AI search is all that anyone seems to be talking about in the world of SEO. It’s gone from being something we need to be aware of in the distant future to something that is affecting our businesses now, in the space of just a few months. I’d say this is the biggest change in the way information is discovered online since social media arrived and the biggest shift in search since it started. Instead of scrolling through a list of results and clicking through to the one that we decide best matches what we are looking for, we get a ready made answer which, in theory, at least, is tailored to be exactly what we wanted. This is huge for marketers as it significantly changes our visibility in search results and makes it far less predictable.
But more of that later. Let’s take a step back and look at what’s been happening.
What Is AI Search?
Unlike traditional search engines that rely heavily on keyword matching, AI search uses large language models (LLMs) to try to understand user intent and provide more comprehensive, direct answers that are closely aligned to the actual need behind the user’s search. You’ll have seen these results in the form of AI Overviews at the top of Google’s search results, either above or below the ads. According to Google there are 1.5 billion users of AI overview each month.
In the USA, Google has just rolled out AI Mode to all users, which is taking AI search one step further. This is how Google describes it: This new Search mode expands what AI Overviews can do with more advanced reasoning, thinking and multimodal capabilities so you can get help with even your toughest questions. You can ask anything on your mind and get a helpful AI-powered response with the ability to go further with follow-up questions and helpful web links.
You can see what it looks like in this explainer from Google.
So it seems as if it is like an AI overview with the ability to refine and drill down the way you do in Chat GPT or Gemini. It’s going to be accessed via a new tab under the Google search bar, although I’ve also heard rumours that it could become the default search option. It’s not yet in the UK but we can be sure that it is coming, and soon!
Understanding AI Search
Google’s initial mission was “to organize the world's information and make it universally accessible and useful” To do this, it indexes all the pages on the web and uses an ever-changing algorithm to decide what order to rank these pages in response to a particular search query from a user.
The signals search engines use were originally based on the number of times the search term was used on the page - which inevitably led to keyword stuffing and all sorts of other dodgy practices - anyone remember white text on a white background? Alongside keywords, backlinks were important signals of a page’s popularity, which again led to abuse such as linkfarms and buying meaningless links. Our original explanation of SEO is still one of the most simple ways to understand the basics.
These practices provided the user with a poor experience as marketers focused on rankings rather than actually providing value for potential customers. They also diminished Google’s own usefulness as the results were not always the most relevant. So the algorithm has been refined and enhanced over time, to help users find the content that is going to be most useful to them at the top of the page. The latest incarnations have been focused on Helpful Content and E-E-A-T (Experience, Expertise, Authoritativeness, and Trustworthiness) to encourage marketers to focus on the quality of their content rather than hacks and shortcuts.
AI Search does not rely on indexes or use these algorithms. Instead it uses LLMs to interpret relationships between words, identify entities, and grasp the overall concept of a search. LLMs have been described as working like word association or predictive text. Fundamentally, they are very good at predicting what word comes next and which words are commonly linked together. The signals are therefore different from those used by search engines.
One commentator in Search Engine Land has an interesting take:
What AI is doing in its current, glitchy, overly confident state is reviving the spirit of early search. Not the cluttered, over-optimized, pay-to-play version we’ve been stuck in. The original idea: that the most valuable voice wins. Not the loudest. Not the richest. Not the most frequent.
Google has AI Overviews and now AI Mode, but Bing also has its own version as does Chat GPT and Perplexity. These are being called Answer Engines because they provide the user with a comprehensive answer right in the results page rather than a list of links.
AI Overviews & Intent
AI Overviews, are influenced by a number factors, largely revolving around the principles of delivering accurate, relevant, and trustworthy information.
Intent is an important factor here. AI Overviews tend to be triggered by informational queries where the user is looking for an answer to a question or the solution to a problem rather than trying to find an item to buy or research.
One way of looking at this is by thinking of Google’s business model. It makes most of its money from advertising - i.e. from people looking for things to buy. It doesn’t make money from people just trying to learn something or answer a question. So there is no benefit to Google if people click through to another site to find the answer. If Google keeps them on the search results page it stands more chance of presenting them with a revenue-producing ad later on.
One example of intent and how it affects the search results page is: type in Chicken Curry, you'll see recipes, while searching for Chicken Curry Near Me gives you local Indian restaurants, and Buy Chicken Curry gives a list of ready made curry providers. That is because the intent of all those searches is different. Then type in what is chicken curry? and you get an AI Overview.
Some sectors, such as health, education, and B2B technology, tend to see a higher prevalence of AI Overviews due to the nature of queries and the need for detailed explanations.
Implications For Marketers
So far so good, but how does this shift affect marketers?
Firstly we need to look at how consumer behaviour is changing. A first-of-its kind UX study into how users interact with AI Overviews shows that click-throughs decline when there is an AI Overview on the page. It also finds that most users click show more to expand the AI Overview but the majority don't scroll all the way down.
However, in B2B related searches behaviour appears to be slightly different. The latest study of B2B buyers indicates that 90% actually click on the citation links. In the B2B sector, despite seeing answers in the AI overviews during the majority of searches, users are keen to check the facts and verify sources for themselves.
Zero Click Marketing
Overall there is no doubt that click through rates are declining and zero-click marketing is on the rise. The statistics here are somewhat confusing: while the click rate is lower, Google’s traffic rose by 22% last year, up to 5 trillion search queries. So for many businesses, the total number of clicks hasn’t dropped.
It is now estimated that more than half of all searches end in zero clicks - i.e. the user does not leave the platform they used to perform the search, whether that is Google, TikTok or somewhere else. This is not particularly new - social platforms are keen to keep users on their own platforms, and prefer content that doesn’t link out to other websites, and Google started providing structured snippets and ready made answers in the form of “People Also Ask” sections a few years ago, reducing the need for users to leave the results page.
But with an AI Overview at the top of the page, users may not even see any links without scrolling, and so clicks are likely to decline further. Part of your strategy now is to optimise for visibility and influence as well as actual traffic.
Semantic Search
Semantic search represents a fundamental shift in how search engines operate, moving beyond simple keyword matching to genuinely understand the meaning and intent behind a user's query, as well as the context of the content itself. It analyses the meaning of words and phrases, considering relationships between them and the context of the search.
When it comes to AI Overviews, semantic search is paramount because these AI-powered summaries are designed to provide direct, concise, and highly relevant answers, much like a human would. If a search engine can accurately understand the contextual meaning and user intent it can produce a more relevant answer than if it relied on exact words alone.
By structuring content semantically – focusing on clear topics, answering direct questions, and demonstrating expertise and authority on their topics – websites can signal their relevance to the AI models, making it more likely that their information will be recognised as the most useful and trustworthy source for a given user query, increasing the chances of being featured in an AI Overview.
Google Ads
At the moment, when an AI Overview appears it is usually on informational queries that don’t have ads. However, Google has confirmed that it will be introducing ads in AI Overviews in the USA for searches with some degree of commercial intent, where there are relevant ads. It will use the content and context of the query plus the generated result to trigger ads.
It appears that adverts will need to be using Google Ads’ “AI Essential” recommendations to appear in the Overviews: these are the automated options like broad match, dynamic search ads and performance max campaigns.
Which Content Appears In AI Overviews
From what has been analysed so far by industry experts such as Neil Patel, it seems that AI Overviews mostly glean their information from the top ten or so web pages in the traditional search results. They also mainly choose the sites with the highest domain authority and prioritise local results.
Recent content also takes precedence over older pages and it’s pretty clear that lists, FAQs, how to guides and step by step instructions are used frequently.
The main shift is in how it chooses pages, as it usually looks at the topic and the intent rather than specific keywords.
Finally, don’t forget about video and even social media as they are also likely to be taken into account.
What Should Marketers Do To Show Up In AI Overviews?
Firstly, don’t panic. If you have been creating helpful content focused on your ideal customers’ needs then you’re on the right path.
The aim, more than ever, is to create authoritative and trustworthy content that is optimised for semantic search. And for informational searches in particular, present that information in a way that is easy for AI engines to pick up, digest and spit back out.
Look at individual sections of content that answer specific questions and start with a concise clear answer in the first 50-100 words of the section. In addition:
Use headings to structure each section
Check that all facts are accurate
Quote your sources
Avoid jargon
Use a conversational tone
Visibility in AI Overviews is becoming increasingly important, as these immediate, concise responses capture users’ attention at the top of the search results page. Understanding the mechanics of AI search and how it differs from traditional search is key to maintaining and enhancing online presence.
How To Create An AI Search Strategy?
Optimising content so that it appears in AI Overviews requires a strategic approach that goes beyond traditional SEO, focusing on creating exceptional, user-focused content that is accurate, authoritative, and easily found and understood by AI models.
Advice for B2B marketers specifically, from the research mentioned earlier, is to create digestible, authoritative content that AI tools can cite, with the aim of driving traffic to your own website for buyers to check.
Here are some specific tips to follow
- Do your research - do AI Overviews currently show for the queries your audience use? If so, dig in and work out what sort of information is being shown, what questions are highlighted and what sort of answers are most likely to be used.
- Define your topics - know what your core topics are and stick to them. Don’t create random content for the sake of it. HubSpot recently lost over half of its blog traffic because it was not sufficiently relevant to its core topics .
- Optimise existing pages - find pages that already exist which are relevant which rank well for that topic. Optimise them by looking at specific sections rather than the page or blog post as a whole. Find sections that could answer the questions in the AI Overview and structure them carefully using a hierarchy of headings, and provide clear, concise explanations at the start of the section and use bullet points to expand on the answer.
- Remember to E-E-A-T - Provide in-depth, helpful content that draws on your own experience to make it unique, and answers the user’s questions. Keep it accurate and quote authoritative sources. Use social proof in the form of reviews, testimonials and case studies where possible.
- Understand intent - work out if the query you are targeting is informational or commercial, so that you can provide the right type of content - a how to guide or a pricing page, for example.
- Use schema - adding schema markup on your most important pages can help the engines understand what type of content is on the page. For example you can indicate that is a product page, an event, a local business, FAQs or an article.
- Add media - unique visuals, infographics and video can all help your content to stand out.
- Refresh the content regularly - while in the past you might have updated your blog posts every couple of years, it is advisable to update this content as often as every 3 months according to the CIM’s Graham Hansell.
- Boost your authority - use digital PR to increase your brand mentions in trusted sources. You want your brand to be mentioned next to specific topic-related words and phrases online on a regular basis so that LLMs associate your business with them.
Measuring AI Search Success
At the moment it is not that easy to track appearances in AI results although there are new tools appearing on the market all the time.
Traditional SEO tools like SEM Rush and Ahrefs now incorporate AI Overview tracking, which is extremely useful. Google Search Console does not currently include AIO appearances but Google has said that AI data is coming soon, but to begin with at least, will be included in the overall numbers and will not be shown separately.
There are also various tools for tracking brand mentions, some of which include AI Overviews.
Google AI Citation Analysis is a useful Chrome extension which analyses the citations in an AI Overview and lists where the source page is found (snippet, people also ask, blue link etc
Looking at website traffic, the Acquisition Report in GA4 shows the source of your website visits and includes AI engines such as ChatGPT, Gemini and Perplexity, so that you can see if it is worth optimising content for these Answer Engines.
Conversion rates can be high for traffic from LLMs as users will have already found a basic answer and narrowed down their options and then decided to click through to your website. In that situation they may well be more ready to take the next step towards purchase than a visitor who has just clicked on a link in the search results. It is definitely worth comparing the conversion rates from search engines and answer engines to see if there is a difference.
Grab The AI Search Opportunity Now
AI search brings a monumental change to the digital world, requiring marketers to rethink and adjust their strategies. Marketers need to understand the move from keyword-based search to topic and meaning-based search using Large Language Models (LLMs). This shift focuses on giving users straightforward, full answers on the search results page, which subsequently reduces click-through rates in most cases.
To get results from AI search, you need to make content that is trustworthy, and demonstrates expertise, while ensuring that it is focused on the user, and also easy for AI to understand. This means giving clear answers at the start of sections, using headings in a structured way, keeping facts accurate, quoting authoritative sources and using a conversational tone of voice.
For B2B, continuing to drive traffic to your website will involve creating clear, topic-based, expert content that AI can point to, so that potential buyers can dig deeper on your website. Checking how you appear in AI Overviews, looking at traffic, and using tracking tools is key to changing plans as the technology evolves.
To keep up with current and future changes, we will all have to keep learning and adapting as AI search grows. By getting on board with these shifts and producing useful, user-focused, structured content, marketers can stay visible and relevant in the world of AI search in order to achieve user attention and real engagement.
