Back to blog
    AI SEO8 min read2026-05-17

    How to Optimize Your Website for AI Search (ChatGPT, Perplexity, Gemini)

    More and more customers ask ChatGPT and Perplexity instead of Googling. Here are 7 concrete steps to get your website cited by AI search engines in 2026.

    How to Optimize Your Website for AI Search (ChatGPT, Perplexity, Gemini)

    Over the last 12 months a lot of people quietly stopped Googling and started asking ChatGPT instead. That shift is going to change how SEO works over the next few years.

    Ask ChatGPT "what is the best marketing agency in Aarhus?" and you do not get 10 blue links the way Google gives you. You get one answer with a handful of cited sources next to it. If your business is one of those sources, you have a shot at the client. If it is not, you are invisible. Most Danish businesses have not noticed this is happening yet.

    What is AI search optimization (GEO, AEO, LLM-O)?

    You will run into three terms for roughly the same thing. GEO, or generative engine optimization, means writing content so it can become a source for generative AI like ChatGPT, Gemini, and Claude. AEO, answer engine optimization, is the broader category and covers Google AI Overviews, Bing Copilot, Perplexity, and You.com. LLM-O, large language model optimization, is the newest label and focuses on structuring content so an LLM can reliably extract and cite it.

    The labels overlap. What they ask for is the same: your content needs to be structured, factual, and easy for a model to quote.

    Why now is a good time for Danish businesses

    ChatGPT already sits in the top 10 most-visited websites in Denmark, and Perplexity and Gemini keep climbing. A business that prepares its site for AI search now buys itself a one-to-two-year head start before this becomes the default.

    There is almost no competition for GEO and AEO terms in Danish right now. That means it takes less work to rank, you can lock in first-mover citation status for common Danish business questions, and you are not fighting a crowded field for those citations.

    7 steps to optimize for AI search

    1. Let AI crawlers in

    Start with your robots.txt. AI search engines run their own bots, and those bots need to be able to fetch your content:

    User-agent: GPTBot
    Allow: /
    
    User-agent: ClaudeBot
    Allow: /
    
    User-agent: PerplexityBot
    Allow: /
    
    User-agent: Google-Extended
    Allow: /
    
    User-agent: Applebot-Extended
    Allow: /
    

    If you use Cloudflare, check Security, Bots, Block AI bots. When that toggle is on, it overrides everything above at the edge no matter what your robots.txt says. It defaults to on for new Cloudflare zones, so turn it off unless you actually want to block AI search.

    2. Add an llms.txt

    llms.txt is a new standard aimed at AI search engines. It is a Markdown file at the root of your domain that describes who you are, what you offer, and where the important pages live. Think of it as robots.txt for content rather than crawl rules.

    You can see ours at opus-studio.dk/llms.txt. It has a short intro, the list of services with links, a portfolio with client URLs, all the main pages in English and Danish, and contact details. An LLM indexing your site reads this file as a condensed brief, which saves it time and raises the odds it cites you accurately instead of guessing wrong about your business.

    3. Schema markup (LocalBusiness, Service, Article)

    AI models are trained on structured data, so Schema.org markup makes it easier for them to work out who you are and what you sell. The types that matter most here are LocalBusiness (address, phone, hours, coordinates, reviews), Service paired with Offer (your exact prices and service categories, like the pricing page where every package carries Offer schema with a DKK price), Organization (founder plus sameAs links to your social profiles), Article or BlogPosting for blog content, and BreadcrumbList for page hierarchy.

    Ask ChatGPT "how much does a website cost in Aarhus" and it will reach for a page that lists explicit prices in Schema markup before it reaches for one that just says "contact us for a quote". Marking up your data turns the page into a source the model can quote, not just a document it skims.

    4. Write content as questions and answers

    LLMs like to cite short, clear answers to specific questions. Build your key pages with FAQ sections where the question and answer sit right next to the structured data.

    Bad: "Our agile methods enable fast delivery at high quality." Good: "How long does a landing page take to build? A standard landing page is finished in 1 to 2 days."

    Give them numbers, dates, and direct answers, and skip the marketing fluff. A model cannot cite a claim it does not understand, and it does not understand a general statement with nothing concrete behind it.

    5. Pre-render the page (SSG or SSR)

    If your site is a pure React or Vue SPA with no server-side rendering, an AI bot sees an empty HTML shell and has nothing to cite. A growing number of models do not run JavaScript when they collect data, so anything that depends on client-side hydration is effectively invisible to them.

    We use vite-react-ssg, which pre-renders every route to a static HTML file at build time. ChatGPT, Perplexity, and Google-Extended get the full content on the first request, with no wait for React to hydrate. You get the same effect with Astro, Next.js (App Router with SSG), SvelteKit, and the other modern frameworks.

    6. Authority signals (About, sameAs, reviews)

    LLMs lean toward sources that have a clear author and organization behind them. A few things help here: an About page with real detail about the founder and team, consistent sameAs links in your Organization schema (LinkedIn, Instagram, Google Business), LocalBusiness schema with address, phone, and geo coordinates, reviews tied to real names on Google Business Profile or Clutch, and mentions on other credible domains through guest posts, interviews, or podcast appearances. The more evidence there is that you are a real, local business, the more willing an AI is to name you when someone asks a local question.

    7. Cite concrete numbers and dates

    AI models reward verifiable facts. Instead of "we build fast websites", write "our sites load in under 1.5 seconds on Slow 4G (measured with Lighthouse)". Instead of "experienced team", write "in business since 2023, 15+ projects delivered".

    This is what GEO comes down to: a specific number is worth citing, an adjective is not. A model will pick the sentence with the figure in it almost every time.

    How to check if you are being cited

    The simplest test costs nothing. Open ChatGPT, Perplexity, and Google AI Overviews and ask things like "What are the best web agencies in Aarhus?", "How much does a website cost in Denmark in 2026?", and "Who is [your business name]?".

    If your name never comes up, at least you know where you stand. Being mentioned is good, and having a link to your content cited is better still. The three tools surface citations differently: Perplexity uses numbered references, ChatGPT in web search mode lists its sources at the bottom of the answer, and Google AI Overviews shows them in a carousel below the answer.

    Run this check once a month for 5-10 key questions in your niche. It is the AI version of reading a GSC report.

    What NOT to do

    A few things actively hurt you here. Keyword stuffing is one: models detect it and quietly downweight your authority, so write for people rather than for the bot. Hidden text and cloaking carry the same cost and add risk from the classic search engines on top. Generic corporate language like "our expertise and synergy" is poison for citability, because there is nothing in it to quote. Hallucination-bait is another trap, so do not claim you are "the best agency in the world" if you cannot back it up, or an AI will either skip you or cite you wrong. And do not block AI bots by accident. Block them if you genuinely do not want to be cited, but let them in if you want to show up in AI answers.

    What now

    AI search optimization does not replace classic SEO, it extends it. Everything good SEO already does, load speed, schema, solid content, real authority, works for AI too. What changes is the emphasis on concreteness: AI wants facts, not marketing.

    If you want help putting this in place, take a look at our SEO and AI Search package or get in touch at opus-studio.dk/contact. Our Pro tier at 1,125 DKK/mo covers GEO and AEO optimization, schema markup for AI models, and monthly citation monitoring across ChatGPT, Perplexity, and Google AI Overviews.