Picture the scene. It’s a week before your mother’s birthday, you need to find a gift, and you know she likes chocolates. In the past, you would have trekked to multiple shops or conducted tedious online searches, flitting through a marathon of browser tabs to find the perfect bars. Today, you can upload a list of chocolates you’ve previously given her to an AI assistant. Within seconds, it curates personalized recommendations with no overlap, checks local availability and pricing, and provides tailored advice like “Rare albino blanco cacao: intense raspberry-lime snap. She hasn’t tasted Peru yet.” This AI-driven commerce experience is rapidly becoming the new reality of shopping.
This shift represents a fundamental change in how companies and products are visible in the marketplace. If you’re not referenced by Large Language Models (LLMs) and the AI systems that power them, then you may as well not exist. The $80bn SEO industry is in the crosshairs of this disruption, and entire web architectures designed for humans and search engine crawlers will need to be refashioned to serve AI agents.
This change represents a clear threat to brands that aren’t able to convince LLMs that they have digital authority and legitimacy. Indeed, the risk of being left behind is very likely if organizations treat this purely as a marketing challenge rather than something that demands senior management attention.
The changes in web traffic clearly show how powerful AI referrals have become, with traffic from AI services growing sevenfold since 2024. Admittedly, this is from a low base, and whereas organic search still produces around half of global internet traffic, AI platforms produce just 0.15%. The pace of change is extremely rapid, however. For instance, in the US, Adobe reported that AI-driven traffic to retail websites jumped 12x between July 2024 and February 2025. What’s more, this trend is even stronger among heavier users of generative AI, where over a third have already started using AI to find things instead of traditional search. By July 2025, this momentum grew to a staggering 4,700% year-over-year increase in AI-driven retail traffic.
While we’re in the early stages of this transformation, it’s clear that we can’t bury our heads in the sand and continue with business as usual. Search Engine Optimization (SEO) and Pay Per Click advertising may have been the default way of attracting customers over the past two decades, but this $80bn industry is having the ground shaken beneath its feet. We’re entering what we might think of as Act II of search, where we replace SEO with Generative Engine Optimization (GEO). It’s an act in which success isn’t measured by where we appear on a search results page, but by whether AI systems cite, reference, and recommend your content.
“By contrast, GEO is built on language, and this distinction is a seismic shift in both how we look at digital marketing and, perhaps most importantly, how users behave.”
The so-called Page Rank system that underpinned traditional search was built on links, with the connections between websites used to gauge authority. By contrast, GEO is built on language, and this distinction is a seismic shift in both how we look at digital marketing and, perhaps most importantly, how users behave.
With traditional search, users type short keyword phrases, such as “single-origin chocolate award winners” or “chocolate bar shops near me,” and receive a list of links to click through. With AI assistants, the interaction is conversational: “I want to buy three new single-origin bars for my mother. She’s open to exploring new chocolates. This is my budget. What do you suggest?”
From a marketer’s perspective, there are four particular differences to be aware of. First, the goals are very different. Whereas with SEO, the aim is to rank as high as possible in the search engine results, with GEO, the aim is to be referenced in the AI-generated responses.
Second, the success metrics are very different. With SEO, marketers are concerned with things like page rankings, click-through rates, and organic traffic. With GEO, we’re more concerned with citation frequency, reference rates, and brand mentions.
There are also differences in how one goes about building authority. With SEO, the key was to secure backlinks from high-authority domains to ensure your Page Rank was as high as possible. In contrast, GEO prioritizes E-E-A-T signals: Experience, Expertise, Authoritativeness, and Trustworthiness.
Last, but not least, are the differences in the user journey. With SEO, there’s a clear pattern of clicking the search result, browsing your website, and hopefully converting. With GEO, the user may get the information they require without ever visiting your website, and this may even require a purchase being completed at arm’s length.
| Dimension | Traditional SEO | Generative Engine Optimization (GEO) |
|---|---|---|
| Primary goal | Rank high on search engine results pages (SERPs) to earn clicks | Be cited or referenced in AI-generated responses |
| Success metric | Click-through rates, page rankings, and organic traffic | Reference rates, citation frequency, brand mentions in AI responses, share of voice |
| Optimization focus | Keywords, backlinks, page speed, mobile-friendliness | Content clarity, authority signals, contextual relevance, structured data |
| Content strategy | Keyword-rich content optimized for crawlers | Natural language, comprehensive answers, authoritative sources |
| Authority building | Backlinks from high-authority domains, domain age | E-E-A-T signals: expertise, authoritativeness, trustworthiness, first-hand experience |
| Technical factors | Page speed, mobile optimization, meta tags, schema markup | Structured data, semantic clarity, extractable facts, schema markup |
| Content format | Blog posts, landing pages optimized for specific keywords | FAQ sections, direct answers, structured content, comparisons, statistics |
| Discoverability | Search engine crawlers index based on keywords and links | AI training data, real-time web retrieval, third-party citations |
| User journey | Click → Browse → Convert on your site | AI generates an answer → User may or may not visit the site → One-click buying options? |
Probably the most important of these differences is in the way in which SEO and GEO evaluate the quality and authority of content.
Probably the most important of these differences is in the way in which SEO and GEO evaluate the quality and authority of content. Traditional SEO leaned heavily on things like keyword density, backlink profiles, mobile readiness, and even page load speeds. In contrast, GEO prioritizes what researchers call “semantic authority”, which is when expertise is demonstrated through content depth, factual accuracy, and real-world application rather than link-based metrics.
Originally developed by Google as part of the company’s Search Quality Rater Guidelines, E-E-A-T now sits at the heart of GEO. Google says that trustworthiness is the key part of E-E-A-T, and while it was also a factor in traditional SEO, its role is even more important for GEO.
AI assistants surface products and content they:
Indeed, research using the GEO-BENCH benchmark demonstrates that GEO methods, such as the inclusion of citations, quotations from relevant sources, and statistics, notably boost source visibility by over 40% across various queries. The top-cited domains in ChatGPT in the US are Reddit, Wikipedia, Amazon, Forbes, and Business Insider. To achieve success in this GenAI environment, it’s vital that brands are able to appear in these kinds of authoritative contexts, as these are where AI models learn to trust the information.
Not all AI platforms work the same way, and understanding these differences is crucial for GEO strategy. The key distinction lies in whether the platform relies solely on training data or can access real-time web information.
Training data-only models: If an AI assistant works only off training data, the information it has access to has a fixed cut-off date. These models are limited to what they learned during training.
Real-time search models: ChatGPT and Claude search online only when the topic is too new or the user explicitly toggles “web search.” Perplexity is search-first by design, always going online first. Gemini AI Mode sits on top of Google Search and goes online quite often.
This distinction affects optimization strategy:
| Aspect | Training data models | Real-time search models |
|---|---|---|
| Goal | Be represented in training datasets | Be cited in real-time web retrieval and AI-generated summaries |
| Primary strategy | Historical presence in high-quality, authoritative sources | Current presence in authoritative, AI-crawled sources; product data feeds |
| Update frequency | Less important unless retraining occurs | Frequent updates maintain freshness; live feeds for products |
| Citation strategy | Inclusion in trusted, high-quality websites during training time | Ongoing mentions on authoritative sources; first-party product feeds |
| Crawlability | Less relevant for training data sourcing | Critical for visibility and parsing by AI-enhanced crawlers |
| Content style | Well-structured, high-authority content with clear answers | Structured, extractable, answer-style content; direct responses to questions |
Despite clearly being in the early stages of GEO, there is a clear initial interest among consumers.
Despite clearly being in the early stages of GEO, there is a clear initial interest among consumers. Data shows that 38% of US consumers were using AI for shopping in July 2025, with over half expected to use it by the end of the year. Retailers are rapidly responding to this trend with innovations of their own. Amazon has Rufus, Walmart has Sparky, China’s Taobao has Wenwen, and France’s Carrefour has a Hopla ChatGPT plugin. During Amazon’s 2025 Prime Day, AI showcased multiple use cases, from Rufus answering natural-language questions to AI-generated shopping guides comparing products.
There is also a lot of innovation in the transaction layer. In April 2025, ChatGPT rolled out shoppable product carousels and opened a merchant-feed program where merchants can directly feed live first-party information into the platform. It followed up five months later with its in-app Instant Checkout feature. Microsoft’s Copilot Merchant Program and Perplexity’s Buy with Pro feature offer similar capabilities.
So, what does this mean for leaders? It’s clear that the AI landscape is changing incredibly quickly, and there isn’t a clear and definitive playbook for GEO yet. Despite this, there are still steps leaders can take to position themselves for success.
The first step is to know where you currently stand. This isn’t straightforward, as LLMs identify themselves differently from traditional search engines. Begin by tracking referral traffic from ChatGPT, Perplexity, Gemini, and Claude. Run synthetic queries related to your products and services across multiple AI platforms to understand whether your brand appears in AI-generated responses.
There are a number of new tools to help you do this, with GEO monitoring services like Goodie, Profound, and Daydream available to help you systematize this process. Perhaps most importantly, investing in these tools will help to ensure that AI is treated as a marketing channel rather than some kind of experimental side project.
In a “zero-click” environment, organizational search success depends on how AI perceives your content and your brands. As a result, you need to show off your expertise via case studies and user stories. You should highlight your expertise and your credentials, both of your brand and of your key thought leaders. Strive for factual accuracy and transparency in all of your content, with author schema markup and structured data to make this information machine-readable.
For GEO to be effective, organizations must increase their thought leadership. White papers, research, and expert analysis are all things that AI models like, especially when they contain facts and statistics that AI can lift verbatim. You should also include social proof via ratings and testimonials to show AI not just what’s on offer but why it matters.
To succeed with GEO, it’s far more important what others say about you than what you say about yourself. When AI systems see your brand referenced in multiple authoritative contexts, they’re more likely to include you in responses. This means coordinating PR and marketing teams to build third-party citations systematically, get published in industry publications, be interviewed on podcasts, contribute to sites like Wikipedia, and participate in industry forums and communities.
There are a number of factors that go into ChatGPT and other LLMs selecting products, and as with SEO, these factors are under constant change. As such, you should test and experiment continuously to find what works and what kind of content is most successful. Benchmark against competitors regularly and make sure the AI perception of your brand aligns with your intended positioning. You should also track engagement metrics from AI referrals and iterate your strategy based on your findings.
Jason Forbes, Co-founder and CEO of AI brand optimization platform Xeo360, highlights five key objectives to track:
The shift from SEO to GEO represents the most significant change in digital marketing strategy since the rise of search engines themselves.
The $80bn SEO market is changing as we speak. The shift from SEO to GEO represents the most significant change in digital marketing strategy since the rise of search engines themselves. Organizations that recognize this transition early and optimize for AI-driven discovery will gain a significant competitive advantage.
The good news is that sites and organizations with genuine expertise can now go toe to toe with established players by showing they have authority and credibility. The core of E-E-A-T in GEO shows that authenticity, expertise, and trustworthiness matter more than ever.
To succeed in this new era requires us to become part storyteller and part data scientist, with compelling narratives backed by authoritative data that can be trusted and cited by the LLMs. This transformation is already underway, so the question isn’t whether to adapt, but when and how quickly.
History has provided us with numerous examples of fundamental changes in how customers find us, from the Yellow Pages to Google, from desktop to mobile. This history tells us that it’s not always the best products that win, but rather those who spot the mood music early and mobilize their resources decisively. The AI-driven search era is here, and it’s time to act. Will you lead this transformation, or will you explain to stakeholders why your brand has become invisible to the next generation of customers?
For multinationals, especially in strategically sensitive sectors such as energy, finance, and technology, traditional defense policies offer inspiration as they develop robust AI strategies. ...
To truly transform, organizations must overcome strategic and conceptual pitfalls. Here is why so many leaders continue to get it wrong. ...
CHROs must lead AI adoption in HR, from performance reviews to strategy, balancing risks and rewards while enhancing organizational impact....
Is your organization AI-ready? IMD professors and researchers predict 2026 trends in artificial intelligence. Read on and prepare to innovate – or be left behind. ...
Explore first person business intelligence from top minds curated for a global executive audience