AI SEO: What Actually Changed and What’s Still Just SEO
TL;DR
- Most of what’s sold as “AI SEO” is traditional SEO wearing a new label. An Ahrefs study from July 2025 found that 76% of AI Overview citations pull from pages already ranking in Google’s top 10, meaning your existing SEO work is the foundation for AI visibility.
- AI search visitors convert at 4.4x the rate of traditional organic visitors according to Semrush research, but they still represent roughly 1-2% of total site traffic for most businesses. The volume is small. The value per visit is enormous.
- Use the SEO-AEO-GEO prioritization framework in this article to figure out which layer matters most for your specific business type, because spreading effort equally across all three is a fast way to accomplish nothing.
- 84% of B2B SaaS CMOs now use ChatGPT, Claude, or Perplexity for vendor discovery according to Wynter’s January 2026 survey, up from zero in 2024. If you sell to other businesses, AI search visibility isn’t optional anymore.
I spent most of 2025 watching smart marketers panic about AI SEO. They’d read some LinkedIn post about “Generative Engine Optimization,” immediately create a Notion doc titled “GEO Strategy 2026,” and then have absolutely no idea what to put in it. I know this because I did the exact same thing in March.
Here’s what six months of testing, reading studies, and talking to people smarter than me eventually made clear: roughly 80% of what gets labeled “AI SEO” is just… SEO. Good, boring, been-doing-it-for-a-decade SEO. The other 20% is genuinely new, and that’s the part worth your attention.
This article won’t rehash the same “AI is transforming search!” overview you’ve read twelve times already. Instead, I’ll break down what actually changed, what didn’t, and give you a decision framework for where to spend your time. Because the worst thing you can do right now is treat every new acronym as an equal priority.
Why 80% of “AI SEO” is just SEO with a rebrand
Google’s Danny Sullivan said it plainly at WordCamp US in August 2025: “Good SEO is good GEO.” He wasn’t being dismissive. He was stating a technical fact about how retrieval-augmented generation (the system powering AI Overviews, ChatGPT’s web search, and Perplexity) actually works.
Retrieval-augmented generation (RAG) is the process where an AI model searches the web, retrieves sources, and then synthesizes an answer from those sources. And where does it retrieve those sources from? Search engines. Google. Bing. The same indexes your traditional SEO targets.
The data backs this up. Ahrefs analyzed thousands of AI Overview results and found that 76% of cited pages already rank in Google’s top 10. That’s not a coincidence. It’s the architecture. AI search engines are not building some parallel universe of content authority. They’re reading the same scoreboard you’ve been trying to climb for years, then summarizing the winners.
Lily Ray, VP of SEO Strategy at Amsive, tested this directly. In her MozCon presentation in late 2025, she shared that articles her team published using standard SEO best practices (no special “GEO tactics”) were cited by multiple AI platforms within hours. No fancy chunking strategy. No prompt engineering. Just clear, well-structured, authoritative content.
“The tactics to earn AI search visibility are really just evolutions of existing SEO and marketing best practices.”
— Lily Ray, VP of SEO Strategy and Research at Amsive (MozCon 2025 Presentation)
So why is an entire cottage industry selling GEO as something fundamentally different? Partially because new acronyms sell consulting packages. Partially because there genuinely are some new elements (we’ll get to those). But mostly because people confuse new surfaces with new fundamentals.
Think of it like this. When Instagram launched shopping features, nobody rebranded ecommerce as “IGO” (Instagram Commerce Optimization). You just adapted your existing product pages and distribution strategy to work on a new surface. That’s what’s happening with AI search. Same content. Same authority signals. New place your content shows up.
The 20% that’s genuinely new (and actually matters)
OK, so I just spent several paragraphs arguing that AI SEO is mostly traditional SEO. But “mostly” is doing some heavy lifting there. Here’s what’s different enough to change your workflow.
The measurement layer is brand new. Before AI search, you tracked rankings and clicks. Now you need to track whether your brand gets mentioned in AI-generated answers, how frequently, and in what context. Seer Interactive’s September 2025 study found that brands cited in AI Overviews earned 35% more organic clicks and 91% more paid clicks than brands that weren’t cited. Being named in the answer matters, even if users don’t click your specific citation link.
The buyer journey starts differently in B2B. Wynter’s January 2026 survey of 101 mid-market B2B SaaS CMOs found that 84% now use AI tools like ChatGPT, Claude, and Perplexity for vendor discovery. That’s up from 24% in 2025 and literally zero in 2024. Even more telling: 68% of those CMOs start their vendor searches in AI tools before touching Google. If you’re a B2B company and your brand doesn’t show up when someone asks ChatGPT “What are the best tools for [your category]?”, you might never enter the consideration set.
Zero-click behavior has accelerated. About 60% of searches on traditional search engines now end without a click according to Bain & Company research from 2025. When AI summaries appear specifically, only 8% of users click a traditional link according to Pew Research. That means your content needs to do its job (build brand awareness, establish authority, answer questions) inside the AI summary itself, not just on your website.
The conversion math is flipped. Despite lower volume, Semrush’s AI search traffic study found that the average AI search visitor is worth 4.4x more than a traditional organic visitor. Users who click through from AI answers are typically further along in their research. They’ve already gotten the overview from the AI. They’re visiting your site for a specific reason.
| What Changed | Why It Matters |
|---|---|
| New measurement: AI visibility and share of voice | You can be “invisible” in AI even while ranking #1 on Google |
| B2B buyers start with AI before Google | 68% of CMOs search AI tools first (Wynter, Jan 2026) |
| 60% of searches yield zero clicks | Your brand message needs to work inside AI answers |
| AI visitors convert 4.4x higher | Lower volume, much higher value per visit |
| AI search traffic up 527% YoY | Still small (1-2% of total), but growing fast (Semrush) |
The SEO-AEO-GEO framework: where to actually spend your time
Let me save you from the thing I see most teams doing wrong. They read about SEO, AEO, and GEO, treat them as three separate strategies requiring three separate workstreams, and end up doing all three poorly. These aren’t three different games. They’re three layers of the same game, and the order you tackle them matters.
SEO (Search Engine Optimization) is your content ranking on Google and Bing through traditional organic search. This is the foundation. Nothing above it works if this layer is broken.
AEO (Answer Engine Optimization) is your content being cited as a source by AI answer engines like ChatGPT, Perplexity, and Google AI Overviews. AEO is what happens when your SEO is good enough that AI retrieval systems pick your pages as references.
GEO (Generative Engine Optimization) is your brand being named, recommended, and accurately described within AI-generated answers, even when your specific page isn’t cited. GEO is the reputation and brand-awareness layer.
Here’s how I think about priority:
- Fix SEO first. If your pages don’t rank in the top 10-20 organically, AI systems probably aren’t retrieving them. Remember: 76% of AI Overview citations come from existing top-10 results. Your technical health, content quality, and backlink profile are the entry ticket.
- Structure for AEO second. Once you rank, make your content easy for AI systems to extract. Self-contained answers under clear headings. Definitions in the first sentence of a section. Statistics with named sources. This isn’t new magic. It’s the same stuff that wins featured snippets.
- Build GEO through brand authority third. Get mentioned on authoritative third-party sites. Earn coverage in industry publications. Build a presence on Reddit, YouTube, and Wikipedia (where LLMs pull heavily from). This is digital PR and reputation management, and it’s always been part of good marketing.
Pro Tip: Before spending a dollar on GEO-specific tactics, ask your target query in ChatGPT, Perplexity, and Google AI Mode. If your competitors show up and you don’t, check whether you rank for the same queries on Google first. In most cases I’ve seen, the AI visibility problem is actually an organic ranking problem wearing a costume.
Who should weight each layer differently? It depends entirely on your business model.
| Business Type | SEO Priority | AEO Priority | GEO Priority | Why |
|---|---|---|---|---|
| B2B SaaS ($5M+ ARR) | High | High | Very High | 84% of CMOs now research vendors via AI (Wynter). Being named in the answer is everything. |
| Local service business | Very High | Medium | Low | Local SEO and Google Business Profile still dominate. AI Overviews appear mostly on informational queries, not “plumber near me.” |
| E-commerce (DTC) | Very High | Medium | Medium | Product pages need organic rankings. AI shopping features are emerging but early. |
| Content publisher | High | Very High | Medium | 88% of AI Overview triggers are informational queries (Semrush). Your content IS the answer. Being cited is survival. |
| Solo consultant / freelancer | Medium | Medium | High | People ask AI “Who’s the best [your specialty] consultant?” Your name needs to be in that answer. |
How to structure content that AI systems actually cite
Here’s where a lot of “AI SEO” advice goes sideways. People obsess over “chunking” content for AI. They restructure entire pages into rigid Q&A formats thinking it’ll improve AI citability. Lily Ray addressed this directly in her MozCon talk, pointing out that AI models don’t see your formatting. They see tokens. They decide how to slice your content, not you.
What you can control is clarity. Content that gets cited across AI platforms tends to share a few traits I’ve noticed after reviewing hundreds of AI-generated responses:
- Self-contained statements. Each important claim stands alone without requiring the reader to have read the previous paragraph. AI engines extract individual passages, not entire articles.
- Named entities everywhere. “Revenue grew 40%” gets ignored. “Acme Corp’s revenue grew 40% in Q3 2025 according to their earnings report” gets cited. Specificity signals reliability.
- Definitions in the first sentence. When you introduce a concept, define it immediately. AI systems love pulling the sentence that follows “X is…” because it cleanly answers the user’s question.
- Statistics with attribution. Every number paired with who reported it and when. AI systems are more likely to surface claims that come with built-in credibility markers.
Do these sound like featured snippet best practices from 2019? They should. Lily Ray made this exact point. The SEO industry has been doing what’s now called “AEO” for years under different names. FAQ schema, how-to markup, definition-style content, concise answer paragraphs. All of it predates the GEO hype cycle.
What’s worth adding to your existing process is a simple audit step: after publishing, drop your target query into ChatGPT and Perplexity. Check if your content appears. If it doesn’t, don’t assume you need a “GEO overhaul.” Check your organic rankings first. Check whether competitors who are cited have stronger backlink profiles or more comprehensive content. Usually the fix is the same one you would have applied in 2023.
The metrics that actually matter now (and the ones to stop obsessing over)
Search Engine Land published a useful piece in February 2026 noting that 24% of CMOs were using AI for vendor research in 2025, and Wynter’s updated survey shows that jumped to 84% by January 2026. That rate of change means your measurement framework needs to evolve, but not in the way most people think.
You don’t need to abandon traditional SEO metrics. You need to add an AI visibility layer on top of them. Here’s what I track now that I didn’t track 18 months ago:
AI Share of Voice. When someone asks an AI tool about my client’s category, does our brand get mentioned? How often compared to competitors? Tools like Semrush’s AI Visibility Toolkit, Profound, and Peak AI now track this. It’s the new version of “where do we rank?”
AI-Referred Traffic and Conversion. Set up a custom channel in GA4 that groups referrals from ChatGPT, Perplexity, Claude, and Copilot domains. Even if the volume is small today, tracking it lets you spot the growth curve early. For one B2B client, AI referral traffic went from undetectable in June 2025 to 3% of total pipeline-influenced visits by January 2026.
Branded Search Lift. When AI tools recommend your brand, people often navigate directly to your site or search your brand name in Google. Watch for correlated lifts in branded search volume alongside AI visibility gains. This is the metric that bridges the gap between “we got mentioned in ChatGPT” and “it actually drove business.”
What I’ve stopped stressing about: raw organic click volume in isolation. Seer Interactive’s data shows organic CTRs dropped from 1.76% to 0.61% on queries where AI Overviews appear. That number is going in one direction. Fighting it is like arguing with weather. The smarter play is making sure your brand presence inside those AI summaries does the awareness and trust-building work that a click used to do.
The uncomfortable truth about AI-generated content and SEO
While we’re being honest about what’s changed, let’s talk about the elephant. AI content now accounts for over 17% of pages in Google’s top 20 search results according to Originality.ai’s tracking. That’s up from 2.27% in 2019. Google hasn’t banned AI content. They’ve said quality and helpfulness matter regardless of how content is produced.
But here’s what I’ve observed: the AI content that ranks is rarely “generate article, hit publish.” It’s AI-assisted content where a human with actual expertise directs the structure, adds proprietary data, injects real opinions, and edits for accuracy. The gap between “AI-written” and “AI-assisted” is the gap between commodity content and content that earns citations.
Why does this matter for AI SEO specifically? Because AI answer engines prioritize content that demonstrates E-E-A-T (Experience, Expertise, Authoritativeness, Trustworthiness). A purely AI-generated article with no named author, no original data, and no specific experience signals gets treated as background noise. An article that cites original research, quotes named experts, and takes specific positions based on real-world testing is exactly what these systems want to surface.
The irony isn’t lost on me. To rank in AI search results, you need to produce content that’s clearly not just AI output. You need the human fingerprints that AI struggles to fake: named experience, proprietary data, genuine opinions, and the kind of specific, weird details that only come from actually doing the thing you’re writing about.
Frequently Asked Questions About AI SEO
Is AI SEO a completely different discipline from traditional SEO?
No. AI SEO builds directly on traditional SEO fundamentals. Ahrefs research from July 2025 found that 76% of AI Overview citations come from pages already ranking in Google’s top 10 organic results. Google’s Danny Sullivan confirmed at WordCamp US in August 2025 that “good SEO is good GEO.” The new elements are measurement (tracking AI visibility), content structuring (making information easy to extract), and brand reputation across the web.
How much traffic do AI search platforms actually drive?
For most websites, AI search platforms like ChatGPT and Perplexity drive roughly 1-2% of total site traffic as of early 2026. Semrush’s 2025 AI Traffic Report found AI search traffic is up 527% year-over-year, but from a small base. The real story is quality, not quantity: Semrush found that the average AI search visitor is worth 4.4x more than a traditional organic search visitor in conversion value.
Should I block AI crawlers from accessing my site?
Blocking AI crawlers means AI platforms can’t cite your content, which removes you from an increasingly important discovery channel. Only about 11% of website owners completely block AI scrapers according to AIOSEO’s 2026 statistics roundup. Unless you have a specific monetization reason (like a major news publisher protecting paywalled content), blocking AI crawlers likely costs you more visibility than it protects.
What tools should I use to track AI visibility?
Several tools now track how often your brand appears in AI-generated answers. Semrush’s AI Visibility Toolkit, Profound, and Peak AI are the most established options. For a quick manual check, simply enter your target queries into ChatGPT, Perplexity, and Google’s AI Mode and note whether your brand or URLs appear. Lily Ray’s team at Amsive recommends setting up a custom “AI Search” channel in GA4 to group traffic from LLM referrer domains.
Do I need to create different content for AI search versus Google?
You don’t need separate content for AI search. The same content that ranks well on Google, specifically content that’s well-structured, authoritative, factual, and provides clear answers to specific questions, is what AI systems retrieve and cite. The main difference is ensuring each important statement in your content is self-contained and makes sense without surrounding context, since AI engines extract individual passages rather than linking to full articles.
There’s a version of AI SEO that’s genuinely worth your time. It’s the version where you keep doing great SEO, add a measurement layer for AI visibility, and make deliberate moves to get your brand mentioned across the web. That’s it. No secret decoder ring required.
If you’d rather have a team handle the strategy, execution, and measurement across all three layers (SEO, AEO, and GEO), LoudScale works with growth-stage companies on exactly this kind of integrated search visibility.
But honestly? Start with the framework above. Fix what’s broken in traditional SEO. Audit your AI visibility on 10 queries that matter to your business. Then decide where the gaps actually are before you invest in anything with a shiny new acronym.