More and more users now put their questions straight to ChatGPT, Gemini or Google's AI Overviews instead of typing short keywords the old way. The question for anyone running a website has shifted from "how do I rank on page one" to "how do I get AI to pull my site into its answers". This article answers that in hands-on terms, from writing content that answers directly and using schema markup to building your brand entity and using AI to help write SEO content the right way.
Does SEO still matter in the AI era?
It does, and in many ways it matters more than before, because AI search does not create answers out of thin air. These systems pull from web pages they can find and have judged trustworthy. AI Overviews is built on Google's index, while ChatGPT and Perplexity call up web search results whenever users ask about anything that needs current information. Pages with solid SEO fundamentals that rank near the top stand a far better chance of being chosen as a source than pages search engines cannot reach.
What has really changed is the scoreboard. Rankings and clicks used to be the whole game. Now you also have to count citations and mentions inside AI answers. Some users do get everything they need without ever clicking through, but a brand that keeps appearing in those answers becomes the first name people think of when it is time to buy. Neglecting SEO now means disappearing from both the search results page and the AI answer at the same time.
What is SEO for AI called?
The industry still uses several names, and they overlap quite a bit.
- AEO (Answer Engine Optimization) focuses on shaping content to be picked as "the answer", such as winning a featured snippet or being read aloud by a voice assistant.
- GEO (Generative Engine Optimization) focuses on getting your content cited inside AI-generated answers on AI Overviews, ChatGPT Search and Perplexity.
- LLMO (Large Language Model Optimization) takes the widest view: making sure language models know and understand your brand correctly, both from their training data and from real-time web search.
Different names, same core job: make your content as easy as possible for machines to understand, trust and reuse. If you want to go deeper into each approach, see Relevant Audience's AI SEO and GEO services.
How do you do SEO for AI?
The whole playbook comes down to six steps, in the order you should tackle them.
1. Put the answer in the first paragraph
AI favours content that answers a question in a self-contained way. Turn your headings into the questions people actually ask, then open the first paragraph with a direct answer in two or three sentences before expanding below. This helps readers in a hurry, and it lets models lift a passage into an answer without distorting its meaning.
2. Structure the page so machines can read it
Use H2 and H3 headings to break content into short sections, one idea per heading. Use bullet lists for steps or comparisons, and tables when data has several dimensions. Language models process content in chunks; a clearly structured page gets split correctly and cited more accurately than one long wall of text.
3. Add complete schema markup
Schema is the labelling layer that tells machines what a page is, who wrote it and when it was updated. At minimum you want Article, FAQPage, Organization, and Person for pages that show an author. This data helps search systems connect your content to your brand's identity faster, and it also underpins rich results on the search page.
4. Make your brand entity unmistakable
AI understands the world through entities, the identities of things, not just keywords. A brand with messy signals, such as a name spelled several ways, addresses that differ across platforms, or no About page explaining what it does, is hard for a model to pin down. Make sure your name, business description and contact details match everywhere: your own site, social media and your Google Business Profile.
5. Let AI crawlers reach your site
Check robots.txt to make sure you are not accidentally blocking AI bots such as GPTBot, ClaudeBot or PerplexityBot. If the bots cannot get in, your chance of being cited is zero. Alongside that, keep the classic technical basics in shape: an up-to-date sitemap, fast page loads, and main content that renders without leaning too heavily on client-side JavaScript.
6. Build mentions beyond your own site
AI models give weight to brands that are talked about across credible sources: media, review sites, question-and-answer threads and third-party articles. Digital PR, answering questions in relevant communities and earning references from other sites all feed your AI visibility directly, much the way backlinks feed your Google rankings.
The basic building blocks of SEO in the age of AI
None of the techniques above will move anything if the foundations are broken. Before chasing new tactics, get this checklist to pass first.
- Your site must index normally. Check Google Search Console to confirm target pages are not stuck behind noindex or a canonical pointing the wrong way.
- Speed and mobile experience. Slow pages lose out with users and with bots that have limited crawl time.
- E-E-A-T. Show real experience and expertise: author names, bios and the sources behind your information.
- Systematic internal links. Connect articles to the relevant service pages so machines understand what your site is an authority on.
- Content freshness. Keep figures, data and years in key articles up to date, because AI search tends to pick the fresher source.
Using AI for SEO writing, done properly
The other side of this topic is using AI as a content production assistant, which is workable and which plenty of professional teams already do. The issue is not whether to use it but how. Google has stated plainly that it does not penalise AI-generated content in itself; what it acts against is low-quality content mass-produced to manipulate rankings.
The safe pattern is to let AI help with research, outlining and a first draft, then have a person fact-check every claim, add perspective or data that cannot be found anywhere else, and edit the language into your brand's voice. Content copy-pasted wholesale from an AI answer tends to be flat, near-identical to hundreds of other sites, and gives no one a reason to cite it. And if your specific goal is getting your brand to appear inside ChatGPT, there is a dedicated approach on the ChatGPT SEO service page.
In short, SEO for AI is not about throwing out the old playbook and starting over. It builds on strong SEO with answer-first writing, machine-readable structure, schema, a clear brand entity and credibility beyond your own site. If you want a team that covers both worlds, Google rankings and AI visibility alike, Relevant Audience can map it out through our SEO services for businesses in Thailand.
