Skip to content
AI Chatbots Are Changing Search Behaviour
AI chatbots like ChatGPT and Bing Chat have rapidly emerged as new tools for information discovery, altering traditional search habits. Users are increasingly posing questions to these conversational agents, which can directly provide answers or recommendations. In fact, recent data shows ChatGPT has become a notable traffic referrer for websites. Education, technology, and software development sites have seen a boost in referral traffic from ChatGPT, which by November 2024 had sent visitors to over 30,000 unique domains searchengineland.com. This indicates that ChatGPT-driven searches are actively leading users to click through to source websites in many cases.
However, the way users interact with chatbots differs from classic search engines. According to a Semrush analysis of millions of ChatGPT interactions, the chatbot answered about 54% of user queries without using any external search, relying solely on its trained knowledge. The remaining 46% of queries did trigger a live web search as part of ChatGPT’s process searchengineland.com. This means nearly half the time ChatGPT is fetching real-time information from the web – a behaviour more akin to a search engine – and providing answers with links or sources. Notably, ChatGPT queries tend to be much longer and more detailed than typical Google queries (averaging 23 words per prompt, versus ~4 words for search engine queries) searchengineland.com. Users often ask open-ended, complex questions in chat form, which signals a shift toward more conversational search behaviour.
The rise of chat-based search is also beginning to show up in market share statistics. While Google is still the dominant player by far, some analyses suggest ChatGPT may account for roughly 4% of the global “search” market share by early 2025 alphametic.com. Hundreds of millions of users now turn to ChatGPT for answers each month. (For scale, Google still saw about 6.5 billion unique visitors worldwide in December 2024, compared to 566 million for ChatGPT searchengineland.com.) ChatGPT’s user base skews younger and more tech-savvy – for example, it’s especially popular among students – whereas Google retains broader appeal across demographics searchengineland.com. This growing usage of AI chatbots for finding information means web traffic patterns are evolving, as users diversify beyond traditional search engines.
Google’s AI-Powered Search Initiatives (SGE and “AI Mode”)
Google has responded to the chatbot trend by infusing its own search engine with AI capabilities. In mid-2023, Google introduced the Search Generative Experience (SGE) in select markets – an experimental feature that displays AI-generated overview answers at the top of search results for certain queries. These AI overviews synthesize content from multiple web sources to answer the query directly on the results page. While helpful for users, publishers quickly grew concerned about the impact on clicks, since users might get their answer from the overview and not click through to the source. Early reports have validated those concerns: Google’s AI-generated answers have already had a “very negative impact on click-through rates” from the search results to websites searchengineland.com.
In other words, if your webpage’s information is used in an AI summary but not clicked, you lose traffic. A recent study even found that not appearing in Google’s AI overview for a topic (when competitors do) can significantly hurt a page’s visits – highlighting how critical it is to be included in these AI-driven results.
This tension hit a flashpoint in early 2025 when education company Chegg sued Google, blaming the AI overview feature for a decline in its business. Chegg claims Google’s AI answers for student queries are using Chegg’s content to satisfy users without sending them to Chegg’s site, thereby “blocking” a major portion of Chegg’s usual traffic searchengineland.com. The complaint argues that Google’s dominance forces content providers to give up their data, only for Google’s AI to use that content to keep users on Google searchengineland.com. This legal dispute underscores the growing friction between content publishers and AI-driven search features that aggregate information from across the web.
Despite these challenges, Google is doubling down on AI. The company is reportedly testing a next-generation search experience called “AI Mode.” Leaked details in February 2025 suggest AI Mode will provide a “persistent place” for users to ask open-ended or exploratory questions and receive AI-generated answers with the ability to ask follow-ups 9to5google.com 9to5google.com. In essence, it brings a ChatGPT-like conversational interface directly into Google Search. Internal descriptions say Google will “intelligently research for you – organising information into easy-to-digest breakdowns with links to explore content across the web” searchengineland.com.
Unlike a one-off snippet, AI Mode is designed for multi-turn interactions (e.g. asking a broad question, then refining or drilling deeper with follow-up queries). It’s aimed at queries that traditional search struggles with – such as advice, comparisons, or highly specific questions that benefit from synthesis searchengineland.com.
Importantly, Google’s AI Mode is reportedly powered by Gemini 2.0, Google’s latest large language model, which promises more advanced reasoning abilities 9to5google.com. A screenshot of the prototype shows an “AI” toggle or tab in the Google interface, alongside familiar options like Images or News 9to5google.com. On mobile, users may even have voice input for asking the AI follow-up questions 9to5google.com. While still in testing (with Google employees in the US for now 9to5google.com), CEO Sundar Pichai has hinted at major search changes on the horizon, noting that “2025 is going to be one of the biggest years for Search innovation yet” 9to5google.com. If AI Mode rolls out broadly, it could transform how users interact with Google – and by extension, how websites get traffic from Google.
The advent of these AI-driven search features means the familiar ranking dynamics are in flux. One analysis found that the pages cited in Google’s AI overviews often don’t match the top organic results, and the AI-chosen sources can be quite volatile searchengineland.com searchengineland.com. In fact, about 40% of the time, an AI summary will cite a page that wasn’t even in the top 10 traditional results for that query searchengineland.com. This decoupling of AI results from the classic SEO rankings presents a new challenge: even if you rank well in normal search, you’re not guaranteed to be featured in the AI answer, and vice versa.
Moreover, the content of AI summaries can change frequently, independent of stable organic rankings searchengineland.com. Google’s own advice to site owners is to monitor which queries their content is appearing (or not appearing) for in AI results, as this may require a different optimisation approach searchengineland.com searchengineland.com.
SEO Impact: Challenges and Opportunities
For SEO professionals and content creators, the rise of AI chatbots and generative search is a double-edged sword. On one hand, these AI platforms represent new traffic channels. ChatGPT, Bing Chat, and other AI search tools can drive visitors to your site if your content is cited or recommended in their answers. For example, ChatGPT has been especially effective at referring traffic to sites focused on learning, coding, and research searchengineland.com.
Some niche websites that never dominated Google’s page-one rankings might find that they can attract an audience through AI recommendations instead. This opens opportunities for quality content to be discovered in new ways, beyond the traditional SEO playbook of climbing Google’s ranks.
On the other hand, AI-driven search poses new challenges and uncertainties. A major concern is the potential loss of clicks. If an AI summary or chatbot directly provides the information a user needs, the user might not visit any website at all. Early data from Google’s SGE confirms that many users are satisfied by the AI snippet, leading to significantly lower click-through rates on the organic results below searchengineland.com.
This “zero-click” trend isn’t entirely new (Google’s featured snippets and instant answers have long reduced clicks for certain queries), but AI takes it to another level by aggregating answers from multiple sources and engaging the user in a longer dialogue. Content publishers worry about being robbed of traffic and credit – a sentiment powerfully voiced in the Chegg vs. Google case, where Chegg argues Google’s AI uses its content to answer student questions without students ever visiting Chegg’s site searchengineland.com searchengineland.com.
There’s also the matter of attribution and accuracy. Some AI answers (like Bing Chat) do include footnotes or citations linking to source pages, which can funnel curious users to the original content. Other implementations (including Google’s early AI overviews) have been more opaque, summarising without clear attribution. This makes building brand awareness and trust more difficult – if users get the info from an AI, they might not even know which site provided it. SEO now isn’t just about getting ranked; it’s about getting credit in an AI-generated answer. Ensuring your brand is mentioned or your page is the one cited by the AI is becoming a key goal.
Nevertheless, where users find value, traffic tends to follow in some form. Chatbots excel at handling broad or complex queries, but they often still guide users to external links for more detail or for actions like purchases. A user asking a chatbot for the best laptop might read the AI’s summary then click a provided link to a review site or retailer. In this sense, having your site recommended by an AI can be as valuable as a top ranking on page one – it might even be the only recommendation shown. As AI search experiences mature, we may see new metrics of success, such as “conversation share” (how often a brand or page is mentioned in AI dialogues) in addition to classic search ranking.
Optimisation Strategies for AI-Driven Search
With search models evolving, SEO strategy is evolving in tandem. Experts are increasingly referring to “Generative Engine Optimisation (GEO)” – essentially, optimising your content to be favoured by generative AI models searchengineland.com. While the fundamentals of good content remain, here are some key optimisation strategies emerging for this new landscape:
- Focus on Comprehensive, Problem-Solving Content: Aim to create content that actively supports learning, problem-solving, and creative tasks – the kind of in-depth material a chatbot would find useful to draw from searchengineland.com. Rather than thin pieces targeting single keywords, provide thorough answers, explanations, or guides. This increases the chances an AI will use your content to formulate its response.
- Ensure Content Is Structured for AI Consumption: Use clear headings, concise paragraphs, and structured data (schema markup) where appropriate. Remember that AI models “read” the entire web to generate answers. Making sure your brand’s content can be easily understood and accurately cited by LLMs (large language models) is increasingly critical searchengineland.com. For instance, if you answer a common question in a well-formatted way, an AI summary might pull directly from that answer and credit your site.
- Build Authority and Mentions Across the Web: Authority is king in the AI era. Large language models are more likely to trust and quote information from sources that have established expertise and credibility. To that end, build your brand’s topical authority by publishing expert content and original research. Engage in digital PR to get your brand mentioned on high-authority sites and forums. Even an unlinked brand mention in a respected publication can increase your presence in the AI training data and knowledge graphs searchengineland.com searchengineland.com. The more an AI has “seen” your site or brand referenced in context with reliable information, the more likely it will surface your content.
- Optimise for Bing and Other AI-Integrated Search Engines: Google isn’t the only game in town for AI search. Microsoft’s Bing not only has its own user base, but it also powers ChatGPT’s live web results and other AI applications alphametic.com. In practice, this means ranking well on Bing can directly lead to your content being recommended by ChatGPT in conversational answers. Bing’s algorithm has its own nuances, so audit your SEO for Bing’s requirements (such as using Bing Webmaster Tools, leveraging social signals, etc.). By improving on Bing, you effectively double-dip – gaining visibility on Bing itself and within any AI that uses Bing’s index or data alphametic.com.
- Monitor AI Search Performance: Just as you track keyword rankings on Google, start tracking how and when your content appears in AI-driven contexts. Some tools and experimental techniques (like querying chatbots with specific prompts) can help identify if your site is being cited. Because AI results can be volatile – one month an AI might cite your page, the next month it might choose a different source searchengineland.comsearchengineland.com – continuous monitoring is important. If you “lose” an AI citation spot, investigate why. It could be that a competitor published more up-to-date info, or the AI model updated. Staying on top of these shifts will be an emerging part of SEO work.
- Adapt Your SEO Metrics and Expectations: Finally, success in an AI-driven search world may need rethinking of metrics. Instead of focusing only on clicks and traditional rankings, consider metrics like implied reach (how many queries your information is used in, even if no click) or engagement with AI results. For example, if your content is frequently summarized by Google’s AI or cited by Bing Chat, that exposure has value, even if it doesn’t always translate into a click.
Over time, search engines might provide publishers more feedback or credit for AI usage of their content. Until then, balance your efforts between attracting human clicks and feeding the AI answers. Both aspects can drive awareness and, ultimately, business outcomes.
The Road Ahead
As of early 2025, AI chatbots and generative search engines are no longer just novelties – they are mainstream tools influencing how people find information. Google’s upcoming AI Mode and similar innovations promise to blur the line between a search engine and a personal research assistant. For users, this means more convenience and conversational interactivity. For website owners and SEO practitioners, it means adapting to a world where being the source of answers might matter more than being the top blue link. The core challenge is ensuring your content remains visible and valuable in these new interfaces.
In this rapidly evolving landscape, staying informed is crucial. Major players are continually updating their algorithms and AI capabilities. (Even Google’s algorithms for AI snippets work independently from regular search rankings, so both need attention searchengineland.com.) The companies whose content strategies can ride this wave – by embracing AI-driven optimization while continuing to deliver quality, trustworthy information – are likely to thrive. In summary, AI-driven search is changing the game, but it’s an opportunity as much as a threat. By focusing on authoritative content, technical optimisation, and cross-platform visibility, brands can ensure they are part of the conversation – whether it’s happening between a user and a search box, or a user and an AI chatbot searchengineland.com. The tools may be evolving, but the goal remains the same: connect with your audience by providing the information they need, wherever and however they seek it.
