What is Generative Search?

Generative search represents the most significant shift in how people find information online since Google introduced PageRank in 1998. Traditional search engines return a list of links ranked by relevance. Generative search creates original responses by synthesising information from multiple sources and presenting it directly to users. Google’s Search Generative Experience, Bing’s integration with ChatGPT and newer platforms like Perplexity exemplify this transformation. The change affects everything from how businesses approach SEO to how consumers expect to interact with information online.

The technology relies on large language models trained on vast datasets. These models don’t simply match keywords to documents. They understand context and construct answers that directly address user intent. When someone searches for “best way to remove red wine stains from carpet”, a traditional search engine returns ten blue links to cleaning blogs. A generative search engine might provide a step-by-step method drawn from multiple sources, complete with timing recommendations and material specifications. This difference seems subtle but its implications for website traffic and user behaviour are profound.

What makes generative search particularly disruptive is its ability to satisfy user queries without requiring clicks to external websites. Publishers who once relied on search traffic now face a future where their content feeds AI responses but generates no visits. The New York Times lawsuit against OpenAI illustrates how content creators view this development. Their articles train the models and inform the answers, yet they receive no traffic, no ad impressions and no subscription opportunities. The economic model that sustained online publishing for two decades faces existential questions.

How Generative Answers Change User Expectations

People searching for information have grown accustomed to immediate, comprehensive responses. Generative search accelerates this expectation. Users no longer want to click through five different websites to piece together an answer. They expect the search engine to do that synthesis work for them. This shift mirrors broader changes in digital consumption patterns where convenience consistently trumps almost every other consideration.

The interface design of generative search tools reflects this priority. Perplexity presents answers at the top of the page with source citations appearing as footnotes. Google’s AI Overviews appear above traditional search results, often answering the query completely before users scroll to the organic listings. Bing’s Chat mode (now integrated into Copilot) presents a conversational interface where follow-up questions feel natural. Each design choice prioritises the generated answer over the traditional link list, fundamentally changing the user’s relationship with search results.

These changes create a particularly interesting challenge for commercial queries. Someone searching “best running shoes for flat feet” represents significant commercial intent. Traditional search allows advertisers to bid for visibility through paid placements and SEO practitioners to compete for organic rankings. Generative search consolidates that commercial opportunity into a single AI-generated response. The shoes recommended by the AI become the default choices, whilst everything else requires active exploration from users who increasingly expect not to need active exploration. This concentration of influence makes the algorithms behind generative search more commercially significant than any search ranking system that came before.

The Importance of Source Attribution in Establishing Trust

Generative search engines face persistent criticism about accuracy and hallucination. These systems can present false information with the same confidence they present facts. Source attribution serves as the primary defence against this weakness. When Perplexity cites specific articles for each claim in its answer, users can verify the information if they choose. When systems fail to provide clear attribution, they ask users to trust AI-generated content without verification pathways.

The legal and ethical implications extend beyond user experience. Publishers argue that proper attribution should include prominent links that drive traffic back to source material. They view the current implementation of citations as insufficient compensation for content usage. Some proposals suggest mandatory click-throughs or revenue sharing arrangements. The parallel to how Google News created abbreviated summaries but still drove substantial traffic to news sites seems obvious, yet generative search proves far more capable of answering queries completely.

Different platforms handle attribution with varying degrees of transparency. ChatGPT’s browse mode shows which websites it consulted but doesn’t link specific claims to specific sources. Google’s AI Overviews include expandable sections showing source websites but the presentation minimises their visibility. Perplexity makes citations more prominent but still positions them as optional verification rather than required reading. The design patterns reveal each platform’s priority: user convenience wins over publisher interests in nearly every case.

Why Trust Signals Determine AI Source Selection

Language models trained to generate search responses learn to prioritise certain types of sources. Academic papers, government websites, established news organisations and recognised experts receive preferential treatment in most systems. The reasoning mirrors traditional search algorithms but operates at a different level. Rather than calculating authority through backlinks, AI systems appear to weight sources based on patterns in their training data about what constitutes reliable information.

Building the signals that AI systems recognise as trustworthy becomes a new focus for digital strategy. Author credentials and expertise markers like professional qualifications or institutional affiliations carry weight. Publishing in peer-reviewed journals or established industry publications creates associations with quality. Even simple factors like domain age and the presence of detailed “about” pages that establish organisational credibility seem to influence AI source selection.

The challenge for smaller businesses and newer websites seems obvious. They lack the historical authority signals that AI systems favour. Breaking into generative search results requires alternative approaches. Developing genuine expertise and documenting it thoroughly helps. Contributing expert commentary to established publications builds credibility. Creating original research or data that becomes the primary source for specific claims establishes authority even without decades of online presence. The path proves narrower but remains accessible for organisations willing to build real expertise rather than just SEO tactics.

New Generative Search Metrics for SEO Strategy

Search engine optimisation as a discipline must adapt to a reality where ranking first means less than it used to. Being the top organic result still matters, but if Google’s AI Overview answers the question using your content without sending traffic your way, the achievement feels hollow. SEO professionals now consider not just how to rank but how to become the authoritative source that AI systems cite and reference.

Content strategies increasingly focus on depth and expertise. Generic blog posts that repackage common knowledge offer little value in a world where AI can aggregate that knowledge instantly. Original research, unique datasets, expert interviews and proprietary methodologies become more valuable because they represent information AI cannot simply synthesise from existing sources. A financial advisory firm publishing original market research creates content that AI must reference rather than replace.

The technical aspects of SEO still apply but the context shifts. Structured data helps AI systems understand and extract information from your pages. Clear, well-organised content makes it easier for language models to parse and cite your work accurately. Fast loading times and mobile optimisation matter because generative search systems often access content the same way users do. The fundamentals remain sound but the end goal changes from attracting clicks to becoming the authoritative source that AI systems trust.

Structural Changes to Traditional Search Results

Generative search doesn’t eliminate traditional search results but it changes their role. Google still displays ten blue links below its AI Overviews. Bing shows conventional results alongside its chat interface. The question is whether users bother scrolling past the AI-generated answer. Early data suggests that when generative search provides a satisfactory answer, click-through rates to traditional results drop significantly.

This creates split search ecosystems. Simple informational queries get answered by AI and generate minimal external clicks. Complex queries, breaking news, commercial research and situations where users want multiple perspectives still drive traffic to websites. The types of queries that convert well for advertisers and publishers shift. Someone who clicks through to your website after reading an AI overview likely has more specific needs or stronger commercial intent than someone arriving from a traditional search result.

Publishers and website owners face difficult decisions. Do they optimise for AI citation even if it reduces traffic? Do they restructure content to be less complete so users must visit their site for full information? Do they focus entirely on query types that AI handles poorly? Each approach carries risks. The most sustainable strategy probably involves creating content valuable enough that AI systems must reference it whilst building direct relationships with audiences through newsletters, communities and proprietary tools that search engines cannot replicate.

Methods for Identifying Intent in Natural Language Queries.

Traditional keyword research assumes people search using short phrases. Someone looking for Italian restaurants nearby might type “Italian restaurant Surrey”. Generative search encourages more conversational queries. The same person might ask “which Italian restaurants in Surrey have outdoor seating and accommodate large groups”. The increased specificity changes what content ranks and how websites should structure information.

Long-tail keywords become even more significant. Specific questions with clear intent match perfectly with how generative search works. A blog post titled “Do Air Fryers Use Less Electricity Than Conventional Ovens” targets a precise query that generative search can answer directly. The AI might use that content to inform its response about energy costs, cooking efficiency, environmental impact and comparisons between them. Websites that anticipate these specific questions rather than optimising for broad terms like “air fryer benefits” position themselves better for citation.

The conversational nature of generative search also enables multi-turn interactions. Someone researching holiday destinations might start with “warm places to visit in January”, then follow up with “which has the best beaches” and then “what about family-friendly options”. Each query builds on the previous context. Websites that structure content to address related questions sequentially create more opportunities for ongoing citation throughout these conversations. The old SEO model of one keyword per page gives way to content that addresses clusters of related intent.

How Commercial Intent Gets Captured or Lost

Follow the money and you see why generative search creates such anxiety among advertisers and affiliate marketers. Someone searching for “best CRM software for small business” represents immediate commercial value. They’re comparing options and likely approaching a purchase decision. Traditional search lets multiple vendors compete for that attention through paid ads and organic rankings. Generative search might recommend three specific options based on the query parameters, effectively choosing winners before the user sees any competitive landscape.

The platforms developing generative search face interesting incentives. Google generates massive revenue from advertising. Bing competes primarily on innovation rather than market share. Perplexity operates a subscription model. Each platform’s business model influences how they handle commercial queries. Google has been notably cautious about using AI Overviews for searches with clear commercial intent, presumably because doing so cannibalises their advertising revenue. Bing shows more willingness to provide direct recommendations, perhaps because capturing any market share from Google proves more valuable than protecting a smaller ad business.

Advertisers adapt by seeking placement within AI-generated responses. Sponsoring the content that AI systems cite becomes one approach. Ensuring product databases and specifications appear in formats that AI can easily reference represents another. Some platforms will likely develop “sponsored recommendations” within generative search results, creating new advertising inventory whilst maintaining the veneer of AI-powered objectivity. The methods will differ but the underlying dynamic remains constant, valuable user attention creates commercial opportunity and commercial opportunity attracts those willing to pay for influence.

For almost two decades, we’ve helped businesses adapt to major shifts in search technology and digital visibility. Operating from our primary office in Horley, Surrey, alongside our Peckham and Hampstead studios in London, we specialise in designing websites and developing marketing strategies for the AI-driven search environment. If you’re looking to adapt your content for generative platforms or need guidance on staying visible as search behaviour changes, we’re here to support you. Reach out to discover how we can keep your business discoverable in this new search reality