Most SEO guides read like they were written in 2019 and lightly updated with a new date. “Create great content. Build backlinks. Optimize your title tags.” Thanks, very helpful.
Here’s what’s actually changed: the search landscape has fractured. Google is still dominant, but ChatGPT, Perplexity, Gemini, and AI Overviews now answer a significant chunk of queries before users ever see a traditional result. Your SEO strategy needs to account for all of it.
I’ve spent the last 18 months helping businesses navigate this shift. What I’ve found is that the fundamentals haven’t died-they’ve expanded. You need a framework that covers both traditional search and the AI layer sitting on top of it.
This is that framework.
The 5-Pillar Framework: Why You Need All Five
Before diving into each pillar, let me explain why a framework matters more than a checklist.
We worked with a B2B SaaS company last year that had impeccable technical SEO. Fast site, clean architecture, perfect structured data. They ranked on page 2-3 for everything. Why? Zero topical authority and a backlink profile thinner than gas station coffee.
Another client had incredible content and strong backlinks but a site so slow and poorly structured that Google couldn’t efficiently crawl half their pages. They were leaving 40% of their traffic potential on the table.
The point: each pillar multiplies the others. Weakness in one creates a ceiling for all five. Here’s the framework:
- Technical + Indexing – Can search engines (and AI crawlers) access and understand your site?
- Topical Authority + Content Systems – Do you demonstrate deep expertise that machines and humans recognize?
- Authority Signals – Do other credible sources vouch for you?
- AI/LLM Visibility – Are you showing up in ChatGPT, Gemini, and Perplexity?
- User Signals + Experience – Do real humans engage with and trust your content?
Let’s break each one down with specific, actionable detail.
Pillar 1: Technical SEO + Indexing
Technical SEO isn’t glamorous. Nobody’s posting TikToks about their XML sitemaps. But it’s the foundation everything else sits on, and in a world where AI crawlers are hitting your site alongside Googlebot, it matters more than ever.
Crawlability and Site Architecture
Think of your site architecture as a filing system. If Google (or GPTBot, or PerplexityBot) can’t find a page within 3-4 clicks from your homepage, that page might as well not exist.
Here’s what a clean architecture looks like in practice:
- Flat hierarchy: No page should be more than 3 clicks from the homepage. I audit sites regularly where key service pages are buried 5-6 levels deep.
- Logical URL structure:
/services/technical-seo/not/pages/services-we-offer/seo-services/technical/ - Internal linking that mirrors topic relationships: Your pillar pages should link to cluster content and vice versa. This isn’t just for users-it’s how search engines understand your topical coverage.
- Clean XML sitemaps: Only include indexable, canonical URLs. I see bloated sitemaps with redirects, noindexed pages, and parameterized URLs constantly. Every junk URL in your sitemap wastes crawl budget.
Tools I actually use for this: Screaming Frog for crawl audits, Sitebulb for visualization, and Google Search Console’s crawl stats report for real Googlebot behavior data.
Core Web Vitals: What Actually Moves the Needle
Let me be blunt: Core Web Vitals are a tiebreaker, not a ranking factor that will catapult you from page 5 to page 1. But when you’re competing for positions 1-5 against sites with similar content and authority, page experience matters.
The three metrics that matter:
- Largest Contentful Paint (LCP): Target under 2.5 seconds. The biggest killer? Unoptimized hero images and render-blocking JavaScript. Switch to WebP/AVIF, implement lazy loading, and defer non-critical scripts.
- Interaction to Next Paint (INP): This replaced FID and is harder to fix. It measures responsiveness across the entire session. Heavy JavaScript frameworks (looking at you, bloated React bundles) are the usual culprit.
- Cumulative Layout Shift (CLS): Set explicit dimensions on all images and embeds. Use
font-display: swapfor web fonts. This is the easiest one to fix and the most commonly broken.
Don’t chase a perfect 100 PageSpeed score. I’ve seen sites with scores of 72 outrank sites with 99 because the content and authority were stronger. Get into the “good” range for all three CWV metrics, then move on to higher-impact work.
Structured Data: Your Machine-Readable Business Card
Structured data has gone from “nice to have” to “essential infrastructure.” Here’s why: AI systems parse structured data to understand entities, relationships, and facts. If your site doesn’t speak this language, you’re invisible to an increasingly important audience.
Priority schema types by business model:
- All businesses: Organization, WebSite, BreadcrumbList
- Service businesses: Service, LocalBusiness, FAQPage
- E-commerce: Product, Review, AggregateRating, Offer
- Content publishers: Article, HowTo, VideoObject
- Multi-location: LocalBusiness with multiple locations, each with unique identifiers
Use Google’s Rich Results Test and Schema Markup Validator religiously. Broken schema is worse than no schema-it can confuse both search engines and AI crawlers.
The New Consideration: AI Crawler Access
This is the part most technical SEO guides miss entirely. GPTBot, PerplexityBot, ClaudeBot, and others are now crawling the web to train and retrieve information. Your robots.txt decisions have real consequences.
My recommendation: allow AI crawlers access to your public content. If you block them, you’re opting out of being cited in AI-generated answers. Some publishers have valid reasons to block (licensing, content monetization), but for most businesses, AI visibility is a net positive.
Check your robots.txt right now. If you see User-agent: GPTBot / Disallow: /, that’s a strategic decision you may not have intentionally made.
Not Sure If AI Crawlers Can Access Your Site?
Our technical audit checks crawler access, structured data, and 140+ technical factors that affect both traditional and AI search visibility.
Pillar 2: Topical Authority + Content Systems
“Just create great content” is the “just be yourself” of SEO advice. Technically true, practically useless.
What actually builds topical authority is a systematic approach to covering a subject area so thoroughly that search engines-and now AI models-recognize you as a definitive source.
Topic Clusters: The Architecture of Authority
A topic cluster isn’t just “write a bunch of articles about the same thing.” It’s a deliberate content architecture:
- Pillar page: A comprehensive resource covering the broad topic (like this guide you’re reading)
- Cluster content: Focused articles addressing specific subtopics, questions, and angles
- Internal links: Bidirectional links between pillar and cluster content, using descriptive anchor text
Here’s a real example. We helped a cybersecurity firm build a topic cluster around “endpoint security.” The pillar page was a 4,000-word guide. We built 23 supporting articles covering specific aspects: EDR vs. antivirus, endpoint security for remote teams, compliance requirements by industry, zero-trust endpoint architecture, and so on.
Within 6 months, they went from ranking for 340 keywords in the endpoint security space to over 2,100. The pillar page went from not ranking to position 3 for “endpoint security” (monthly volume: 8,100). But here’s the kicker-their content also started appearing in AI Overviews and Perplexity citations for related queries.
Entity-Based Content Strategy
This is where the game has fundamentally shifted. Google and AI models don’t just match keywords anymore-they understand entities (people, companies, concepts, products) and the relationships between them.
What this means for your content strategy:
- Build your brand as a recognized entity: Consistent information across your site, social profiles, Wikipedia (if notable enough), Wikidata, Crunchbase, and industry directories.
- Reference and link to authoritative entities: When you mention tools, methodologies, or research, link to the source. This helps AI systems understand context.
- Create content that connects entities: “How [Your Product] Integrates with [Known Platform]” is stronger than generic feature descriptions because it creates entity relationships AI can parse.
Topical authority isn’t about volume-it’s about coverage and depth. Twenty well-structured articles that comprehensively cover a topic will outperform 200 thin posts every time. Build content systems, not content calendars.
Content Quality Signals That Actually Matter
Google’s helpful content system and AI training data preferences both reward the same things:
- First-hand experience: Share original data, case studies, screenshots, and results. “We tested this with 47 client sites” beats “experts say” every time.
- Specificity over generality: Name the tools. Share the numbers. Give the exact steps. Vague advice signals you’ve never actually done the work.
- Clear point of view: AI can generate bland, balanced summaries better than any human. Your competitive advantage is having strong, defensible opinions backed by experience.
- Updated information: Content freshness matters especially for topics that evolve. Build an update schedule into your content operations.
One more thing: stop writing for word count. I’ve seen 1,200-word articles outrank 5,000-word “ultimate guides” because they answered the searcher’s question directly and completely. Match your content length to the topic’s complexity and the searcher’s intent.
Pillar 3: Authority Signals
I need to address something head-on: there’s a narrative floating around that “backlinks are dead” or “links don’t matter in the AI era.” This is wrong. Demonstrably, measurably wrong.
Backlinks remain one of the strongest ranking signals in Google’s algorithm. And the same authority signals that help you rank in traditional search also influence whether AI systems cite you. Here’s why: AI models are trained on web data, and the sources they learn to trust are-surprise-the same ones that have strong authority signals.
Backlinks Still Matter (But Quality Over Quantity Has Never Been More True)
The link building landscape has changed, though. Here’s what works now:
- Digital PR: Creating genuinely newsworthy content, data studies, and expert commentary that earns links from publications. This is harder than buying links from PBNs, which is exactly why it works.
- Original research: Survey your customers. Analyze your data. Publish findings. We ran a data study for a fintech client analyzing 10,000 small business financial decisions-it earned 340+ linking root domains in 3 months.
- Expert-sourced content: HARO is largely dead, but platforms like Connectively, Qwoted, and Featured.com still connect journalists with sources. The key is responding quickly with genuinely useful, specific quotes.
- Strategic partnerships: Co-create content with complementary (non-competing) businesses. Webinars, joint research, co-authored guides-all generate natural, contextual links.
Link schemes, PBNs, and mass guest posting on low-quality sites will catch up with you. Google’s spam detection has gotten significantly better, and a manual action or algorithmic demotion can wipe out months of work overnight. I’ve seen it happen to agencies who should know better.
Brand Mentions and Co-Occurrence
Unlinked brand mentions-places where your brand is discussed without a hyperlink-also function as authority signals. Google has patents around using brand mentions as implicit links, and AI models definitely pick up on brand co-occurrence patterns.
If your brand consistently appears alongside authoritative sources in industry discussions, AI systems learn that association. Practical ways to build this:
- Contribute expert commentary to industry publications
- Participate in podcast interviews (show notes create mentions)
- Speak at conferences and events (speaker bios create mentions)
- Engage in community discussions on platforms like Reddit, LinkedIn, and industry forums-genuinely, not spammy self-promotion
Brand Search Volume as a Ranking Signal
This is underappreciated. When people search for your brand name, it signals to Google that you’re a real, recognized entity. High brand search volume correlates strongly with higher rankings for non-branded terms.
How to build brand search demand: invest in brand marketing. Podcast ads, YouTube sponsorships, event sponsorships, thought leadership-anything that makes people aware of your brand so they search for it directly. This is the bridge between marketing and SEO that most pure-play SEO agencies completely miss.
Track your branded search volume in Google Search Console. If it’s flat or declining while you’re investing in SEO, you have a brand awareness problem that will cap your organic growth. The best SEO strategy in the world can’t overcome brand obscurity.
Pillar 4: AI/LLM Visibility
This is the genuinely new frontier, and where most businesses are completely asleep at the wheel.
ChatGPT, Gemini, Perplexity, and Copilot are answering an enormous and growing volume of queries. When someone asks “what’s the best project management tool for remote teams” in ChatGPT, your brand is either mentioned or it isn’t. And unlike Google rankings, you can’t see your “position” in a dashboard.
How AI Systems Choose What to Cite
Understanding the mechanics helps you optimize for them:
- Training data: LLMs learned from massive web crawls. If your brand was well-represented in high-quality content when training data was collected, you have an advantage. This is historical and hard to change retroactively.
- Retrieval-Augmented Generation (RAG): Tools like Perplexity and ChatGPT’s browsing mode pull real-time information from the web. This is where your current SEO efforts directly impact AI visibility.
- Source authority: AI systems preferentially cite sources they assess as authoritative. The same signals that build authority for traditional SEO work here.
For a detailed breakdown of how to optimize specifically for AI citation, I wrote a companion guide: How to Get Your Brand Featured in ChatGPT and Perplexity.
Practical Steps for AI Visibility
Here’s what you should be doing right now:
- Create citation-worthy content: Original data, unique frameworks, definitive guides that AI systems want to reference when answering questions in your space.
- Strengthen your entity presence: Ensure your brand information is consistent and present across Wikipedia, Wikidata, Crunchbase, LinkedIn, and authoritative directories.
- Implement comprehensive structured data: This helps AI systems parse your content accurately.
- Allow AI crawlers: Check robots.txt for GPTBot, PerplexityBot, and other AI crawler restrictions.
- Monitor AI mentions: Manually query ChatGPT, Perplexity, and Gemini for terms in your space. Tools like Otterly.ai and Peec.ai are building automated monitoring, though the space is still maturing.
Where Does Your Brand Stand in AI Search?
We audit your visibility across ChatGPT, Perplexity, and Gemini for your target queries-and build a strategy to get you cited.
The Relationship Between Traditional SEO and AI Visibility
Here’s what most people miss: strong traditional SEO is the primary driver of AI visibility.
When Perplexity answers a query, it searches the web and cites results-often from top-ranking pages. When ChatGPT browses, it pulls from authoritative sources. When Gemini generates an AI Overview, it draws from top-ranking content.
The relationship is circular and reinforcing: good SEO → high rankings → more likely to be crawled and cited by AI → more brand visibility → more brand searches → better SEO. If you had to choose between “optimizing for AI” and “doing great SEO,” choose the latter. It achieves both.
Pillar 5: User Signals + Experience
Google has spent two decades trying to measure whether searchers are actually satisfied with search results. Their ability to do this has gotten frighteningly good.
The Signals That Influence Rankings
Google doesn’t publicly confirm using “user signals” as ranking factors, but the evidence from leaked documents, patents, and observable ranking behavior is overwhelming:
- Click-through rate (CTR): Pages with higher-than-expected CTR for their position tend to move up. This is why title tags and meta descriptions matter beyond just keyword inclusion-they need to compel clicks.
- Pogo-sticking: When users click a result, immediately return to the SERP, and click a different result, that’s a strong negative signal. It means your page didn’t satisfy the query.
- Dwell time and engagement: Longer time on page, scrolling depth, and interaction signals suggest content satisfaction. Short visits with quick bounces suggest the opposite.
- Task completion: For certain queries, Google can infer whether the user accomplished their goal. If someone searches “how to remove background from image” and then doesn’t search further, the result likely solved their problem.
How to Actually Improve User Signals
This isn’t about tricks. It’s about genuinely serving the user better than competing results.
Write better title tags. Not keyword-stuffed monstrosities-titles that promise specific value and deliver it. “7 Link Building Strategies (That Worked in Our Tests)” outperforms “Link Building Strategies: The Ultimate Guide to Building Links” every time.
Match search intent precisely. If someone searches “best CRM for small business,” they want a comparison with recommendations, not a 2,000-word essay on what CRM software is. Sounds obvious, but most content misses intent alignment.
Front-load value. Put the answer, the key insight, or the most important information near the top. Users who find immediate value stay and read more. Users who have to wade through filler leave.
Use Microsoft Clarity (it’s free) to watch real user sessions on your top pages. You’ll learn more about content engagement in 30 minutes of session recordings than from any analytics dashboard. I’ve caught issues like broken interactive elements, confusing layouts, and content sections that every user skips-things analytics can’t show you.
The UX Factors That Matter for SEO
- Mobile experience: Over 60% of searches happen on mobile. If your content is painful to read on a phone, you’re losing the majority of your audience.
- Page speed (yes, again): Slow pages don’t just hurt Core Web Vitals scores-they hurt engagement. Every 100ms of latency costs real engagement.
- Content formatting: Walls of text kill engagement. Use headers, bullet points, images, callout boxes, and white space to make content scannable.
- No intrusive interstitials: Full-screen pop-ups on mobile, especially on entry, are both a ranking negative and a user experience disaster.
Putting It All Together: Implementation Priority
If you’re feeling overwhelmed, here’s how I’d prioritize for a business starting from scratch or overhauling their approach:
Month 1-2: Technical foundation. Fix crawlability issues, implement structured data, ensure AI crawlers have access, get Core Web Vitals into the green. You can’t build on a broken foundation.
Month 2-4: Content system. Map your topic clusters, create your pillar pages, and start building cluster content. Focus on your highest-value topic area first.
Month 3-6: Authority building. Launch digital PR campaigns, build industry relationships, create linkable assets (original research, tools, data studies).
Ongoing: User experience optimization. Continuously test and improve title tags, content quality, and page experience based on real user behavior data.
Ongoing: AI visibility monitoring. Track your presence in AI-generated answers, optimize content for citation-worthiness, and stay current as the AI search landscape evolves rapidly.
Don’t try to do everything at once. I’ve watched businesses spread themselves so thin across all five pillars that they make no meaningful progress in any of them. Pick one or two areas where you’re weakest, make measurable progress, then expand. Sequential intensity beats parallel mediocrity.
The Integration Layer: How AI Search Changes the Game
Let me be direct about what’s different now versus even two years ago.
The search journey has fragmented. A potential customer might:
- Ask ChatGPT for recommendations
- Search Google to validate what they heard
- Read reviews on G2 or Reddit
- Ask Perplexity for a comparison
- Finally visit your site
You need to be present at multiple touchpoints, not just Google position 1. This is why the framework has five pillars, not three. Traditional SEO (pillars 1-3) feeds AI visibility (pillar 4), and user experience (pillar 5) determines whether any of it converts.
The businesses winning right now aren’t the ones doing one thing extraordinarily well. They’re the ones executing competently across all five pillars and continuously improving.
For a deeper look at how traditional SEO and AI optimization work together, read our guide on AEO vs SEO: why your business needs both.
Measuring What Matters
You can’t improve what you don’t measure, and the metrics for this expanded framework go beyond traditional SEO KPIs.
Traditional SEO Metrics (Still Essential)
- Organic traffic (Google Search Console, not just GA4-GSC shows impressions for queries you rank for but don’t get clicks on)
- Keyword rankings for target terms
- Organic conversions and revenue
- Technical health scores from regular crawl audits
- Backlink acquisition rate and quality
AI Visibility Metrics (New and Critical)
- Brand mentions in AI-generated responses (manual monitoring + emerging tools)
- Citation frequency in Perplexity results
- Inclusion in Google AI Overviews for target queries
- AI referral traffic (check your GA4 for traffic from chat.openai.com, perplexity.ai, etc.)
User Signal Metrics
- CTR by query and page (Search Console)
- Engagement rate and time on page (GA4)
- Core Web Vitals field data (CrUX report)
- Session recordings and heatmaps (Microsoft Clarity)
AI search optimization isn’t a replacement for SEO-it’s an expansion of it. The 5-pillar framework works because each element reinforces the others: technical excellence enables content discovery, quality content attracts authority signals, authority signals drive AI citations, and great user experience converts all that visibility into business results.
What Happens If You Ignore This
I’ll end with something I’ve observed directly: companies that treat AI search as “someone else’s problem” are already losing ground.
One of our clients in the financial advisory space came to us after noticing a 23% decline in organic leads over 8 months-despite stable Google rankings. The traffic was going to AI answers. People were getting their questions answered in ChatGPT and never clicking through to the website.
We rebuilt their content strategy around the 5-pillar framework, with particular emphasis on making their content citation-worthy for AI systems. Within 5 months, they were being cited in ChatGPT and Perplexity responses, their AI referral traffic grew from essentially zero to 12% of total organic, and their traditional rankings improved too because the same improvements that drive AI visibility also strengthen conventional SEO.
The window to establish AI visibility is open now, but it won’t stay open forever. As more businesses catch on, the competition for AI citations will intensify just like the competition for Google page 1 did years ago. The early movers will have a significant, compounding advantage.
The question isn’t whether to adapt. It’s how quickly you can get started.
