
Artificial Intelligence Optimisation: AIO vs SEO Explained
Updated November 2026. We first published this guide in July 2025. Since then, the evidence on what earns AI citations has firmed up, and some of our earlier advice, llms.txt in particular, didn't hold up. This version reflects what we actually do now. If you're newer to the topic, start with Your website isn't broken. It's just being found differently now.
AIO, AEO and GEO all describe the same goal: getting your website found, understood and cited by AI tools like ChatGPT, Perplexity, Claude and Google's AI Overviews. It builds on SEO rather than replacing it. The work that measurably helps is unglamorous: letting AI crawlers in, keeping key pages current, giving them specific evidence worth quoting, and making sure your structured data is accurate.
AIO, AEO and GEO: what the terms mean
The industry hasn't settled on one name, so you'll see three used interchangeably:
- AIO (Artificial Intelligence Optimisation) - the broadest term: preparing a website for AI-driven discovery.
- AEO (Answer Engine Optimisation) - getting your content used and cited in AI-generated answers.
- GEO (Generative Engine Optimisation) - the academic term, from the research paper that first measured what works.
At Dot The i Studio we use AEO day to day, and call the service AI Search Readiness. The label matters far less than the work.
How AEO relates to SEO
For Google, the answer is simple. Google's Gary Illyes has said that normal SEO is what gets content into AI Overviews, because AI Overviews and AI Mode draw on the same index as regular search (Search Engine Land, July 2025). If your SEO foundations are weak, fix those first.
ChatGPT, Perplexity and Claude are different. They run their own crawlers, build their own indexes, and choose which handful of sources to cite in each answer. That's where the AEO-specific work comes in.
What actually moves AI citations
This is the order we work in on every AI Search Readiness audit, because it's roughly the order of impact.
1. Make sure AI crawlers can reach your site
Every major AI provider now runs separate crawlers for training and for search. OpenAI, for example, uses GPTBot for model training and OAI-SearchBot for ChatGPT search, and its crawler documentation is explicit that OAI-SearchBot is the one to allow if you want to appear in search results. Anthropic (ClaudeBot and Claude-SearchBot) and Perplexity (PerplexityBot) follow similar patterns.
The common mistake is blocking a training bot, or a whole category of "AI bots", and accidentally removing the site from AI search altogether. The second is a CDN or firewall rule that blocks these crawlers without ever showing up in robots.txt. If an engine can't fetch your page, nothing else on this list matters.

2. Keep your important pages genuinely current
AI tools lean towards recently updated sources. A refresh needs to be real: new information, updated examples, a revised section. Changing the date on an unchanged page doesn't count, and the date in your structured data should match when the content actually changed. For pages you want cited, a quarterly review is a sensible rhythm.
3. Give each page something specific worth citing
The clearest evidence here comes from the original GEO research by Princeton, Georgia Tech and IIT Delhi, presented at KDD 2024 (Aggarwal et al.). Across 10,000 test queries, adding statistics, citing credible sources and including relevant quotations lifted a page's visibility in AI-generated answers by up to 40%. Keyword stuffing, the old SEO reflex, did little or nothing.
In practice, that means replacing vague claims with real numbers, linking to the sources behind them, and attributing expertise to named people. It also helps to name your business right next to your claims and results. AI tools often paraphrase without crediting the source, and a brand name sitting beside the fact gives them something to carry through.
4. Make sure your structured data tells the truth
Schema markup helps engines understand who you are, what you offer and where you operate. Its value now is understanding and attribution rather than fancy search results, which makes accuracy the whole point. The errors we see most often: duplicate organisation entries, out-of-date addresses and map coordinates, relative URLs, and leftover placeholder values. We found and fixed several of these on our own site this year.
5. Then the fundamentals
Clear page titles, a logical heading structure, sensible internal links and genuine FAQs still matter for people and machines alike. Write the whole paragraph well, not just the first sentence: AI tools read the full passage and extract from wherever the substance is.
What we've stopped recommending
llms.txt as a visibility tactic. In the first version of this guide we described llms.txt as "the robots.txt for AI". That was wrong. It doesn't control crawling (robots.txt does that), Google has said it doesn't use it, and none of the major AI search providers has said it uses the file to decide what to cite. We still keep one on our own site and set one up for clients, because it's cheap and may prove useful for AI agents. But it isn't a reason anyone will cite you, and we won't sell it as one.
Formula-driven "answer-first" blocks. Short, clear answers still help readers, but padding every section with a 40-word summary followed by filler doesn't serve anyone. Specific, well-sourced content does.
A five-step checklist for your site
- Check crawler access. Confirm OAI-SearchBot, Claude-SearchBot, PerplexityBot and Googlebot are allowed, and make a deliberate decision about training bots. Check any CDN or firewall too.
- Refresh your priority pages. Start with the three to five pages that matter most, and add genuinely new information.
- Add evidence. Real figures, linked sources, named people, and your business name next to your results.
- Audit your schema. Look for duplicates, wrong addresses, relative URLs and placeholders.
- Test it. Ask ChatGPT, Perplexity and Google's AI Mode the questions your customers ask. Note who gets cited and whether your business is described accurately. Repeat every quarter.

How to measure progress
- Citations: whether your site appears as a source for the questions you care about, and how often.
- Accuracy: whether AI tools describe your business, services and location correctly.
- Referral traffic: visits from chatgpt.com, perplexity.ai and similar sources in your analytics.
- Classic search metrics: impressions, clicks and rankings, which still feed Google's AI features.
The short version
- AIO, AEO and GEO are the same goal with different labels. AEO builds on SEO.
- Check crawler access first. It's the cheapest fix with the biggest consequences.
- Fresh, specific, well-sourced pages are what get cited.
- Accurate schema helps engines understand and credit you.
- llms.txt is fine to have, but it won't get you cited.
Common questions
What’s the difference between AIO, AEO and GEO?
They're three names for the same goal: getting a website found, understood and cited by AI tools. AIO is the broadest term, AEO focuses on being cited in AI-generated answers, and GEO is the academic term from the research that first measured what works.
Which term should I use day-to-day?
Whichever your audience understands. We use AEO, and call our service AI Search Readiness. The label matters far less than the work: crawler access, fresh content, specific evidence and accurate structured data.
Does AIO replace traditional SEO?
No. AEO builds on SEO. Google has said normal SEO is what gets content into AI Overviews, and strong foundations help with ChatGPT, Perplexity and Claude too. AEO adds crawler access, freshness and citable evidence on top.
Do I need to rewrite my existing content for AIO?
Usually not. Most of the gains come from refreshing your most important pages with genuinely new information, adding specific figures and sources, and fixing structured data errors. A focused audit tells you which pages are worth the effort.
What is llms.txt and does my site need one?
llms.txt is a proposed file that summarises a website for AI tools. It doesn't control crawling, Google has said it doesn't use it, and no major AI search provider has said it uses the file to choose sources. It's cheap to maintain and may help AI agents, so we keep one, but it won't get you cited.
How do I check if my content appears in AI search results?
Ask ChatGPT, Perplexity and Google's AI Mode the questions your customers ask, and note which sources are cited and how your business is described. Check your analytics for referral visits from AI tools. Repeat quarterly, since citations change as engines refresh.
Should I be worried about AI tools using my content without credit?
It's a fair concern, and you can opt out of AI training by blocking training crawlers such as GPTBot and ClaudeBot. Blocking their search crawlers is different: it removes you from AI answers entirely. Most businesses are better off allowing search crawlers and naming their brand alongside their key claims.
How does AIO relate to local SEO?
Local SEO and AIO reinforce each other more than people realise. AI tools weigh signals like location, business name consistency, Google Business Profile accuracy and structured data when answering location-based queries. Strong local SEO foundations make a business more likely to be described and cited correctly in AI answers, and AIO improvements feed back into traditional local rankings.
Want your business showing up in AI answers?
We audit sites in the order that matters - crawler access, freshness, evidence, then structured data - fix what's worth fixing, and tell you plainly what isn't. See how our AI Search Readiness work runs, or start a conversation.



