Established 28 Years
Flexible Packages 
Based on the South Coast
27 March 2025

How to Get Your Business Website Selected for Recommendations in AI Overviews and ChatGPT


Optimising Your Website for Citations in AI-Powered Search

While the digital and tech industries have been quick to explore AI for business in its entirety, that’s virtual assistants, AI data processing and of course generative AI in search, product recommendations and chat functions, other industries have been more reserved in the uptake of using AI and AI-driven search. In the first quarter of 2025, noticeable trends indicate that both businesses and consumers across the board are gaining confidence in generative AI in SEO. More people are turning to AI Overviews alongside traditional search for research and recommendations looking for answers that your business can provide. As confidence in AI search grows, so does the opportunity for businesses to use AI-powered search as a platform to receive service, product or advice recommendations via generative AI engines such as AI Overviews, ChatGPT, Claude and Perplexity, in addition to traditional SEO.

How Can Generative AI-Search Help Businesses?

Since generative AI in SEO made its breakthrough in early 2023, use cases were initially faced with controversy and resistance, however, over the past 6–9 months, generative AI learning systems have evolved to become the quickest and most personalised version of search available to the consumer. In 2025, use cases have become so advanced that business have been able to create:
  • A wine merchant established citations from its wine database in providing personalised wine recommendations
  • An automobile trader developed generative AI-driven owner's manual advice and car troubleshooting guides for user queries.
  • A private healthcare provider offered health advice based on the description of symptoms.
  • A waste management company was able to use generative AI to supply information to users so they can more easily identify recyclable goods and dispose of them appropriately.
Studies investigating customer search habits have observed that AI search queries are becoming more complex and highly specific or personalised. The examples above, supported by research, demonstrate that new search habits can be utilised by businesses who want to provide exceptional advice and serviced based content as part of their TOF organic strategy. Advancements in some generative engines are making it (almost) possible to purchase products directly from the Chat Experience Results Page (CHERP). When this becomes a widely adopted feature in AI Search Engines will have the potential to be optimised for the full sales funnel (TOF, MOF and BOF).

How Do AI Models Select Websites?

While traditional search engines use sophisticated natural language processing (NLP) to infer meaning from web pages, AI-powered search models go a step further by aggregating and synthesising information into direct answers. AI systems do not "search" in the traditional sense. Instead, they retrieve, synthesise, and present information based on various signals such as domain authority, content depth, structured data, and relevance. Platforms like ChatGPT and Google's AI Overviews provide concise, synthesised answers, often pulling from various web sources. To have your website featured in these AI-generated responses, it's essential to adapt and optimise your content accordingly. Google's AI Overview, for example, aggregates high-ranking and high-trust sources, while ChatGPT with browsing capabilities uses Google’s and Bing’s index to fetch information, so by achieving high organic rankings, via an SEO strategy, your business also increases the likelihood of your content being cited by AI tools. Studies indicate that a significant portion of AI citations come from pages ranking within the top 12 positions on search engines. Therefore, focusing on keyword optimisation, quality backlinks, and user engagement metrics is essential. Factors influencing AI-generated citations include:
  • Content relevance: Does the content offer depth and expertise relative to the user queries?
  • Authority and trustworthiness: Is the domain cited by reputable sources within a similar niche or industry?
  • SEO fundamentals: Does the page rank well in traditional search?
  • Structured data: Is the content easily digestible for AI agents and crawlers?
 

1. Optimise for Traditional SEO Principles

We are now at a time where use cases of AI-search have filtered down to the everyday consumer and it’s now being used by your ideal customers to research, compare, and make purchasing decisions faster than ever before. AI-optimised search, such as AI overviews, is featured at the very top of traditional search, in utilising this platform you have the potential to create one of the most efficient online experiences, generate additional TOF opportunities for your existing SEO strategy and guide users straight to the answers, products, or services they need. This means higher click-through rates, improved conversions, and an overall better customer journey. In a recent blog, I explored how AI-powered search for SMBs is making it easier to compete in SERPs for visibility against larger dominant brands. Unlike traditional search, where large brands often dominate the top spots, generative AI engines prioritise credibility, accuracy, and relevance over pure ranking position. In fact, recent AI models are more likely to pull sources from the top 12 search results rather than just the top three, giving smaller scale businesses a real opportunity to gain visibility. AI-search models assess factors around domain authority, content quality, and user engagement to determine which sources to reference in their responses. Therefore, aligning your website's content with these criteria is essential for inclusion in AI-generated citations.​ Although AI-generated search results operate differently from traditional search engines, organic ranking remains a crucial factor:
  • High-ranking keywords are naturally integrated.
  • Your website is optimised for page experience, mobile usability and Core Web Vitals.
  • Backlink strategies to improve domain authority.
  • Structured data and Schema Markup is implemented to improve content readability for AI agents and crawlers.

2. Do Generative AI Engines have a Preferred Writing and Language Style?

AI tools cite content from domains with a history of high-quality, informative, and well-researched material. AI models favour content that is comprehensive, accurate, and authoritative. Producing in-depth articles that thoroughly address topics, incorporate reliable data, and provide valuable insights can position your website as a trusted source for AI references. In our own research, conducted at UTOPi,  we have tested theories of AI citation selection and found certain sentence and paragraph structures are favoured by AI and are rewarded with selection. We were able to manipulate a theory that explains that AI search models, such as ChatGPT, use encoders trained on a large set of text data to recognise the patterns and relationships between words in different contexts. These data sets typically follow a similar sentence and context structure, so in replicating this sentence and content structure in your website content, you are aligning your content with patterns the AI model has been taught to recognise, aiding the processing of your content. This is just one of the techniques we use in our Generative Engine Optimisations (GEO) services. In a related ongoing study at UTOPi, we are continuing our linguistic analysis of ChatGPT and AI Overviews, with a specific focus on AI-generated chat responses. Since AI-search models are trained on structured language patterns during the encoding process, their decoded responses are expected to reflect these patterns, maintaining coherence and alignment with the original training data. Our research aims to determine whether reverse engineering the sentence and content structure of AI-generated responses can aid the AI model’s ability to interpret page content. Additionally, our findings indicate that certain sentence and paragraph structures, which are not present in generative AI outputs, tend to be processed less accurately when incorporated into website content. As our study progresses, we seek to identify a linguistic fingerprint for widely used AI-powered search engines. This involves analysing language style, the presence or absence of specific words and character combinations, the frequency of function words (e.g., "of," "for"), and metrics such as average and maximum sentence length, word length, and vocabulary distribution. Establishing these characteristics will provide valuable insights into optimising website content to align with the linguistic patterns AI models are trained to recognise. Some key findings so far include:
  • Sentences often contain multiple clauses: A lot of sentences contain conjunctions such as "however," "therefore," and "in addition."
  • Paragraphs are structured and concise: Typically 3–5 sentences per paragraph.
  • Character count per paragraph: Ranges from 200–500 characters. This also helps readability.
  • Varied sentence length: While most sentences are medium to long (15–30 words), shorter sentences are often used for explanations or core information.
  • Formal and academic TOV: Responses by default favour professional and precise wording over colloquial or conversational terms.

3. Establish Authority and Topical Authority

Authority and topical authority are another very important element to getting website citations in AI Overviews and ChatGPT-generated responses, as these systems prioritise sources that demonstrate deep expertise, credibility, and trustworthiness. Unlike traditional search algorithms that rely heavily on link equity and keyword matching, AI-driven search mechanisms assess the semantic depth, coherence, and reliability of a source before integrating it into a response. This type of processing highlights the need for a well-defined niche of expertise within your website rather than attempting to cover a broad spectrum of topics with superficial knowledge. When your business website commits to a focused subject matter, it not only enhances its topical authority but also helps to position your brand as a recognised leader in its industry. Google's algorithms and, by extension, AI search engines reward sites that consistently produce in-depth, high-quality content within a specific domain. This is largely due to the way AI models assess content, relying on entity recognition, knowledge graph associations, and semantic relevance to determine credibility and alignment with user intent. A well-defined niche enhances a site's visibility within AI-powered search by reinforcing its association with key entities in the knowledge graph, ensuring that it is recognised as a reliable and authoritative source in generative search results.

The Power of a Topical Focus in Natural Link Acquisition

Focusing on a niche creates a compounding effect in the way of consistently publishing high-quality, interrelated content, you create a dense internal knowledge network that strengthens your authority within that topic. This depth signals to AI-driven search engines that your site is a definitive source, increasing the likelihood of citation in AI-generated responses. A strong topical focus encourages natural link acquisition, as other businesses within the industry , journalists, or bloggers are more likely to cite a highly authoritative resource rather than a generalist website that lacks a clear specialisation. From the viewpoint of Generative AI models, being a master of a niche is significantly more valuable than being a ‘Jack of all trades’. A diluted content strategy that attempts to cover multiple disciplines weakens a site's perceived expertise, making it less attractive to AI-driven search models that seek the most authoritative and relevant sources. Instead, by concentrating on a specific field and continuously reinforcing expertise through original research, case studies, expert interviews and thought leadership, a brand can solidify its authority in both human and AI-driven search.  

4. Use Schema Markup and Structured Data

By implementing structured data, you provide AI agents with a clear, machine-readable framework that helps content discoverability and improves semantic understanding. Within each schema, there is a list of properties and items which are specific entities which feed into AI agent logic to process the page content. Think of schema markup like a restaurant menu. Imagine an AI search engine as a diner walking into a restaurant, looking for the best meal to order. If the restaurant has a well-structured menu with clear categories (starters, mains, desserts) and detailed descriptions (ingredients, dietary info), the diner can quickly decide what to order. However, if the menu is unstructured—just a long, unordered list of dishes with no context—it takes longer for the diner to figure out their options, and they might choose to leave. Schema works the same way. It gives AI search engines an organized "menu" of your website's content, ensuring that it can quickly extract relevant details and present them to users. Without schema, AI models must rely on vague signals and context clues, which can make content less likely to be selected for AI-generated responses.
  • Article schema for blog content.
  • FAQ schema for direct answer retrieval.
  • How to Schema for AI-driven responses that guide users through processes
  • Product Schema for e-commerce and product-based businesses with product details such as: product name, price, and availability
  • Person and Organisation schema to establish credibility

What are our key learnings for how to become cited for AI-Search Models? 

At UTOPi we believe the key to being cited in AI Overviews and ChatGPT responses isn’t limited to chasing traditional algorithms. Traditional SEO practices have a large role to play however SMBs have an opportunity to get ahead in AI recognition by creating content that genuinely services consumer problems. Search engines and AI models are becoming increasingly sophisticated in identifying sources that provide real value, meaning businesses that focus on depth, expertise, and user-centric content will naturally rise above the sources which offer repetitive and most of the time non-engaging content. By mastering a niche, structuring content, and correcting technical SEO fundamentals, your website can position itself as a go-to resource not only for traditional search but also for the future of AI-driven search. We offer our Strategic Development Plan which is our hybrid service combining SEO, GEO and CRO. One aim within this plan is to develop your brand as a thought leader, build your website’s authority, and help you define a niche topical focus to not onyu increase your visibility in AI-generated results but help you to establish long-term brand credibility in your industry. As we take our generative AI learnings forward into Q2 2025, we are set to help businesses prepare for this shift towards the future of SEO. Drop us a message or give us a call, and we can discuss SEO opportunities specific to your website and your business goals

MORE READING...

Recent Articles

17 April 2025
Building a Woman Owned SEO Agency and Developing the Organic Optimisation Model
27 March 2025
How to Get Your Business Website Selected for Recommendations in AI Overviews and ChatGPT
20 March 2025
Information Architecture's Influence on SEO and AI
7 March 2025
Are You Harming Your Website By Chasing Backlink Success?
UTOPi Digital Optimisation is a division of Cammegh Davies Fleming. Established for 28 years & founded in expertise and creativity 
LET'S START
Privacy PolicyTerms & Conditons
Copyright © UTOPi Digital Optimisation 2025