How To Dominate AI Search in 2026

Is your Google Search Console showing impressions climbing whilst your clicks quietly fall off a cliff?

You’re not imagining it, and you’re not doing anything wrong. It’s happening to almost every website in almost every sector — and it’s the clearest sign yet that the rules of search marketing have fundamentally shifted.

In this session, Charlie Marchant, CEO here at Exposure Ninja, presented our agency’s State of AI Search report findings alongside a practical playbook for getting your brand visible in AI answers across ChatGPT, Google’s AI Overviews, Perplexity, and more. This article covers everything she shared — the forces behind the traffic drop, the three pillars of AI search, and the specific steps your team can take to turn things around in 2026.

If you’re a senior marketing leader trying to make sense of why the game has changed and what to do about it, this is the clearest picture of the current situation we can offer.

Executive Summary

• AI is eating your top-of-funnel traffic. Three forces are suppressing organic click-through rates: AI Overviews on Google, AI chatbots like ChatGPT, and Google’s AI Mode. AI Overviews alone correlate with a 34.5% reduction in click-through rates.
• The buyer journey is compressing dramatically. What previously took 15–30 days and 20+ touchpoints can now happen in a single 5-minute conversation on ChatGPT — taking a prospect from awareness to a personalised product recommendation without ever visiting your website.
• Top-of-funnel content now exists to influence AI answers, not to generate clicks. The old mindset — more articles equals more sessions equals more conversions — no longer holds. Visibility inside AI answers is the new metric that matters.
• 82% of AI citations come from third-party websites. Your own site is only referenced around 9% of the time. Digital PR, reviews, and external mentions are now the biggest single lever in your AI search strategy.
• There are three pillars to address: technical foundations, clear positioning and content, and digital PR and citations. All three are measurable, and real results are already being seen — including a 243% increase in AI platform traffic for one Exposure Ninja client.

Why Are Your Impressions Up but Your Clicks Down?

The divergence is happening everywhere. Impressions rising, clicks falling — not just in one sector, but across the board. The cause is a set of interconnected AI forces that have changed what users see when they search, and what they actually click on.

The first and most significant force is Google’s AI Overviews — the AI-generated summaries that now appear above organic results on most informational and many commercial searches. They dominate the above-the-fold experience on mobile in particular, and their impact on traffic is well-documented.

“AI Overviews correlate with almost a 35% reduction in click-through rates. This is significantly huge. Some industries are affected more than others, and it overwhelmingly affects top-of-funnel searches and some middle-of-funnel searches.” — Charlie Marchant, CEO, Exposure Ninja

The second force is the rise of AI chatbots as primary search interfaces. ChatGPT is now a top-5 website globally by monthly visits, sitting just behind Google, YouTube, Facebook, and Instagram. This is not a niche tool used by technologists — it is a mainstream search behaviour that is accelerating rapidly. According to Exposure Ninja’s State of AI Search Report, 75% of people are using AI platforms more than they did a year ago, and 80% say at least half of their purchase decisions are made within AI tools.

The third force is Google’s own AI Mode — a ChatGPT-style conversational interface available directly within Google, increasingly promoted in search results and especially in Google Discover on mobile. Google is actively pushing users towards it because it knows the direction of travel.

The result is a search landscape where your organic content is being seen by more people than ever — impressions are up because AI Overviews are drawing on your pages — but fewer of those people are clicking through, because AI is answering their question without them needing to visit your site.

This is not a temporary blip. It is the new baseline. The question is what you do about it.

How Has AI Changed the Buyer Journey?

For years, the buyer journey moved through a predictable funnel. Awareness at the top — broad, informational searches. Consideration in the middle — comparisons, reviews, alternatives. Decision at the bottom — brand-specific terms, pricing, demos. Each stage happened over days or weeks, across dozens of touchpoints, with your content playing a role at each one.

AI has compressed that journey significantly — and in some cases, collapsed it entirely into a single conversation.

Charlie illustrated this with a skincare example. A user can open ChatGPT, describe their skin concern, and receive not just an explanation of potential causes but a personalised product recommendation, a comparison of specific ingredients, and a ranked shortlist of products — all before they’ve visited a single website. What traditionally took 15–30 days and 20 or more touchpoints can now happen in five minutes.

“The top of funnel is existing to influence AI answers, not to generate clicks. This is the new challenge that we have, the new challenge we need to address in our search marketing strategies.” — Charlie Marchant

This creates a profound change in how you should think about content. Publishing top-of-funnel articles to drive traffic to your blog no longer works the same way. Those articles may still be useful — but their primary value is now whether they get cited in AI answers, not whether they generate a click. A brand that appears repeatedly in AI responses during the research phase of a buyer journey builds familiarity and trust even when no click occurs. Charlie called this brand stickiness: the searcher sees your name and logo repeatedly across AI answers and organic results, begins to associate you with reliable answers on that topic, and arrives at your website later in the funnel already warmed up.

The implication for marketing leaders is clear. The job of top-of-funnel content has changed. You are now writing to influence AI, not to attract clicks. Measuring success purely through sessions and organic traffic will increasingly misrepresent your true search visibility.

What Are the Three Pillars of AI Search?

Getting your business to show up in AI answers — and to be spoken about accurately and positively — requires work across three interconnected areas.

Pillar 1: Technical Foundations

None of the content or PR work matters if AI can’t access and understand your site. Charlie was direct about the hierarchy: technical SEO is the base of the pyramid. If it fails, everything built on top of it fails too.

The key requirements are a clean website architecture — logical, well-organised, easy to navigate — so that AI crawlers can traverse your pages in the same way a human would navigate a library. Schema markup is also critical: AI reads the structured data on your pages to understand what you do and what each page is about. Missing or outdated schema is a missed opportunity to communicate clearly with AI systems at a technical level.

Broken links and 404 errors present a specific problem in the AI era. When AI agents browse websites autonomously — as in ChatGPT’s agent mode — they behave like impatient users. They exit broken pages without exploring further. A site riddled with errors is less likely to be cited as a reliable source.

One often-overlooked issue Charlie flagged: Cloudflare. If your site uses Cloudflare and the settings haven’t been reviewed in the past year, there is a reasonable chance that AI crawlers are being blocked by default. Cloudflare updated its default crawl settings approximately eight months ago, and many businesses missed it.

Pillar 2: Clear Positioning and Content

AI looks across your entire online presence — your website, your social profiles, your PR mentions, your reviews — and tries to assemble a coherent picture of what your business does. If you describe yourself five different ways across five different platforms, AI becomes uncertain about the source of truth and may represent your business inaccurately or inconsistently in answers.

The fix is simple in principle: one clear sentence describing what your business does, used consistently everywhere.

Content planning for AI search also works differently from traditional SEO. Charlie framed it as a Kill, Keep, Create decision for every piece of content on your site:

• Kill content that is generic, repetitive, or so top-of-funnel that AI can summarise it perfectly without needing your page — “What is X” explainers with no brand perspective, rehashed definitions, anything with no connection to what your business actually sells. This content may be actively confusing AI about your topical focus.
• Keep and update high-performing content and mid-performers worth improving: comparisons and alternatives, pricing and use-case pages, reviews and awards, high-intent FAQs. These are exactly what AI chooses to cite.
• Create entity-rich reference pages — about specific products, ingredients, concepts, or your brand itself — original data and case studies, opinionated frameworks, and best-of articles. These formats get surfaced repeatedly in AI answers.

“Judge your top-of-funnel content by its potential to be cited in AI, not only the traffic potential. You want to be thinking about both how likely it is to get traffic from Google, and how likely it is to be visible in AI.” — Charlie Marchant

Content structure matters too. AI uses a technique called query fan-out: when a user submits a complex natural language prompt, AI doesn’t just search for those exact words. It breaks the query down into multiple related sub-queries and searches across all of them simultaneously. A prompt about things to do in Nashville this weekend might fan out into independent bookstores, breweries, parks, off-the-beaten-path restaurants, and food tours — none of which appeared in the original query. For your content to be surfaced in answers your business cares about, it needs to have the depth and breadth to answer those downstream queries, not just the surface-level keyword you originally targeted.

Charlie’s colleague Abigail Johns, our Head of Digital PR and Content, has published a blog post template covering exactly what AI-optimised content needs to look like. The broad structure includes a heading that combines primary keyword with unique value, an introduction that answers the core question within the first few sentences, a table of contents with jump links, clear natural-language subheadings, key data or quote callout boxes, concise paragraphs, multimedia elements, an FAQ section with schema markup, and CTAs.

Pillar 3: Digital PR, Citations, and Reviews

This is the largest and, in many ways, the most important of the three pillars — and the one most businesses have underinvested in. The reason is sobering: when AI generates an answer that cites sources, it draws on third-party websites 82% of the time. Your own website accounts for only around 9% of citations. URLs that include your brand name account for another 9%.

This means that for all the work you do on your own site, the majority of your AI visibility is determined by what other people say about you elsewhere on the internet.

Charlie was clear about what this means in practice: digital PR is now a direct AI search tactic, not just a brand awareness or backlink strategy.

How Do You Get Into AI Overviews?

AI Overviews on Google represent the most visible and most immediately actionable AI search opportunity for most businesses. Charlie walked through a five-step workflow using Semrush* that any marketing team can implement.

Step 1: Find the AI Overviews you already own. In Semrush’s Organic Research tool, filter by Domain Ranks and AI Overview. This shows you the queries where your content is already being cited in an AI Overview, so you know what to protect and maintain.

Step 2: Find the AI Overviews you don’t own yet. Switch the filter to Domain Doesn’t Rank and AI Overview. This gives you a list of queries where an AI Overview appears but your site isn’t the cited source — your target list.

Step 3: Filter to top-10 positions. AI Overviews are significantly more likely to surface content already ranking on the first page of Google. Focus your efforts on keywords where you are already in the top 10 — these are the opportunities where a content update is most likely to win you the AI Overview citation. Keywords where you’re not on page one require an SEO campaign first.

Step 4: Build your optimisation list. Export the filtered results: keyword, position, current URL, volume, and intent. Prioritise keywords with both commercial and informational intent — these are more likely to attract buyers further down the funnel, not just researchers.

Step 5: Track your owned AI Overviews period-on-period. As you update content and make changes, monitor the count of AI Overviews your domain owns across your target regions. One Exposure Ninja client tracked this across US, UK, and India and saw increases of 107%, 314%, and 115% respectively between Q2 and Q3.

The key technical principle behind all of this: content that already ranks well on Google has the highest chance of also winning the AI Overview. It’s not a separate game — it’s an extension of the same one.

Why Does Content Strategy Need to Change for AI?

The standard keyword-first approach to content — identify a search term, write a page, measure traffic — is not wrong, but it is incomplete for AI search. AI doesn’t rank pages by keywords. It identifies brands it considers authorities within a topical area and surfaces information from them across a wide range of related queries.

This changes what content success looks like. A single well-ranking blog post targeting one keyword is less valuable in an AI context than a cluster of well-organised, interlinked content that covers an entire topical area with depth and authority. A veg box company that publishes on farming practices, seasonality, recipes, environmental impact, supply chain transparency, and customer stories is building AI topical authority in a way that a single “best veg box delivery” page never could.

Researching what prompts your customers are using in AI is the starting point Charlie recommends. Average Google searches run to 3–4 words. Average ChatGPT prompts average around 60 words, with far more context, nuance, and natural language. Your content needs to be answering the way people actually ask questions in AI, not just the keywords they typed into Google in 2022.

The tool she recommends for finding those prompts is Peec AI*, which allows you to input your website and receive a suggested list of prompts to target — longer-tail, conversational versions of the queries relevant to your business. Perplexity is also a useful free tool for understanding query fan-out: put in a prompt you want to rank for, and Perplexity will show you the sub-queries it searches in the background to assemble its answer. That breakdown is your content brief.

Why Is Digital PR the Biggest Lever in AI Search?

When 82% of AI citations come from third-party sources, the logic follows directly: the most efficient way to become visible in AI answers is to become the brand that authoritative publications write about.

Charlie shared two examples involving ZUGU, a client our team ran a digital PR campaign for. Articles were placed in ZDNet and Wired featuring the brand’s iPad Air cases. Perplexity cited the ZDNet article the day after publication. Gemini cited the Wired article the day after that. Months later, when Charlie ran the same searches again, both articles were still being referenced in Google’s AI Mode, with ZUGU’s cases dominating the results for iPad case queries.

The overall result: a 243% increase in AI platform traffic for ZUGU, and 123% over their revenue target.

The strategy that produced these results had three components: a technical SEO audit that brought site health from 54% to 78%, product and collection page optimisation, and a targeted digital PR campaign focused on tech and Apple-focused publications with product-led pitches. All three pillars working together.

For businesses looking to build their own citation profile, Charlie outlined several approaches:

• Journalist request platforms — Qwoted and HARO (Help a Reporter Out) allow you to respond to journalists actively looking for sources on relevant topics. These placements tend to be high-quality and appear quickly.
• Direct outreach to editors — particularly effective for getting included in regularly updated “best of” lists that AI is likely to continue citing over time.
• Press releases and PR stories built around original data — if you have customer data, internal research, or case study results, packaging these as news creates citation-worthy content that journalists will link to.
• Partnerships, affiliates, and influencers — first-person accounts and reviews from real users are heavily favoured by AI systems, making this channel more strategically valuable than it may have appeared previously.
• Awards, directories, and customer reviews — all contribute to AI topical authority, particularly when they cluster around the specific areas you want to own.

A key strategic point: digital PR for AI search needs to be topically focused. Getting mentioned in a vague lifestyle article is less valuable than being cited in a comparison piece directly relevant to your sector. AI is pattern-matching across your citation profile to determine what topical area you have authority in. The more your citations cluster around your core area, the stronger that authority becomes.

If you want to understand which publications you’re currently being cited in — and which competitor-relevant articles you’re missing from — tools like Peec AI and Profound can pull that data automatically, showing you a list of sources across all relevant prompts and flagging where you appear and where you don’t.

How Do You Measure AI Search Results?

The perception that AI search is impossible to measure is a myth. The data is available — it just requires some setup.

Google Analytics 4 — AI referral traffic and conversions. In Google Analytics 4, you can create a custom channel group called “AI Platform Traffic” by adding conditions for each major AI platform: chat.openai.com, chatgpt.com, gemini.google.com, copilot.microsoft.com, edgeservices.bing.com, and perplexity (contains). Once set up, this channel will appear in your Traffic Acquisition report and track both sessions and conversions attributable to AI referrals. This captures clicks directly from AI platforms — it won’t capture “dark traffic” where someone researched in ChatGPT and later typed your URL directly into a browser, but it gives you a real, trackable baseline.

Looker Studio — AI traffic dashboards. If you’re already using Looker Studio for client or stakeholder reporting, the same GA4 data can be surfaced in dashboard form, tracking AI platform sessions over time and splitting by source. For eCommerce businesses with revenue tracking configured in GA4, this will also show AI-attributed revenue and transactions. In one example Charlie shared, AI platform sessions had grown by 3,937% year-on-year for a tracked client.

Prompt visibility tools — Peec AI, Profound, Scrunch AI. These platforms work by running your target prompts across multiple AI chatbots at scale and returning a visibility score: for the answers your business cares about, how often does your brand appear? This mirrors keyword rank tracking in traditional SEO and gives you a comparable KPI to track over time. For businesses just starting out, Peec AI offers a free trial and can be useful for generating an initial prompt list as well as tracking visibility.

AI sentiment — Profound. Beyond visibility, AI sentiment tracks how positively AI answers describe your brand. Charlie showed an example from Sandals Resorts: 72.8% positive sentiment, with AI citing honeymoon perks and streamlined booking experience. The 27.2% negative sentiment was almost entirely around venue restrictions and extra costs. That breakdown tells the business exactly what objections to address in their content and PR work to improve their AI representation over time.

On reporting, Charlie’s recommendation is to anchor everything to your bottom line. Traffic metrics matter less than conversion metrics. If AI is driving highly qualified visitors to your website — people who have already done their research in ChatGPT and arrived ready to enquire — your conversion rate from that traffic should be strong, even if the volume appears modest. Focus on leads and revenue, not just sessions.

If you’re unsure where to start with any of this, requesting a free marketing review at exposureninja.com/review will give you a clear view of your current AI search visibility and the most significant opportunities for your specific business.

Next Steps

This Week

• Check whether AI crawlers can access your site. If you use Cloudflare, log in and verify your crawler settings. If AI bots have been blocked by the default settings change that rolled out approximately eight months ago, this is your single highest-priority fix.
• Set up your GA4 AI Platform Traffic channel group. Create a custom channel in GA4 Admin → Channel Groups, adding conditions for each major AI platform source. This gives you a baseline and makes every future improvement measurable.
• Search for your business in ChatGPT and Perplexity. Ask the questions your customers ask about your category. Does your brand appear? Are competitors named instead? What does the answer say about your business if you do appear? This is your current AI visibility audit.

Next 30 Days

• Run a Kill/Keep/Create review of your top 20 content pages. Use GA4 to identify what gets zero traffic and no rankings, and begin pruning content that creates topical confusion. Update your best-performing content to meet AI-optimised content structure standards.
• Audit your schema markup. Check your highest-priority service and product pages in Google’s Rich Results Test. Update and expand schema where it is missing, outdated, or minimal.
• Run a Semrush AI Overview gap analysis. Follow the five-step workflow to identify keywords where an AI Overview exists but your site isn’t cited. Prioritise your top-10 ranking pages as the first optimisation targets.

Next 90 Days

• Develop your prompt targeting list. Using Peec AI or a similar tool, identify the conversational prompts your ideal customers are entering into AI platforms. These become your content and PR targeting brief.
• Launch a targeted digital PR campaign. Identify the publications currently being cited in AI answers for prompts relevant to your business. Prioritise getting into regularly updated “best of” and comparison articles in those publications — these are the citations most likely to drive ongoing AI visibility.
• Request a free marketing review at exposureninja.com/review to get Exposure Ninja’s full diagnostic on your current AI search visibility, technical foundations, and the specific opportunities most likely to move your numbers.

Resources Mentioned

• Exposure Ninja State of AI Search Report — The research report Charlie drew on throughout this webinar, covering AI platform usage statistics, consumer purchasing behaviour in AI tools, and benchmarks for AI search visibility. Free download at exposureninja.com/state-of-ai-search-report.
• Semrush* — Used for the AI Overview gap analysis workflow (Steps 1–5). The Organic Research tool allows you to filter by AI Overview presence and domain ranking status to identify keyword opportunities.
• Peec AI* — Prompt tracking and citation analysis tool. Recommended for generating a target prompt list and tracking your brand’s visibility across AI chatbot answers at scale. Free trial available.
• Profound — AI visibility and sentiment tracking platform. The Sandals Resorts sentiment analysis example in the webinar was from Profound. Tracks how positively AI platforms describe your brand and breaks down themes in positive and negative mentions.
• Scrunch AI — An alternative prompt tracking and citation analysis tool, similar in function to Peec AI.
• Answer the Public — Useful for mapping the query fan-out around a topic. Charlie recommends it as a starting point for understanding the sub-queries AI is likely to run when a user puts in a broad prompt relevant to your business.
• Qwoted — Journalist request platform for responding to media queries and securing placements in relevant publications. A low-effort way to begin building the third-party citations that AI favours.
• HARO (Help a Reporter Out) — Similar to Qwoted. Connects brands and subject matter experts with journalists looking for sources.
• Looker Studio — Reporting platform used to visualise AI platform traffic and conversions from GA4, including sessions over time and platform split.

In Conclusion

The search landscape has shifted enough that continuing to optimise purely for Google clicks is no longer a complete strategy. AI Overviews, ChatGPT, and Google’s AI Mode are intercepting the research process at the top and middle of your funnel. Your customers are making significant parts of their purchase decisions before they ever visit your website. The brands that end up recommended in those conversations are winning customers you’ll never see in your Google Analytics.

The good news is that this is not a mystery. The three pillars — technical foundations, content and positioning, and digital PR — are concrete, measurable, and actionable. The ZUGU results (243% AI traffic increase, 123% over revenue target) are a proof point, but they are not unusual. Businesses that move on this early while most of their competitors are still watching their click-through rates decline will have a structural advantage that compounds over time.

The first step is knowing where you stand. Request a free marketing review and our AI Search team will give you an honest assessment of your current AI visibility and the clearest path to improving it.

Is your marketing underperforming?

Request a website and marketing review and our team will tell you how to improve your marketing.

Request a Free Review

*Disclaimer: Exposure Ninja may get a commission through the marked links above, at no cost to you