The Decline of Search Traffic

Over the last couple of years, publishers, website owners, and e-commerce brands have been complaining that traffic from search is declining. Search is still one of the main channels for discovering brands and information — DataReportal (2025) confirms that — but numerous independent studies show that click-through rates (CTR) from search results are falling.

At first, social search was blamed. More and more people discover products and news directly on social media platforms without using search engines at all. Younger audiences especially turn to TikTok, Instagram, and Reddit to find recommendations and real experiences, bypassing Google entirely. And it’s not just social anymore — according to Harvard Business Review (June 2025), 58 % of consumers now use generative AI tools for product or service recommendations, up from 25 % in 2023.

Now there’s another player on the stage — AI (artificial inteligence). Tools like ChatGPT, Perplexity, and Google’s AI Overviews compile and summarise data directly from online content. There’s still no comprehensive research on how much these generative AI tools affect website traffic, but the market reaction says a lot. When Google introduced AI-generated summaries in May 2024, publishers could measure the change instantly — Digital Content Next reported notable traffic declines, and The Daily Mail publicly shared sharp drops in their search referrals Digiday 2025 November.

Main Channels for Online Brand Research, Datareportal 2025

Content as Data Source

If people are clicking less and less and don’t go to the source of the information — the news site, blog, or any other website — then what happens?

That’s the question more and more creators, publishers, and brands are starting to ask. When users get their answers from AI summaries or chat-style interfaces, they no longer need to visit the original source. Content still fuels the system — only now, it’s less read and more used.

In this emerging environment, articles, blogs, and websites are becoming training material and reference data for generative tools. The same pieces once written for human readers now increasingly serve as input for AI models that summarise, rephrase, or quote — without necessarily driving any traffic back.

It’s hard to tell what this means in the long run. But one thing is certain — content is shifting from being consumed by people to being consumed also by machines.

So this is the moment to talk about GEO — Generative Engine Optimisation.

‍

GEO, LLMO and the New Rules of Visibility

If SEO was about how search engines find and rank content, then GEO — Generative Engine Optimisation — is about how AI tools find and use it.

GEO focuses on making your content visible, understandable, and credible — not just to humans or Google’s crawler, but to the AI systems that now read, summarise, and reuse information. These engines don’t simply index pages anymore; they interpret meaning, context, and reliability.

In parallel, marketers and researchers are beginning to use another term — LLMO, or Large Language Model Optimisation. It’s essentially the same concept, viewed through the lens of training and prompting AI models such as ChatGPT, Perplexity, Gemini, and Claude.

While traditional SEO focused on improving clicks and rankings, GEO and LLMO focus on recognition and inclusion. Instead of competing for a top position on a search results page, brands are now competing for a place inside an AI-generated answer.

Some marketers call this space AI SEO or Answer Engine Optimisation, but the goal is the same: to help AI systems understand, trust, and represent your content accurately.

So how does it work?

Generative engines like ChatGPT, Perplexity, and Google AI Overviews rely on vast datasets to decide which content to summarise or cite. That selection depends on factors such as:

  • Authority: Is your website or brand trusted across the web?
  • Structure: Is your content clean, well-organised, and easy for machines to parse?
  • Context: Does it answer questions directly and concisely?
  • Attribution signals: Are your brand mentions, metadata, and schema markup consistent across platforms?

Unlike SEO, GEO doesn’t reward you with clicks — it rewards you with inclusion and citation. Visibility now happens in summaries, not search results.

This shift means content strategy is evolving from chasing rankings to earning references. It’s no longer just about who reads your article — it’s about whether AI uses your content when people ask questions.

‍

GEO in Practice — What Brands Can Start Doing Tomorrow

Most marketers don’t need a new acronym — they need a plan.
The truth is, GEO or LLMO isn’t something separate from what you’re already doing. It’s about making sure your content is not only searchable, but also usable by AI.

‍

Here’s where to start — things you could begin tomorrow:

1. Make your content easy for AI to understand

  • Write in clear, structured formats — short paragraphs, direct answers, bullet points, FAQs.
  • Use semantic headings (H2, H3) and descriptive titles so meaning is obvious.
  • Add schema markup (FAQPage, HowTo, Product, Article) — this helps both search engines and AI models understand context.

2. Build your brand authority

  • Ensure your business information (name, address, profiles) is consistent across the web.
  • Get credible mentions, links, and citations — AI models learn trust through repetition and authority.
  • Publish content that adds something original — unique data, insights, or experiences. Generative models prioritise distinctive value, not duplicates.

3. Optimise for “answerability”

  • Identify the questions your audience asks most and answer them directly on your pages.
  • Use question-based headings (“How does it work?”, “What’s the best option?”) — these are often picked up in summaries.
  • Keep one-sentence summaries or TL;DR sections at the top of key pages — AI models use these snippets.

4. Monitor where your brand shows up

  • Try Perplexity.ai, ChatGPT, or Google AI Overviews and search your own brand or topic.
  • Track whether your brand is being mentioned, cited, or missed — that’s your early Share of Model.
  • Note gaps where you’re invisible — those topics are your GEO opportunities.

5. Don’t abandon SEO — expand it

  • Keep technical SEO solid (site speed, crawlability, mobile-first).
  • GEO builds on SEO — it doesn’t replace it. Think of SEO as your foundation and GEO as your next layer.

‍

The Takeaway

You don’t have to reinvent your content strategy — just make it more readable, structured, and trustworthy for both humans and machines.
If SEO helped people find you, GEO helps AI understand you — and that’s quickly becoming just as important.

‍

‍

FAQ

What is GEO in marketing?
GEO, or Generative Engine Optimisation, helps your content appear and be cited in AI-generated summaries and chat answers.

Is GEO replacing SEO?
No. GEO builds on SEO — it extends visibility from search results to generative AI systems.

How can brands measure their GEO impact?
Monitor where your brand appears in AI tools like ChatGPT, Perplexity, or Google AI Overviews, and track brand mentions or citations — your emerging Share of Model.

Where should marketers start?
Start by structuring your content clearly, using schema markup, and ensuring brand consistency across the web — the foundations of both SEO and GEO.

‍

‍

Sources: