The Age of Generative Answers
GEO, AAO, AEO… How AI is reshaping online search and what business leaders should be doing right now
Image by Growtika, Unsplash
Did you ask ChatGPT for Christmas gift ideas yet? Or tried an AI shopping assistant to find the best deals on Black Friday? What was a novelty a year ago is quickly developing and integrating consumers’ behaviors.
What is happening?
Search engines are increasingly answering queries with AI overviews while Large Language Models (LLMs) attract massive user bases. ChatGPT alone now serves 800 million users every week. This is creating a shift in web traffic. According to Adobe, AI traffic to US retail websites jumped 4,700% year over year in July 2025.
And this surge is not just experimental, consumers seem to be truly changing the way they search for information. Already 44% of AI-powered search users say it is now their primary source of insight, ranking before traditional search engines, retailer websites and review platforms. As consumers migrate from traditional engines to conversation-based models, the rules of discovery are changing.
How it works and what it means. From SEO to GEO
When searching on a traditional search engine, we entered a few keywords, scrolled down to select which websites we will visit and often did from the first result page. Traditional search engines, like Google, function essentially as a massive, automated library system. They use software programs (often called “robots” or “crawlers”) to constantly scan the internet and file copies of web pages into a database known as an “index”. When you type a query, the system doesn’t “read” or “understand” the web in a human sense; instead, it instantly scans its index to match your keywords against the pages it has stored. It then ranks these results using hundreds of factors, such as your location or the page’s relevance, to provide you with a list of links where you can likely find the answer yourself. In order to be visible and well indexed, a set of techniques and strategies to perform on traditional search engines developed: Search Engine Optimization (SEO), and rewarded elements like keyword matching and backlink structures.
But LLMs don’t follow the same rules. LLMs use a process known as Retrieval Augmented Generation (RAG), where the model first reformulates your query into a set of simpler sub-queries (easier for the web to process) and then retrieves sources from the web on a traditional search engine. The retrieved documents are then summarized and fed to a response-generating model.
Source: Digital Disruption Chair
“ChatGPT sources 89.4% of its information on Google results pages after the #20 rank”
But when the sub-queries are used to search the web and retrieve information, what factors are influencing the selection of sources and retrieval of information? In a study of June 2025 made by Semrush (SEO tools company) and Statista, we see that AI systems reference sources very differently, with Reddit.com ranking as “top domain cited by LLMs like Chat GPT and Perplexity” (40%), followed by Wikipedia.org (26%) and Youtube.com (23.5%). Another study found that while SEO goal is all about being on Google first page of results, with only 0.63% of Google searchers clicking on a result from the second page, ChatGPT sources 89.4% of its information on pages after the #20 rank.
For products, AI systems do not interpret product pages like humans. They scan for structured attributes, explicit benefits, normalized data, and semantic clarity. Without these elements, products become harder for AI to evaluate and may be categorized incorrectly or excluded from consideration altogether. The credibility, format and structure of the source content heavily influence whether it will be selected and cited by the AI systems.
An erosion of organic traffic is underway
As search online is changing, the rules to be visible and indexed are as well. As seen with the previous figures, search through an AI intermediary is entering consumer’s habits quickly and thus, SEO doesn’t disappear, but is increasingly insufficient as these new ways of searching expand. The most immediate consequence is the erosion of organic traffic.
As generative engines collect, synthesize and deliver complete responses directly to users, click through declines. Gartner expects traditional search volume to drop by 25% by 2026, a heavy loss of traffic for unprepared brands. Even with a perfect SEO strategy, organizations should now also focus on AI visibility, with content structured for machine readability.
Being visible to AI systems: A 3-step playbook
That is why a new discipline is growing, focused on AI systems. GEO, AEO, AAO… The terminology is not quite clear yet and varies, but the most common acronyms are the following:
GEO: Generative Engine Optimization
AEO: Answer Engine Optimization
AAO: AI Agent Optimization
The main objective here is ensuring that the content is not simply crawled but readable, understood, and selected by AI systems. Which requires a different playbook.
First, fresh proprietary and structured content. AI models are hungry for facts, statistics and quotes. Specific factual content can boost visibility up to 40% on certain queries. The content is also preferred when structured in bullet lists, FAQs, tables, than walls of texts, and regularly updated, as they want to provide recent information. You should demonstrate E-E-A-T, which stands for Expertise, Experience, Authoritativeness, and Trustworthiness, by using a persuasive and authoritative tone and integrating technical terms that can help signal relevance to the system. Finally, mirroring how people ask questions in a conversational tone, being very specific, in a well-designed FAQ or with content structured with titles in the form of a question (hence Reddit’s popularity in LLMs sources).
Second, machine readability. Implementing labeled content that machines can understand with schema markup for essential elements such as FAQs, product pricing, reviews and addresses. This categorization increases the likelihood of being pulled into rich snippets. The content matters but the data behind should also be rich in semantic descriptions and detailed attribute taxonomies. The more context is provided, the better the AI can recommend your product.
Third, third-party authority. AI platforms pull information from a diverse array of sources beyond just your owned website, which means visibility depends heavily on your brand’s reputation and presence across the web. It means being present and careful about your brand’s image on AI trusted sources like Wikipedia, leveraging user-generated content on platforms such as Reddit, YouTube, and Facebook, and being cited in reputable media or trade publications which generates positive co-occurrences.
The disruptive rise of commercial agentic AI: from B2C to B2Bots
In April 2025, Amazon launched a “Buy for me” beta feature on its app, enabling autonomous purchase on the users’ behalf on another platform if the product isn’t available on Amazon. In December 2025, AWS made a press release announcing its collaboration with VISA to “enable next-generation agentic commerce”. And just a month after, Amazon sued Perplexity over their agentic shopping tools that were browsing its website. All of these recent developments clearly signal a race and competition to enable agentic commerce, and keep a maximum of control over it. Although there are also signs of collaboration between actors, with the formation of the Agentic AI Foundation (AAIF) on December 9, co-funded by key players of the sector: The Linux Foundation, OpenAI, Anthropic and Block, and with the support of Google, Microsoft, AWS, Bloomberg, and Cloudflare. The objective is to create an open, neutral foundation for the future of AI by unifying open standards for AI agents to develop in an open and interoperable way rather than being locked into proprietary platforms.
And that comes with no surprise, as the rise of agentic AI amplifies the stakes. AI agents browse, compare, negotiate and initiate purchases, acting as an economic proxy for the user, who simply has to express his intent. All could be done without the user having to visit a single webpage. This introduces a powerful layer between merchants and customers, and creates a new balance of power and influence, favors objective value over brand recognition, compares through countless websites and every feature… redefining the entire customer journey. In Reimagining Discoverability: How Generative Engines Bring the Web to You, the BCG explains: “For brands, this means designing content not just for humans but for bots and agents”.
And this is the disruption, the agent being the new customer. And businesses having to adapt to appeal to a hyper rational AI agent. When agents navigate the web instead of humans, they make data harvesting much less valuable and weaken the mechanisms through which platforms have historically built the “shadow profile” used to target and influence consumers. They become the new gatekeeper for user data and attention. Agents collapse search, comparison, and checkout into a single intent-driven workflow. This shifts advantage away from platforms optimized for stickiness and toward systems optimized for interoperability and product-performance.
In the report The end of inertia: Agentic AI’s disruption of retail and SME banking, McKinsey highlights how agentic AI pressures traditional business models: In retail, agentic flows may bypass ads and platforms, threatening retail media networks. In banking, agents undermine inertia-based revenue pools by reallocating deposits and optimizing card usage, reducing spreads and interchange fees, “AI agents won’t care about brand loyalty. They will optimize for outcomes”. Disrupting in many ways current business models.
Why now?
During the 14th Asia Privacy Bridge (APB) Forum 2025, Prof. Jan Ondrus highlighted a convergence of three conditions that make agentic commerce possible today:
Powerful AI models able to interpret intent, break goals into multistep tasks, and act autonomously.
Seamless, instant transactions are increasingly possible through digital wallets and emerging payment protocols, enabling secure automated purchases.
Rich personal data access, allowing agents to build persistent profiles of user values, constraints, and preferences.
Source: Digital Disruption Chair
But the next question is: while agentic commerce is possible, are consumers ready? It’s a question that largely remains unanswered. Even if users are showing a growing appetite for AI tools, agentic commerce goes a step further, especially when it comes to the customer’s experience and to trust, and might not be appealing for every service or product the same way. In a surprising twist, the Strait Times notes that “as AI upends online shopping, physical stores may gain renewed importance”, creating an opportunity for brand marketers to strengthen in-store customer experiences and build brand loyalty outside of the AI flow.
Are we going to see a fresh impulse for physical stores? Time will tell. In a largely digitalized economy, SEO remains foundational, but is from now on just one dimension of a broader challenge: business leaders need to focus on machine-readable credibility across the entire web ecosystem. Visibility will increasingly depend on structured data, third-party validation, semantic clarity, and the ability to demonstrate value in ways that agents can objectively measure and compare. Even if the day is young for agentic commerce, its disruptive power when reaching mainstream usage should encourage you to prepare for a future where your most important customer might be an AI agent.
This week’s curated news:
Industry Leaders Unite to Launch the Next Era of Agentic AI
The Linux Foundation created the Agentic AI Foundation to consolidate Anthropic’s MCP, Block’s Goose framework, and OpenAI’s AGENTS.md into a shared, vendor-neutral standard for interoperable and reliable agentic AI systems.
Read the full news here.
HSBC Partners with Mistral to Scale GenAI Across Operations
HSBC will self-host Mistral’s frontier models to enhance complex document analysis, multilingual operations and client communications, reinforcing finance as one of the fastest-moving sectors in AI adoption.
Read the full news here.
Netflix vs Paramount Clash Intensifies over Warner Bros Discovery
Netflix’s $82.7bn bid for WBD faces a $108bn hostile counter-offer from Paramount, triggering a high-stakes battle involving antitrust risks, Hollywood backlash, Middle Eastern financing and potential White House influence as both sides prepare for a prolonged bidding war.
Read the full news here.
🔍 Dive Deeper into Digital Disruption
Explore the 2025 Digital Disruption Matrix → Your guide to navigating digital transformation with confidence. This annual barometer combines rigorous data analysis with human perspective to rank this year’s most disruptive technologies. Discover how blockchain, AI, Web3, and emerging innovations are reshaping industries.
💡 Never Miss an Insight
To get weekly analysis on emerging technologies and digital transformation delivered to your inbox, follow us on Substack.
Follow the Digital Disruption Chair on LinkedIn for regular insights, interviews, comments and articles on frontier technologies.
Learn more about our programs | Contact us
Bibliography:
Agentic AI Foundation. (n.d.). Linux Foundation Announces the Formation of the Agentic AI Foundation (AAIF), Anchored by New Project Contributions Including Model Context Protocol (MCP), goose and AGENTS.md – Agentic AI Foundation (AAIF). Retrieved December 10, 2025, from https://aaif.io/press/linux-foundation-announces-the-formation-of-the-agentic-ai-foundation-aaif-anchored-by-new-project-contributions-including-model-context-protocol-mcp-goose-and-agents-md/
Aggarwal, P., Murahari, V., Rajpurohit, T., Kalyan, A., Narasimhan, K., & Deshpande, A. (2024). GEO: Generative Engine Optimization (No. arXiv:2311.09735). arXiv. https://doi.org/10.48550/arXiv.2311.09735
Amazon sues Perplexity over “agentic” shopping tool. (2025, November 5). Reuters. https://www.reuters.com/business/retail-consumer/perplexity-receives-legal-threat-amazon-over-agentic-ai-shopping-tool-2025-11-04/
Amble, Z. C., Seema. (2025, May 28). How Generative Engine Optimization (GEO) Rewrites the Rules of Search. Andreessen Horowitz. https://a16z.com/geo-over-seo/
Binder, A. (n.d.). Generative Engine Optimization (GEO): The Future Of Search Is Here. Forbes. Retrieved December 9, 2025, from https://www.forbes.com/councils/forbesagencycouncil/2025/01/02/generative-engine-optimization-geo-the-future-of-search-is-here/
Dean, B. (2019, August 27). We Analyzed 4 Million Google Search Results. Here’s What We Learned About Organic CTR. Backlinko. https://backlinko.com/google-ctr-stats
Derow, R., & Robnett. (2025a, April 10). The Future of Discoverability. BCG X. https://www.bcg.com/x/the-multiplier/the-future-of-discoverability
Derow, R., & Robnett. (2025b, July 28). Reimagining Discoverability: How Generative Engines Bring the Web to You. BCG X. https://www.bcg.com/x/the-multiplier/how-generative-engines-bring-web-to-you
Ducoudray, M. (2025, June 24). Qu’est-ce que le Generative Engine Optimization (GEO) ? https://www.eskimoz.fr/geo/
Gartner Predicts Search Engine Volume Will Drop 25% by 2026, Due to AI Chatbots and Other Virtual Agents. (2024, February 19). Gartner. https://www.gartner.com/en/newsroom/press-releases/2024-02-19-gartner-predicts-search-engine-volume-will-drop-25-percent-by-2026-due-to-ai-chatbots-and-other-virtual-agents
GEO (Generative Engine Optimization): Allié ou remplaçant du SEO ? (2025, November 26). https://blog.hubspot.fr/marketing/geo-generative-engine-optimization
Girardeau, M. (2025, May 15). Agent commerce: Quand l’IA redéfinit l’acte d’achat. Capgemini France. https://www.capgemini.com/fr-fr/perspectives/blog/agent-commerce-ia-redefinit-acte-dachat/
Guide détaillé de la recherche Google | Google Search Central | Documentation. (n.d.). Google for Developers. Retrieved December 9, 2025, from https://developers.google.com/search/docs/fundamentals/how-search-works?hl=fr
How AI is disrupting shopping. (2025, December 9). The Straits Times. https://www.straitstimes.com/opinion/how-ai-is-disrupting-shopping
Introducing shopping research in ChatGPT. (2025, November 25). https://openai.com/index/chatgpt-shopping-research/
Ondrus, Pr. J. (2025, October 16). How AI Agents Could Reshape Digital Privacy [Keynote]. The 14th Asia Privacy Bridge (APB) Forum, 2025, Seoul.
Pandya, V. (2025, August 21). Adobe: Generative AI-powered shopping rises with traffic to U.S. retail sites up 4,700%.Adobe for Business. https://business.adobe.com/blog/generative-ai-powered-shopping-rises-with-traffic-to-retail-sites
Pendyala, A., & Sakowski, M. (2025, November 12). Decoding the Future of Retail: Embracing AI Shopping Agents | AWS for Industries. https://aws.amazon.com/blogs/industries/decoding-the-future-of-retail-embracing-ai-shopping-agents/
Perplexity Hub—Shopping That Puts You First. (2025, November 25). https://www.perplexity.ai/fr/hub/blog/shopping-that-puts-you-first
Pol, T. (2025, November 17). Generative Engine Optimization: The New Era of Search. Semrush Blog. https://www.semrush.com/blog/generative-engine-optimization/null
Schumacher, K., & Roberts, R. (2025, October 17). Agentic commerce 6 areas flipcards. https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-agentic-commerce-opportunity-how-ai-agents-are-ushering-in-a-new-era-for-consumers-and-merchants
Silliman, E., Boudet, J., Robinson, K., & Oppong, D. (2025, October 16). Winning in the age of AI search | McKinsey. https://www.mckinsey.com/capabilities/growth-marketing-and-sales/our-insights/new-front-door-to-the-internet-winning-in-the-age-of-ai-search#/
Srinivasan. (2023, May 10). Search ads today and in the future. Google. https://blog.google/products/ads-commerce/google-search-ads-generative-ai/
Staff, A. (2025, April 3). Amazon’s new “Buy for Me” feature helps customers find and buy products from other brands’ sites. https://www.aboutamazon.com/news/retail/amazon-shopping-app-buy-for-me-brands
Statista Research Department. (n.d.). Top web domains cited by large language models (LLMs) in June 2025. Statista. Retrieved December 9, 2025, from https://www.statista.com/statistics/1620335/top-web-domains-cited-by-llms/?srsltid=AfmBOorCWO1ITGTfvQJjCLVb3vKVtKup6DIeexDYp2LTskiBUHo9rv-F
Statista Research Department. (2025, November 28). Ranking positions of search results cited by large language models (LLMs) as of June 2025. Statista. https://www.statista.com/statistics/1620412/llm-cited-search-results-ranking-positions/
Sundararajan, R., & Jeenah, U. (2025). Agentic AI’s disruption of retail and SME banking | McKinsey. https://www.mckinsey.com/industries/financial-services/our-insights/the-end-of-inertia-agentic-ais-disruption-of-retail-and-sme-banking
The end of SEO as we know it? Agentic commerce and banking’s next digital frontier. (2025, September 17). PwC. https://www.pwc.com/us/en/industries/financial-services/library/agentic-commerce-and-banking.html
Visa and AWS Enable Next-Generation Agentic Commerce Capabilities. (2025, December 1). US Press Center. https://press.aboutamazon.com/aws/2025/12/visa-and-aws-enable-next-generation-agentic-commerce-capabilities