Blog

Staying Visible When Nobody Googles Anymore

25.08.2026

SEO and GEO for Radio Stations Explained

Sofia SEO, AI Assistant - SEO/GEO Expert for the Radio Creator AI-Tools

Imagine a listener asking ChatGPT: "Which radio station in my area has the best traffic updates?" The AI answers - with or without your station's name. Maybe even with a link, but to the wrong landing page.

That's exactly the reality radio stations, media houses and publishers now have to deal with. Because search as we've known it for the last twenty years is changing radically.

Why this topic is urgent right now

The numbers speak for themselves: zero-click searches - queries that end without a single click on a result - are increasing dramatically. As soon as an AI overview appears on Google, the click-through rate on the first organic result drops even further - studies report declines of between 34 and 61 percent, depending on methodology and query. In Germany, according to a Sistrix analysis from February 2026, 265 million clicks per month are lost directly to Google's AI overview.

The media industry across Europe is feeling this pressure too. How to deal with AI giants like Google and ChatGPT is one of the top topics at every media conference.

Some major media houses are already forging their own path: Germany's "Spiegel" has been partnering directly with the AI search engine Perplexity AI since 2024, making its own content more visible there and sharing in the advertising revenue.

Axel Springer, meanwhile, has partnered with OpenAI. But deals like these are only realistic for the very biggest players. For most radio stations and regional media houses, what really matters is this: your own website has to be set up so that AI systems can find, understand and cite it on their own.

SEO and GEO - two terms, one goal

Before we dive into detail, let's clear up the biggest source of confusion: what's actually the difference between SEO and GEO?

SEO (Search Engine Optimization) is probably already familiar to you: it's about getting your website as high as possible in the results list on Google and Bing.
GEO (Generative Engine Optimization) is the term for how well content is understood by AI systems such as ChatGPT, Perplexity, Google's AI Mode or Anthropic's Claude - and how often it gets cited in their generated answers. Instead of a spot on the link list, it's about a spot inside the answer itself.

GEO is not a brand-new, separate discipline that replaces SEO. Google itself makes this clear in its official guide to optimizing for generative AI features: from Google Search's point of view, optimizing for generative AI is simply optimizing for search overall - and therefore still classic SEO. The foundation for visibility in AI search is still the same technical and content-based SEO groundwork that classic Google Search has always relied on. (AI Optimization Guide)

So you don't need to learn an entirely new discipline - just apply the existing fundamentals consistently.

How does an AI search actually "think"?

To understand why some content gets cited and other content doesn't, it helps to look behind the scenes. In its May 2026 Bing blog post, Microsoft describes the fundamental difference between classic search and so-called "grounding" - the process of gathering information for AI answers: classic search asks which pages a human should visit. Grounding asks which pieces of information an AI system can use to construct an answer. In classic search, the smallest unit is the document page. In grounding, it's the individual, verifiable fact with a clear source. (https://blogs.bing.com/search/May-2026/Evolving-role-of-the-index-From-ranking-pages-to-supporting-answers)

Google explains how this works technically: through Retrieval-Augmented Generation (RAG), Google Search's AI features draw on the same core ranking systems used by classic search, retrieve relevant pages from the search index, and then check the specific information on those pages in order to generate an answer from it.

In very practical terms for you as a radio station: for the AI to even consider your content as a candidate for an answer, your page must first be technically discoverable and crawlable, and second, must offer a clearly recognizable, reliable fact or a unique statement that can be cleanly cited.

The building blocks of optimization

Sofia SEO is a new AI assistant within the Radio Creator AI Tools that can examine websites and generate reports from them. She breaks the analysis down into three levels that have to work together: technology, content, and visual presentation.

Technology. This includes the robots.txt file, where you determine whether Google and important AI crawlers such as GPTBot, ClaudeBot or PerplexityBot are even allowed to visit your site. On top of that comes a complete and up-to-date sitemap.xml, clean meta tags (title, description, canonical tag), structured data based on Schema.org that actually matches the page content, and the question of whether there's a critical gap between the raw HTML code and the content that only becomes visible after JavaScript runs - a gap that trips up simple crawlers. No need to be intimidated by all the technical jargon: Sofia SEO explains every analysis step and provides instructions for optimizing your files.

Sofia SEO: Sample Technical Check

Content. Does your page deliver a direct, understandable answer right at the start? Are headings phrased as real questions your users would actually ask? Is there a short, clear FAQ section? Are authors, sources and recency clearly identifiable - the famous E-E-A-T principle (Experience, Expertise, Authoritativeness, Trustworthiness)? Google itself stresses in its guide how important a unique perspective is: a first-hand review or report based on your own experience offers a completely different kind of value than simply summarizing information that already exists elsewhere. For a radio station, that means: an article about a traffic jam on the motorway, written by your own traffic reporter with local insider knowledge, has far better citation chances than a news brief that reads the same everywhere.
Visual presentation. Most users today are on mobile. Does the most important content show up on a smartphone screen without scrolling? Does the page's text jump around while ads load (layout shift)? A good user experience factors into the ranking on some search engines as well.

The llms.txt myth

In many forums, the llms.txt file is treated as an insider tip - a simple text file placed in the website's root directory that supposedly points AI systems to the most important content.

Google addressed this head-on in its official guide in May 2026, explicitly listing llms.txt among the things you can safely ignore: you don't need to create new machine-readable files, AI text files or Markdown files in order to appear in Google Search, since Google itself doesn't use them for that purpose. Independent server-log analyses back this up too: major AI search crawlers hardly ever request the file, while the very same bots regularly read robots.txt.

That doesn't mean llms.txt is completely useless - for specialized agentic use cases, where individual AI agents specifically crawl your website, it can serve as an orientation aid. But as a GEO strategy for getting more citations in ChatGPT, Perplexity or Google AI Overviews, current evidence suggests it has no effect. What matters far more is the robots.txt file: allow the relevant AI crawlers there, or you simply won't show up in AI answers at all.

The biggest levers and quick wins

So what actually matters? Google sums it up neatly in its guide: unique, valuable and non-generic content is likely to have a greater long-term impact on AI search visibility than all other recommendations combined. An article titled "7 Tips for the Perfect Playlist" is interchangeable general knowledge. An article titled "Why Our Music Editor Will Never Play This Song Again" delivers a unique, experience-based perspective - exactly what AI systems need in order to cite something.

Concrete quick wins for radio stations:

Check and fix your technical foundation: open up robots.txt for AI crawlers, keep your sitemap current, improve mobile load times.
Make your authors visible: name the real people, roles and experience behind your newsroom on every article - this strengthens E-E-A-T.
Lead with the answer: start articles and show pages with the core message in one or two sentences before going into detail.
Reduce duplicate content: if the same story appears almost identically on several sub-pages, it unnecessarily dilutes your visibility.
Check Search Console: the performance report for generative AI features shows you how your content actually performs within Google Search's AI features.

And equally important: what you don't need to do. Google explicitly debunks three more myths: you don't need to chop content into tiny fragments to make it easier for AI to understand. You don't need to write in some special "AI language." And you don't need excessive, special Schema.org markup just for AI search - structured data is still useful for rich results, but it isn't a requirement for being cited.

Write for your users - not for the search engine

A common misconception: writing LLM-friendly content doesn't mean writing for robots instead of people. It means structuring your content clearly enough that both listeners and machines can grasp it effortlessly. A few tried-and-tested tricks from newsroom practice:

Phrase subheadings as real questions your listeners would actually ask - for example "When does the new morning show start?" instead of "Schedule".
Add a short FAQ block with three to five concise question-and-answer pairs at the end of show or topic pages.
Use concrete numbers, dates and facts instead of vague phrasing - "on air since 1998" instead of "for many years now".
Make recency visible: a publication date and, where relevant, an update date shown directly on the article.

Grounding pages: extra pages just for AI?

An interesting, more advanced approach comes from SEO expert Hanns Kronenberg, SEO manager at Chefkoch.de, with his project Grounding Page. The basic idea: instead of hoping an AI will somehow piece together the right facts about your company from scattered sources, you create a deliberately clear, fact-based reference page - about your station, your brands, your most important concepts. These pages are explicitly not hidden technical files, but perfectly normal, human-readable pages that additionally carry machine-readable structure (via Schema.org and JSON-LD, for example).

The thinking behind it: classic SEO optimizes documents for keywords. AI SEO under this model optimizes entities - clear, unambiguous definitions of who you are, what you do, and what sets you apart from others - for stable, accurate mentions in AI answers. For a radio station, this could be a well-maintained "About Us" page that states unmistakably: broadcast area, frequencies, founding year, music format, reach, presenter line-up - facts an AI can actually work with instead of guessing or hallucinating.

Important context: the Grounding Page approach itself is described as an open framework that doesn't require any additional infrastructure - the structure can just as easily be applied to an existing "About Us" page. It isn't an officially confirmed ranking factor from Google, OpenAI or Perplexity, but a plausible, increasingly discussed practice that draws on several independent studies of how RAG systems behave. Still, for media houses with multiple brands, shows or locations, the core idea is worth adopting: clear, unambiguous fact pages never hurt - whether you call it a "Grounding Page" or simply a well-made company page.

How Sofia SEO supports you

Sofia SEO, the new AI assistant within the Radio Creator AI Tools, takes care of both areas at once: classic SEO - how well your site is found on Google & co. - and GEO, how well your content gets cited in AI search systems like Google AI Overviews, ChatGPT, Perplexity, Claude, Gemini or Copilot.

Sofia doesn't just hand you a raw checklist of data - she puts things into context. Precisely because the SEO and GEO space is full of half-truths and outdated tips - see llms.txt - Sofia tells you what's actually important, what's only a minor bonus factor, and what's a myth.

What Sofia can specifically do for you:

Quick analysis for targeted questions like "Is our homepage readable for AI bots?" or "Is important text getting cut off on mobile?"
Full report for an entire page or website section: technical, content, visual and GEO findings, prioritized at the end by effort and impact - available as a chart or as a ready-made PDF.
AI visibility check: give Sofia a search term and your domain, and she'll run an AI search across OpenAI, Anthropic, Google and Perplexity to check whether and how you're mentioned there. Good to know: this is a spot check, not a fixed ranking position - it's best repeated regularly with slightly varied phrasing.
Matomo integration: if Matomo is set up on your site, Sofia can even show you what share of your traffic comes from AI assistants rather than classic search engines - your most important success metric for GEO.

Example of a visibility check

Run an AI visibility check for the search term "AI tools for radio stations". 
Check whether our website "radio-creator.com" is mentioned and cited.

Prompts to Try

Check whether our homepage is readable for ChatGPT and Perplexity. 

Compare the mobile and desktop views of our show page - is important content getting cut off there? 

Create a complete SEO and GEO report for our news category. 

Show me how much traffic is coming from AI assistants and how that has developed over the last three months. 

Which of our article pages have a high bounce rate, and why might that be?

Now is the time to act

Search is being rebuilt right now - from a list of results to a direct answer. Whoever keeps their technical foundations clean, makes real expertise visible, and ignores the myths, gains a real advantage. The effort required is manageable once you know where to start - and that's exactly what Sofia SEO helps you do.