TECHNICAL SEO (heading hierarchy, internal links, schema opportunity, canonical, mobile-readiness)
4. USER EXPERIENCE (above-the-fold value, scannability, CTA clarity, content structure)

For each dimension:
- Current score with reasoning
- Specific fixes ranked by impact (high/medium/low)
- Before → after examples for title tag, meta description, and H1

Classify the overall rewrite level needed:
- Light touch (80-100): minor tweaks
- Moderate (65-79): restructure + add sections
- Major (40-64): significant rewrite keeping best parts
- Complete (<40): start fresh with new angle

Run a technical SEO audit for my site.

Site type: [blog / ecommerce / SaaS / local business / publisher]
CMS/platform: [WordPress / Shopify / Next.js / custom / etc.]
Site size: [number of pages]
PageSpeed scores: [paste LCP, INP, CLS numbers if available]
Known issues: [anything you've already identified]
Monthly organic traffic: [if known]

Audit these 8 categories and flag issues as Critical / High / Medium / Low:

1. CRAWLABILITY — robots.txt, XML sitemap, crawl budget, orphan pages, redirect chains
2. INDEXABILITY — canonical tags, meta robots, duplicate content, thin pages, index bloat
3. CORE WEB VITALS — LCP optimization (hero images, server response), INP fixes (JavaScript execution, event handlers, hydration), CLS prevention (image dimensions, font loading, dynamic content)
4. SECURITY — HTTPS, mixed content, security headers (HSTS, CSP)
5. URL STRUCTURE — descriptive slugs, flat hierarchy, parameter handling
6. MOBILE — responsive design, tap targets, viewport configuration
7. STRUCTURED DATA — current schema audit, missing opportunities (avoid HowTo and FAQ unless gov/health)
8. JAVASCRIPT RENDERING — render-blocking resources, async/defer usage, hydration issues

For each issue found:
- What's wrong (with specific evidence)
- How to fix it (platform-specific instructions)
- Expected impact on rankings
- Implementation difficulty (easy / moderate / hard)

End with a 15-minute monthly maintenance checklist I can follow.

Design a topic cluster to build topical authority for: [main topic]

My site: [URL or description]
Business goal: [what conversion does this cluster drive?]
Current content: [list any existing articles on this topic]
Target audience: [who and what stage of awareness]

Deliver a complete cluster architecture:

1. PILLAR PAGE
   - Target keyword and search intent
   - Recommended format (ultimate guide / hub page / comparison matrix)
   - Suggested word count (3000-5000)
   - H2 outline covering the topic comprehensively
   - Internal linking strategy (how supporting articles feed into this)

2. 10 SUPPORTING ARTICLES — for each one:
   - Target keyword + monthly search volume estimate
   - Search intent (informational / commercial / comparison)
   - Content format (how-to / listicle / case study / data study / comparison)
   - Word count target
   - How it links to the pillar (anchor text suggestion)
   - How it cross-links to other supporting articles
   - AEO defensibility score (can AI Overviews replace this? If yes, what angle makes it defensible?)

3. LINKING MAP — visual text diagram showing how all pieces connect

4. 3-PHASE ROLLOUT PLAN
   - Phase 1 (Foundation): pillar + 3 highest-impact articles
   - Phase 2 (Authority): next 4 articles targeting commercial intent
   - Phase 3 (Coverage): remaining articles filling topical gaps
   - Publishing cadence recommendation

I have an existing page that needs improvement.

Page URL: [URL]
Page content: [paste full text]
Target keyword: [primary keyword]
Current performance: [paste Search Console data if available — position, impressions, clicks, CTR]
Date originally published: [when]
Last updated: [when]

Do a full content refresh analysis:

1. CURRENT SCORE (out of 100) across:
   - Keyword optimization (are terms placed correctly?)
   - Content depth (vs. what's ranking above you now)
   - Freshness (outdated stats, dead links, old references)
   - E-E-A-T signals (author credibility, first-hand experience, citations)
   - Structural SEO (headings, internal links, schema)

2. COMPETITIVE GAP — analyze what the top 3 ranking pages cover that mine doesn't

3. REFRESH PLAN — prioritized list of specific changes:
   - Sections to add (with draft outlines)
   - Sections to cut or merge
   - Stats/examples to update
   - Internal links to add (suggest anchor text)
   - Schema markup to implement
   - Title tag and meta description rewrites (3 options each)

4. AEO HARDENING — make this content resistant to AI-generated answers:
   - Add original data points or first-person experience AI can't fabricate
   - Structure key passages for featured snippet extraction
   - Identify claims that need unique proof or case studies

5. BEFORE vs. AFTER predicted score with reasoning

I want to build programmatic SEO pages at scale.

Business: [what you do]
Data sources available: [databases, APIs, spreadsheets, scraped data]
Target audience: [who needs these pages]
Example query pattern: [e.g. 'best [tool] for [use case]' or '[service] in [city]']
Estimated page count: [hundreds / thousands / tens of thousands]

Design the programmatic SEO system:

1. TAXONOMY DESIGN
   - Primary dimension (e.g., cities, tools, industries)
   - Secondary dimensions for cross-referencing
   - URL pattern template (e.g., /tools/{category}/{tool-name})
   - Title and H1 templates with dynamic variables

2. PAGE TYPE ARCHITECTURE — for each page type:
   - JSON schema defining all required data fields
   - Which fields are unique per page vs. templated
   - Minimum content quality threshold (word count, unique data points)
   - Example of a fully populated page

3. THIN CONTENT SAFEGUARDS
   - Minimum viable data per page (below this → noindex or don't generate)
   - How to inject genuinely unique value into each page
   - Content mix recommendation: resource pages (checklists, ideas) > interactive tools > comparisons
   - Internal linking automation rules

4. INDEX BLOAT PREVENTION
   - Which pages to index vs. noindex
   - Canonical strategy for near-duplicate pages
   - Pagination and faceted navigation handling
   - Sitemap segmentation strategy

5. USEFUL PAGE TEST — for 3 sample pages from my data, show me exactly what each page would contain and honestly assess: is this genuinely useful?

I want my brand to appear in AI-generated search answers.

Brand/site: [name and URL]
Core topics: [what you want to be cited for]
Current AI visibility: [do ChatGPT/Perplexity/AI Overviews ever mention you? Which competitors appear instead?]
Content assets: [your best existing content]
Budget level: [time only / small budget / significant investment]

Build a Generative Engine Optimization (GEO) strategy:

1. QUERY INVENTORY
   - 20 high-value queries where your brand should appear in AI answers
   - Current citation audit: who gets mentioned now for each query?
   - Gap analysis: which queries are winnable vs. dominated?

2. REDDIT STRATEGY (highest ROI for AI citations)
   - Subreddits where your audience asks questions
   - 5 ask-and-answer post concepts (branded account, genuine value)
   - 5 existing threads to answer with domain expertise
   - How to build credibility without being flagged as spam

3. YOUTUBE SIGNAL STRATEGY
   - 3 tutorial/demo video concepts AI models reference
   - Niche creator partnership targets (micro-influencers > celebrities for citation)
   - How to structure video titles and descriptions for AI crawlers

4. CONTENT SIGNALS FOR LLM CITATION
   - Structure passages as direct, quotable answers (not buried in paragraphs)
   - Add original data, named frameworks, and specific numbers AI can cite
   - Implement comparison tables LLMs pull from
   - FAQ sections with concise, authoritative answers
   - Author bylines with verifiable credentials

5. AI CRAWLER ACCESS
   - robots.txt rules for GPTBot, ClaudeBot, PerplexityBot
   - llms.txt file implementation
   - Structured data that helps AI understand your content

6. MEASUREMENT PLAN
   - Monthly citation tracking routine (15 minutes)
   - Which tools to use for monitoring AI mentions
   - Success metrics and timeline expectations

Build a complete local SEO domination plan for my business.

Business: [name, type, e.g. 'Ace Plumbing, residential plumber']
Location: [city/region, service area radius]
Google Business Profile status: [claimed / optimized / not yet set up]
Review count: [number and average rating]
Top local competitor: [name if known]
Current website: [URL if exists]

Deliver a 7-part local SEO plan:

1. GOOGLE BUSINESS PROFILE OPTIMIZATION
   - Every field to fill in (with example text for my business type)
   - Category selection (primary + secondary)
   - Service area configuration
   - Products/services to add
   - Photo strategy (types, quantity, naming convention)

2. LOCAL SCHEMA MARKUP — ready-to-use JSON-LD for:
   - LocalBusiness (with correct subtype for my industry)
   - AggregateRating
   - AreaServed
   - OpeningHoursSpecification
   - GeoCoordinates

3. HIGH-INTENT KEYWORD TARGETS
   - 20 'ready to buy NOW' keywords (emergency, near me, pricing, same-day modifiers)
   - Keyword map: which page targets which keyword

4. GBP POST CALENDAR — 12 posts covering:
   - Service spotlights with seasonal hooks
   - Before/after showcases
   - Customer FAQ answers
   - Limited-time offers
   - Review highlights

5. REVIEW GENERATION STRATEGY
   - When and how to ask (without being pushy)
   - SMS/email templates for review requests
   - Response templates for positive and negative reviews

6. LOCAL CITATION SOURCES — 10 directories specific to my industry

7. 30-DAY ACTION PLAN with weekly milestones (max 30 min/day)