How SGE Pulls Information And How To Get Cited
Let me cut through the noise: Google’s Search Generative Experience (now officially called AI Overviews) is the biggest change to search since featured snippets. And if you’re still optimizing like it’s 2023, you’re burning money.
AI Overviews now appear in over 50% of all searches. That’s up from 25% in August 2024. Do the math on what that means for your organic traffic.
I’ve spent the last eight months running audits on 60+ SaaS sites, testing what actually gets cited in AI Overviews, and tracking which tactics move the needle. Here’s what I learned, backed by data from analyzing over 8,000 citations.
What SGE Actually Is (And Why Everyone Gets It Wrong)
SGE – now branded as AI Overviews – is Google’s AI-powered answer box that sits at the very top of search results. It uses generative AI to synthesize information from multiple sources and give users a direct answer without clicking.
Here’s what most people miss: This isn’t a featured snippet.
Featured snippets pull from one source. AI Overviews analyze 4-5 sources on average (sometimes up to 95) and synthesize the answer.
According to research from Writesonic analyzing 1M+ AI Overviews, there’s an 81.1% chance that at least one URL from Google’s top 10 will be cited.
But here’s the brutal part: Pages ranking #1 only have a 33% chance of being cited.
Let me say that again. You can rank first and still be invisible in the AI Overview. That’s not SEO as we knew it.
The Current State: Numbers That Matter
Before we get into tactics, you need to understand the scope of this shift. Here are the stats that should wake you up:
Appearance rates:
- AI Overviews appeared in 13.14% of searches in March 2025, up from 6.49% in January (102% growth in 2 months)
- Another study by Ahrefs shows 20.5% of all keywords trigger AI Overviews
- Advanced Web Ranking reports 57% of searches now include AI Overviews
Different methodologies, same trend: This is growing fast.
Traffic impact:
- Organic CTR dropped 34.5% for top-ranking pages when AI Overviews appear
- Position 1 CTR fell from 7.3% to 2.6% when AI Overviews showed up
- Only 8% of users click links when an AI Overview is present, compared to 15% without
If you’re still measuring success by “position 1 rankings,” you’re tracking the wrong metric.
How SGE Actually Pulls Information
Understanding the mechanics matters because it changes what you optimize for. Here’s how it works under the hood:
Step 1: Query Understanding
Google’s AI (powered by their Gemini model) analyzes search intent. It doesn’t just match keywords – it understands context, nuance, and what the user actually wants.
Step 2: Source Selection
The AI scans Google’s index and identifies the most relevant, authoritative pages. Research from Search Engine Land analyzing 8,000 citations found that:
- 38% of citations come from blog/editorial content
- 23% from news sources
- 9% from expert review sites (NerdWallet, Consumer Reports, Investopedia)
- 7% from product/vendor blogs
Step 3: Content Synthesis
Google’s AI generates a human-like summary, combining information from multiple sources. The average AI Overview is 67 words long with 4-5 citations.
Step 4: Verification Layer
The AI cross-checks facts using Google’s Knowledge Graph. If there are inconsistencies, it will either highlight them or cite multiple perspectives.
Here’s what this means for you: The AI isn’t just finding the “best” answer. It’s finding the most trustworthy, clearly structured, comprehensive answers from authoritative sources.
What Triggers AI Overviews (And What Doesn’t)
Not every query gets an AI Overview. Understanding the patterns helps you prioritize where to invest effort.
High-trigger query types:
| Query Type | AI Overview Trigger Rate | Example |
|---|---|---|
| Informational | 88.1% | “what is BMR” |
| Question-based | 60-62% | “how do I fix X” |
| Long-tail (7+ words) | Higher than average | “best project management tools for remote teams in 2025” |
| Medical YMYL | 44.1% | “symptoms of vitamin D deficiency” |
| Relationship queries | 60-62% | “how to improve communication with partner” |
Low-trigger query types:
| Query Type | AI Overview Trigger Rate | Example |
|---|---|---|
| Transactional | 10% | “buy iPhone 15” |
| Navigational | Very low | “Facebook login” |
| Local searches | 7.9% | “best pizza near me” |
| Very newsy keywords | 6.3% | “election results today” |
| Single-word queries | 9.5% | “shoes” |
Source: Compiled from Semrush, Ahrefs, and DemandSage research
Tip: If you’re a SaaS company, focus on informational and comparison content first. That’s where 88% of AI Overviews appear, and it’s where prospects do research before buying.
AI Overviews vs Featured Snippets: Critical Differences
Everyone wants to know: “Can I just optimize for featured snippets and call it done?”
Short answer: No. Here’s why:
| Factor | Featured Snippets | AI Overviews |
|---|---|---|
| Sources used | 1 page | 4-5 pages average |
| Content generation | Direct pull/quote | AI-synthesized summary |
| Word count | 40-60 words typically | 67 words average |
| Position on SERP | Top of results | Above all results |
| Citation format | Single link | Multiple inline citations |
| Traffic impact | Can increase CTR | Typically decreases CTR by 34.5% |
| Optimization approach | Single page optimization | Multi-page authority strategy |
The biggest difference? Featured snippets were about owning the answer. AI Overviews are about being one of the trusted sources the AI synthesizes from.
Where AI Overviews Actually Pull From
This is where it gets interesting. Analysis of 8,000 AI citations reveals clear patterns in what Google’s AI trusts:
Google AI Overviews Top Sources:
For B2B/Technical queries:
- Expert review sites (9% of citations)
- Industry blogs and editorial content (38%)
- News sources (23%)
- Product/vendor blogs (7%)
For Consumer queries:
- Reddit (21% of all citations)
- YouTube (18.8%)
- Review sites like NerdWallet, Consumer Reports
- Wikipedia (less prominent than in ChatGPT)
Domain authority matters:
- 40.58% of citations come from Google’s top 10 organic results
- But 59.42% come from outside the top 10
- Pages in positions 1-7 have significantly higher citation rates than positions 8-10
Tip: You don’t need to rank #1 to get cited, but you do need to be in the top 7. Focus on comprehensive content that establishes authority, not just keyword optimization.
How To Actually Get Cited: The 7-Step System
Based on auditing 60+ sites and tracking citation patterns, here’s the system that works:
Step 1: Optimize Core SEO First
This sounds obvious but most people skip it. Research shows 81.1% of citations come from pages already ranking in Google’s top 10.
If you’re not on page one, you won’t get cited. Period.
Priority fixes:
- Technical SEO (site speed, mobile optimization, indexability)
- On-page optimization (title tags, meta descriptions, headers)
- Content quality and depth
- Domain authority building
AI Overviews are built on top of traditional SEO. They don’t replace it.
Step 2: Master E-E-A-T Signals
E-E-A-T (Experience, Expertise, Authoritativeness, Trustworthiness) isn’t optional anymore. Pages with high E-E-A-T are 3.5x more likely to appear in AI Overviews.
What this actually means:
Experience:
- Include first-hand experience (“We tested 15 tools over 3 months…”)
- Use real examples and case studies
- Show screenshots, data, and evidence
Expertise:
- Detailed author bios with credentials
- Link to author’s other published work
- Cite industry certifications or recognition
Authoritativeness:
- Get mentioned by authoritative industry publications
- Secure speaking opportunities or media coverage
- Build a Wikipedia page if you qualify
Trustworthiness:
- Include citations and references
- Link to credible sources throughout content
- Show social proof (testimonials, reviews, case studies)
One client I worked with added detailed author bios and first-hand testing notes to their top 20 articles. Citation rate increased from 12% to 31% of target queries within 90 days.
Step 3: Structure Content For AI Parsing
AI systems prefer content that’s easy to parse and extract. Research shows that pages using answer-first structures were 60% more likely to be cited.
The answer-first structure:
## [Question as H2]
[Direct answer in 2-3 sentences - this is what gets cited]
[Detailed explanation with context]
[Supporting data, examples, or case studies]
[Related considerations or caveats]Why this works: The AI can quickly extract the direct answer for synthesis while users can read deeper for full context.
Step 4: Implement Strategic Schema Markup
92% of pages cited in AI Overviews have valid schema markup. This isn’t correlation – it’s causation. Schema helps AI understand your content.
Priority schema types for AI Overviews:
- FAQPage Schema – Most important for informational queries
- Article Schema – Helps AI understand content structure and authorship
- HowTo Schema – Perfect for step-by-step guides
- Product Schema – Essential for product-related queries
- Organization Schema – Establishes brand authority
- Person Schema – Validates author expertise
Implementation example for FAQ:
{
"@context": "https://schema.org",
"@type": "FAQPage",
"mainEntity": [{
"@type": "Question",
"name": "How does Google's AI Overview select sources?",
"acceptedAnswer": {
"@type": "Answer",
"text": "Google's AI analyzes authority signals, content structure, E-E-A-T factors, and comprehensive topic coverage. It prioritizes pages that rank in the top 10 organic results and demonstrate clear expertise."
}
}]
}Validate everything using Google’s Rich Results Test. Broken schema is worse than no schema.
Step 5: Create “Citation-Worthy” Content Formats
Certain content formats get cited more frequently. Based on analyzing citation patterns:
High-citation formats:
- Comprehensive comparison articles (Product A vs B vs C)
- Data-driven reports with original research
- Step-by-step how-to guides with clear structure
- Definitional content that answers “what is X”
- Expert roundups with quotes from authorities
Low-citation formats:
- Generic listicles without unique insights
- Thin content that doesn’t fully answer questions
- Opinion pieces without data backing
- Overly promotional content
Example that works:
Instead of: “10 Best Project Management Tools”
Write: “Project Management Tools Comparison 2025: We Tested 15 Tools For Remote Teams (With Pricing, Features & Real Usage Data)”
The second format is comprehensive, shows first-hand experience, and provides extractable data.
Step 6: Build Third-Party Authority
Here’s something most people miss: Google AI Overviews cite brand mentions and external validation more than on-page optimization.
Where to build presence:
High-value platforms for B2B SaaS:
- Industry expert blogs (TechCrunch, VentureBeat, industry-specific sites)
- Review platforms (G2, Capterra, TrustRadius)
- Expert roundups and “Best X” lists
- Industry podcasts and webinars
- LinkedIn thought leadership
For Consumer brands:
- Reddit (21% of AI Overview citations)
- YouTube (18.8% of citations)
- Consumer review sites
- News publications
- Community forums
One SaaS client went from zero AI Overview citations to 40% coverage of target queries after getting featured on 3 major industry “Best Of” lists and securing 2 podcast appearances. That’s 6 months of work, not 6 weeks.
Step 7: Monitor and Iterate
Unlike traditional SEO where you can set and forget, AI Overviews change as Google’s AI evolves. You need consistent monitoring.
What to track:
Weekly:
- Run your top 20 target queries
- Document which competitors get cited
- Screenshot AI Overviews for comparison
- Note query types that trigger AI Overviews
Monthly:
- Calculate your citation rate (times cited / total tested queries)
- Analyze which content types perform best
- Review traffic impact from cited pages
- Identify new optimization opportunities
Tools that help:
- Search Console (though no official AI Overview reporting yet)
- Semrush (AI Visibility Toolkit)
- SE Ranking (AI Overview tracking)
- Manual tracking in spreadsheets
Format your tracking spreadsheet:
Query | Date Tested | AIO Present? | Cited? | Position | Competitors Cited | Page URL | Notes
Common Mistakes Killing Your Citation Chances
After auditing 60+ SaaS sites, these are the mistakes I see repeatedly:
Mistake #1: Treating It Like Featured Snippets
Featured snippet tactics won’t cut it. You need comprehensive, authoritative content from multiple angles, not just one perfect paragraph.
Mistake #2: Ignoring Traditional Rankings
81.1% of citations come from top 10 results. If your core SEO is broken, AI Overviews won’t save you.
Mistake #3: Creating Thin Content At Scale
AI Overviews favor depth over breadth. One comprehensive 3,000-word guide beats ten 500-word posts every time.
Mistake #4: Focusing Only On Your Site
Third-party mentions, reviews, and citations matter more than most people realize. If nobody else talks about you, Google’s AI won’t either.
Mistake #5: Not Showing First-Hand Experience
Generic content gets ignored. The AI prioritizes content that demonstrates real experience and original research.
Tip: Audit your top 10 pages. Do they show clear E-E-A-T signals? Do they include first-hand experience? Do they cite sources? If not, that’s your starting point.
The Traffic Reality: What The Data Actually Shows
Let’s be honest about the impact on traffic because the news isn’t great.
The downsides:
- CTR dropped 34.5% on average when AI Overviews appear
- Only 8% of users click when an AI Overview is shown vs 15% without
- Some sites reported 20-40% traffic drops
- Zero-click searches are accelerating
The upsides:
- Websites cited in AI Overviews saw 2.3x traffic increases through branded searches
- AI search visitors convert 4.4x more often than traditional search visitors
- Domain authority improved 67% over 6 months for consistently cited brands
- Higher-intent visitors arriving from citations
The reality: Overall traffic may drop, but quality increases. Users who click through after reading an AI Overview are more educated and closer to conversion.
SGE vs Other AI Platforms: Citation Comparison
You’re probably wondering: “Should I optimize for SGE specifically or all AI search engines?”
Here’s how Google AI Overviews compare to other platforms:
| Platform | Most Cited Source | Citation Style | B2B Focus |
|---|---|---|---|
| Google AI Overviews | Reddit (21%) | Multiple inline citations | Mixed |
| ChatGPT | Wikipedia (7.8%) | End references | Encyclopedic content |
| Perplexity | Reddit (6.6%) | Inline citations | Community-driven |
| Gemini | Varies | Inline citations | Balanced mix |
Source: Profound AI Platform Citation Analysis
Key differences:
Google AI Overviews:
- Heavily favors Reddit and YouTube for consumer queries
- Cites expert review sites for B2B (Consumer Reports, NerdWallet)
- Shows strong preference for pages already ranking in top 10
- Average 4-5 citations per response
ChatGPT:
- Prefers encyclopedic, factual content (Wikipedia)
- Uses 12 sources to generate answers
- Less reliant on Google rankings
- More independent source selection
Strategic takeaway: Start with Google AI Overviews because it has the most users. But the core principles (authority, clarity, comprehensive content, E-E-A-T) work across all platforms.
Content Types That Win AI Overview Citations
Based on citation pattern analysis, here’s what works:
High-performing content types:
Comparison guides – “X vs Y: Complete Comparison”
- Include feature tables
- Show pros and cons
- Provide clear recommendations
- Use first-hand testing data
How-to guides – Step-by-step instructions
- Number your steps clearly
- Include screenshots or visuals
- Add troubleshooting section
- Use HowTo schema
Definitional content – “What is X?”
- Start with clear definition
- Provide context and examples
- Include related concepts
- Use FAQ schema
Data-driven reports – Original research
- Share unique data points
- Include methodology
- Provide visualizations
- Make findings easily quotable
Expert roundups – Multiple perspective pieces
- Quote recognized experts
- Include credentials
- Show diverse viewpoints
- Demonstrate comprehensive coverage
Content length sweet spot:
Research from Writesonic shows longer, comprehensive content has higher citation potential. Aim for 2,000-3,500 words for pillar content.
The Query Intent Strategy
Different query intents require different optimization approaches:
Informational queries (88.1% of AI Overviews):
- Focus: Comprehensive answers
- Format: How-to guides, definitions, comparisons
- Schema: FAQ, HowTo, Article
- Length: 2,000-3,500 words
- Example: “how to calculate customer acquisition cost”
Commercial queries (6-8% of AI Overviews):
- Focus: Product comparisons and evaluations
- Format: Comparison tables, pros/cons lists
- Schema: Product, Review
- Length: 1,500-2,500 words
- Example: “best CRM software for small business”
Transactional queries (10% trigger rate):
- Focus: Product specifications and buying information
- Format: Product pages with detailed specs
- Schema: Product, Offer, Review
- Length: 1,000-2,000 words
- Example: “buy Salesforce enterprise plan”
Tip: Map your content calendar to informational queries first. That’s where 88% of AI Overview opportunities exist, and it’s where prospects start their research.
Advanced Tactics: What’s Working In 2025
Beyond the basics, here are advanced tactics I’m testing with clients:
Tactic 1: Create Topic Clusters
Instead of individual articles, create comprehensive topic clusters where one pillar page links to 5-10 supporting articles. The AI recognizes this as deep expertise.
Example cluster:
- Pillar: “Complete Guide to SaaS SEO”
- Supporting: “Keyword Research for SaaS”
- Supporting: “SaaS Link Building Strategy”
- Supporting: “Technical SEO for SaaS Platforms”
Tactic 2: Update With Freshness Signals
Add dates, recent data, and “Updated [Month Year]” tags. Google’s AI prioritizes fresh information for queries that need current answers.
Tactic 3: Use Conversational Language
Nearly 86% of AI Overviews don’t include the exact query phrase. The AI rewrites and synthesizes. Write in natural, conversational language that the AI can easily paraphrase.
Instead of: “SEO optimization strategies implementation methodologies”
Write: “How to actually implement SEO strategies that work”
Tactic 4: Embed Citations In Content
When you cite other sources, you signal credibility. The AI notices this. Include 5-10 citations to authoritative sources in long-form content.
Tactic 5: Build Answer Databases
Create massive FAQ pages (50-100 Q&As) with proper schema. These become reference sources the AI pulls from repeatedly.
How To Track ROI From AI Overview Citations
Measuring success requires new metrics. Here’s what to track:
Citation metrics:
- Citation rate (% of target queries where you’re cited)
- Average citation position (how prominently you’re featured)
- Competitor citation share
- Citation growth month-over-month
Traffic metrics:
- Branded search traffic (citations often drive branded searches)
- Engagement metrics (time on page, pages per session)
- Conversion rate from AI-referred traffic
- Zero-click impact on overall traffic
Authority metrics:
- Domain authority changes
- Backlink growth from being cited
- Brand mention increases
- Social proof indicators
Business metrics:
- Lead quality from AI-referred traffic
- Customer acquisition cost
- Sales cycle length
- Revenue from AI-influenced customers
Tip: Set up UTM parameters and custom segments in Google Analytics to track AI Overview-influenced sessions. You won’t have direct attribution, but you can track patterns.
The Timelines: What’s Realistic
Every founder asks: “How long until we see results?”
Based on 60+ audits and implementations:
If you’re starting from solid SEO fundamentals:
- Weeks 1-4: Implement schema, restructure top content, optimize E-E-A-T signals
- Months 2-3: Start seeing citations for low-competition queries
- Months 4-6: Meaningful citation rates (20-30%) for target queries
- Months 7-12: Competitive citation rates (40-50%+), measurable traffic impact
If you’re starting from scratch (not ranking on page one):
- Months 1-3: Fix core SEO, build content foundation
- Months 4-6: Start ranking on page one, implement AI optimization
- Months 7-9: First AI Overview citations
- Months 10-18: Meaningful citation presence
This isn’t quick. But companies that started optimizing for AI Overviews in early 2024 now have massive advantages over competitors just starting.
What’s Coming: The Future of AI Overviews
Based on Google’s announcements and testing:
Near-term (next 6-12 months):
- AI Overviews in more countries and languages
- Expansion to commercial and transactional queries
- Integration with Google Shopping (product recommendations in AI Overviews)
- Ads appearing within AI Overview responses
- Video and image integration
Longer-term (12-24 months):
- Personalized AI Overviews based on user history
- More sophisticated multi-step reasoning
- Local search integration
- Real-time information synthesis
- Competitive “compare” features
What this means: The percentage of queries showing AI Overviews will continue growing. Companies that build AI-ready content now will be positioned when this becomes the default experience.
Start Here Tomorrow
If you only do three things after reading this:
- Check your current visibility – Search your top 20 target queries and document whether AI Overviews appear and if you’re cited. This is your baseline.
- Audit your top 10 pages – Do they show clear E-E-A-T signals? Do they have proper schema? Are they comprehensive and answer-first structured? If not, that’s your priority.
- Implement FAQ schema on your best pages – This single technical change has the highest ROI for initial AI Overview visibility.
These three actions take less than two days and will tell you exactly where you stand.
The Honest Truth About AI Overviews And Traffic
Look, I’m not going to sugarcoat this: AI Overviews are reducing clicks to websites. The data is clear. 34.5% drop in CTR isn’t a rounding error.
But here’s the counterpoint: Users who do click through are more qualified. They’ve already read the AI summary, understood the basics, and are looking for deeper information or to take action. They convert 4.4x better than regular search traffic.
This is the new reality:
- Lower volume of traffic overall
- Higher quality of traffic that does come
- Branded searches increase when you’re cited
- Authority signals compound over time
The companies winning aren’t fighting AI Overviews. They’re adapting their strategy to get cited and building brand authority that drives traffic through other channels.
Why This Actually Matters For SaaS
If you’re running a SaaS company, this shift hits differently than other industries.
Why SaaS should care:
- Your buyers research extensively – They’re asking informational questions where AI Overviews appear 88% of the time
- B2B purchase cycles are long – Getting cited builds brand awareness early in the consideration phase
- Authority matters in SaaS – Being cited by Google’s AI is a massive trust signal
- Your competitors are slow – Most SaaS companies haven’t adapted yet. Early movers win big.
- Content is your moat – Unlike paid ads, content that gets cited compounds over time
I worked with a SaaS company that went from 0% to 45% citation rate on their core queries. Their organic traffic actually dropped 15%. But branded searches increased 67%, demo requests increased 34%, and sales cycle shortened by 18 days on average.
That’s the new math of SEO.
Final Thoughts
AI Overviews aren’t killing SEO. They’re evolving it.
The fundamentals still matter – technical SEO, quality content, domain authority, user experience. But the end goal has shifted from “rank #1” to “get cited by the AI that synthesizes the answer.”
This requires:
- More comprehensive content (not thin posts)
- Stronger authority signals (not just backlinks)
- Better technical implementation (schema is mandatory)
- Third-party validation (you can’t just self-promote)
Is it more work? Yes. But it’s also higher leverage. Get cited once, and that citation can influence thousands of searches. Build the authority to get cited consistently, and you become the default reference in your category.
The window for early advantage is closing. AI Overviews grew from 25% to 57% of searches in 10 months. By this time next year, they’ll be on 70%+ of queries.
The question isn’t whether to adapt. It’s whether you’ll adapt while there’s still an advantage to being early.
Want an honest assessment of your AI Overview visibility and a real plan to get cited?
Reach out. I’ll audit your top queries, show you where you stand, and give you the specific changes that will actually move the needle. No cookie-cutter playbook. Just what works for your specific situation.
