Search has split into two fronts: the classic results page and the AI-generated answers inside ChatGPT, Perplexity and Google’s own AI Overviews. Optimizing for only one of them leaves visibility on the table. A page built purely for organic rankings will not automatically get pulled into an AI answer, and a page that is easy for AI to cite may still struggle in Google if the fundamentals are weak.
The stakes are already commercial. According to Semrush, 77% of ChatGPT users in the USA treat the tool like a search engine, and nearly 60% of consumers use AI tools to shop. That means discovery, trust and conversion are moving into conversations, not just clicks.
Two disciplines, one trust problem
Answer engine optimization (AEO) is about getting your content retrieved and cited when someone asks an AI tool a question. Search engine optimization (SEO) is about ranking in Google and Bing. They sound different, but the underlying inputs overlap heavily.
Semrush compared AI answers with Google organic results and found a pattern: AI favors trusted domains, but it does not always cite the exact pages that rank. It often pulls deeper subpages, blog posts or documentation from authoritative sites instead. The overlap varies by platform. Perplexity has the strongest alignment with Google, while ChatGPT has the weakest, though it still leans on high-authority domains. Google AI Mode’s sidebar shows lower overlap with the top 10, while its additional links overlap much more.
The practical point: strong organic rankings and domain authority are still valuable, but they are not the same as being quotable. AI systems reward sites that are easy to parse, reference and lift from.
What actually gets you cited
The source lists six levers, but they cluster around three ideas: original evidence, external trust signals and extractable structure.
- Publish original research and first-hand experience. Data you gathered, tests you ran or expert quotes you secured give AI systems something that cannot be generated from thin air.
- Build backlinks and unlinked brand mentions. Backlinks remain an authority signal. Brand mentions matter especially for large language models because they read the surrounding text and notice when your brand appears across trusted sources.
- Keep entity names consistent. Decide whether you are saying “Google AI Overviews,” “AIOs” or “Google’s AI feature,” then use that term every time. Consistent entities help machines connect your brand to a topic.
- Lead sections with a standalone answer. A first sentence under roughly 40 words that works without surrounding context is far easier for an AI system to lift and cite.
- Use FAQ-style headings. A question heading followed by a direct answer matches how people prompt AI tools.
- Keep content genuinely fresh. Update the facts, examples and screenshots. Changing the publish date alone does not send a real recency signal.
Alex Birkett puts the backlink question bluntly in the guide: do backlinks still matter? “Yeah, a hundred percent.” He adds that LLM optimization is more text-based, so even unlinked mentions on credible sites help because the model reads the text around the prompt.
HubSpot’s State of Marketing Report is the source’s recurring example of this flywheel. It has generated thousands of backlinks, gets mentioned across trusted outlets, and helps HubSpot maintain both a high AI visibility score and millions of organic visits each month.
A monitoring workflow that keeps both honest
Treat SEO and AEO as one dashboard, not two rival projects. Track keyword rankings and AI prompt visibility side by side. Look for the keywords that matter on Google, then identify the prompts people actually ask ChatGPT, Perplexity or AI Mode. Watch whether AI platforms cite, recommend or ignore your brand for each prompt, and compare that with competitors.
The framework is simple: build trust with original research, structure for extraction, earn mentions on trusted sites and check both surfaces before declaring a page a winner.
Source: Semrush Blog
