AI search behavior: fewer clicks, better leads?

A thorough SEO study for CMOs & B2B marketing teams – by SEOritmo

Open your organic Google Search Console today and the story might look ugly: impressions up, clicks down, CTR slipping, branded queries doing weird zig-zags.

If you just skim the charts, it feels like demand is evaporating. But AI search behavior is not that simple, and in most B2B funnels, it’s not that catastrophic either.

In this whitepaper, we observe how the zero-clicks phenomenon might not be damaging  qualified conversions and new business generation as we think.

Clicks are down and look scary. Except they aren’t. 

Over the last 12–18 months, Google has started answering complex, multi-step questions directly on the SERP through AI Overviews and AI Mode.

At the same time, LLM assistants like ChatGPT, Perplexity or Claude have become a probabilistic-search layer where buyers test ideas, shortlist vendors and refine requirements before they ever click a blue link.

The impact on clicks is real for basically every industry out there.

Here are some raw data:

• When an AI summary appears, only about 8% of users click a traditional result, versus 15% when there is no AI summary at all. ¹
• Clicking links inside the summary itself is even rarer, at roughly 1% of visits. ¹
• In parallel, SparkToro’s 2024 zero-click study shows that for every 1,000 Google searches, only around 360–374 clicks go to the open web, and nearly 60% of searches end without any click. ²
• Chat GPT’s response block produces up to 100× the impressions of other surfaces, but still achieves only a 0.01%–1.6% CTR. ³

So yes, users are clicking less.

But that doesn’t mean buyers have disappeared or that organic has stopped working.

Let’s think of the B2B funnel, for instance.

We all know how important the two extreme layers of the funnel are in a typical B2B organic pipeline (entry points at the top, and deal closers/ branded interactions at the bottom).

While everything in the middle is often a fragmented back and forth across multiple touch points.

When AI comes in between this process, one can mainly observe:

• Compression (not deletion) of the upper funnel: for instance, people can ask broad “what is” and “how should we choose” questions inside AI interfaces
• Engaging with the website or keeping an AI-based touch point in mid-funnel activities: decision makers iterate on constraints like budget, tech stack or region, and only click when they’re closer to a decision. By the time they land on the website, they’re more educated, more opinionated and more ready to talk to sales.
• Re-engaging with the website directly from an AI answer, or begin a hybrid/ offline sales interaction as usual: 

In other words, fewer clicks from AI-affected search results do not automatically mean lost demand.

In fact, those fewer clicks might even be better clicks.

We’ll attempt to prove that in the following sections.

Our job now is to stop obsessing about raw traffic and start measuring what matters in an AI-mediated journey: post-AI Overview conversion rate, AI-assisted conversions, branded vs non-branded mix and, above all, lead quality and pipeline.

Why fewer clicks can mean later-stage, higher-intent visits

On classic SEO dashboards, fewer sessions almost always looked bad. In an AI-first world, that assumption doesn’t make sense anymore.

Qualified and motivated buyers simply could arrive later in their journey.

After the organic click, users have already compared categories, filtered options by budget or stack, and clarified what “good” should look like for them.

The story you want to tell internally is simple:

“AI search behavior is not cutting off informational demand, it is compressing the top of the funnel.

You are trading some anonymous early research visits for fewer, better clicks that are closer to opportunity and revenue.

Your organic analytics are lying to you: what to look at instead

Once you accept that AI search behavior compresses the top of the funnel, the question becomes simple: what should we measure instead of raw clicks and sessions?

In practical terms, we need to separate three lenses:

• How pages perform when AI Overviews are present/ when an LLM recommends them
• How AI-assisted journeys show up in analytics
• How the organic mix changes across brand, non-brand and pipeline

The goal is not to build a brand new reporting stack.

It is to rewire the questions you ask of the data you already have in GA4, Google Search Console and your CRM, ideally through an AI-based SEO framework.

You read that right: to spot patterns hidden in the AI mess, you want to use AI itself.

In a Chat GPT’s query fan-out experiment for a client in the domain reselling industry, for instance, we combined human observation with proprietary AI projects to spot and surmise content gaps (derived from the fanout sequence).

At a surface level, what we did is:

1. Explored ChatGPT’s process after a BoFU query to compare domain service providers, observing its grounded search, thinking and comparison process, down to the user asnwer
2. Used GPT model 5.1 to aggregate patterns and spot evident gaps between the fanout sequence – results and the brand content we want to optimize for.
3. Isolated candidate topics, content structures, and quick technical wins that we could for tackle client’s brand and presented them as actionable recommendations, then:
   • Used SEOritmo’s Tech Bender to create related schema markup assets at scale
   • Used SEOritmo’s Make-Rev function to create SEO-optimized content drafts over those same recommendations   

The illusion of query-based “traffic loss” in SERPs

When AI Overviews or AI Mode appear, they rearrange how clicks are distributed on the SERP.

Depending on your industry, this might be more or less impactful, but the CTR loss to expect is substantial regardless.

AIO appearance by industry, Advanced Web Ranking – updated 11 Dec 2025

In fact, position 1 can lose a big chunk of its historical CTR even if your ranking does not move and your content is still the primary source for the AI answer.

From inside your Google Search Console dashboard, this may simply show up as “less traffic for the same keyword.”

To make things messier, Google Search Console does not break out AI Overview impressions or clicks as a separate performance surface (we hope that, one day, Mr. Google will listen to the entire SEO community’s cry), meaning impressions and clicks from AI blend into traditional organic data.

Bottom line, we cannot see how much of the journey took place inside AI Overview before the click. Yet, what is actually happening is a new layer of untracked, pre-click education.

For B2B marketing teams, the symptoms tend to look like this:

1. Sessions down on mid-funnel content, but demo or consultation requests are steady or increasing (if this is not the case, your brand has probably other problems, which we are happy to talk about and study)
2. Branded search impressions up while branded clicks are flat or down, suggesting AI is answering most lightweight brand questions directly.
3. A growing block of “direct” traffic landing on deep-funnel pages, which often hides users coming back from AI chats, shared links or saved results.

So, the hard task of CMOs becomes reframing the story from: “organic traffic is down” to “AI search behavior is hiding a chunk of the journey upstream”.

For this, your SEO specialists will likely have toupgrade their analytical skills when working with GA4 and Google Search Console data, or when working with the Google integrations of the SEO tool in use.

1 – Traffic down, pipeline up

Imagine your non-brand organic sessions drop 20% after AI Overviews roll out for several key queries. At the same time, the percentage of those visits that request a demo jumps and, maybe, the average deal size from organic increases too.

The old narrative would call this a traffic problem. The AI search behavior narrative calls it a qualification upgrade.

2 – Branded search spike, non-brand flat

You keep publishing mid-funnel guides and comparison pieces. Non-brand clicks stay flat or decline slightly, but branded impressions and brand-plus-keyword combinations grow.

This is a classic sign that AI systems are leaning on your content to educate users earlier, and that those users later return with brand-aware queries when they move into vendor selection.

3 – Direct and organic mix shift

Deep-funnel pages such as pricing, implementation or integration docs start receiving more “direct” sessions and more repeat visitors.

In reality, a slice of these are people who first discovered you inside an AI answer, then clicked later through a saved link, a shared chat or a brand search.

On the surface it looks like organic lost visibility. Underneath, organic content is still doing the work, just through AI-mediated discovery.

Post-AI Overview conversion rate

If AI Overviews and AI Mode are taking some of the exploratory clicks, you want to know whether the remaining clicks are better at converting.

The most direct way to see that is to isolate topics and pages that are frequently affected by AI Overviews, then compare their performance to similar, non-AIO topics.

Let’s keep it practical. Here’s a simple workflow you can implement today:

1. Build a short list of strategic queries where you know AI Overviews often appear, especially mid-funnel questions like “how to choose [category] vendor” or “best [solution] for [industry].”

Note – We know, that’s pretty basic type of content there; but basic doesn’t mean bland, unoriginal, or not-working.

Your content team can produce SEO content templates for VS articles, comparative listicles, or even editorial assets with automatic matchof on-page, technical and E-E-A-T best practices.

Just fill the brief, hit play and watch it appear directly in your WordPress or WebFlow website.

2. Map those queries to the landing pages you are trying to rank with, typically comparison pages, solution landings or high-buying-intent guides.

Note – This is oversimplifying, but one simple approach is to filter out exact or similar queries in Google Search Console, check which landing pages show up for each, then investigate user behaviour on them in GA4.

You can exploit a pre-built retrieval of the pages for each queries via SEOritmo’s Google Search Console integration.

3. In GA4, create a segment for sessions landing on those pages from organic search and track key events such as demo requests, trial signups or high-value content downloads.

AI-search-assisted conversions

The hardest part of AI search behavior is that the pre-click journey often happens outside your analytics stack.

Buyers ask questions in AI Overviews or LLM chats, refine their requirements, then arrive through what looks like a “clean” direct or brand search click. On paper, it seems like search played a tiny role. In reality, AI search did most of the education.

Thus, you cannot track every AI interaction. 

Yet, another tactic we use to get as close as possible to this data is using some proxies to understand AI-assisted conversions.

We are talking about:

• BoFU organic or branded interactions
• Spikes in brand+keyword queries
• Returning visitors on BoFU pages

Short paths with late organic or brand interaction

In a GA4’s path exploration, look for journeys where the user:

1. Enters as “direct” or through a vague search
2. Views a small number of high-intent pages
3. Converts more quickly than usual – we recommend Andy Crestodina’s framework to isolate the date and time of conversion when possible isolating timestamps on “thank you”-pages visits, when applicable:
   1. Open a page path report.
   2. Set a long date range (for example, the past 12 months).
   3. Use the search bar to find the URL of your thank-you page (or however the you called named the slug)
   4. Click the blue “+” above the first column to add a secondary dimension, then choose “Date + hour.”
   5. Export the data using the share icon in the top-right corner and download it as a CSV file.

Sudden increases in branded-plus-keyword queries

When you start surfacing in AI Overviews for “how to choose [category]” type terms, you will often see growth in queries like “[your brand] vs [competitor]” or “[your brand] pricing.”

Those can be AI-conditioned buyers coming back to validate what they saw in the AI answer.

Higher share of returning visitors on deep-funnel pages

People who explore through AI might click your brand, bounce quickly, then return later through a saved link, a shared chat or a direct visit.

In GA4, keep an eye on both channel source and returning users on pricing, implementation and integration pages.

Branded or non-branded: that is the matter

Before AI search behavior changed the rules, many SEO teams treated branded and non-branded traffic as separate universes.

Non-brand was about discovery; brand was a loyalty or performance marketing problem. With AI sitting between the user and the SERP, those lines blur.

AI Overviews and LLM assistants are now doing a lot of the non-brand discovery work.

If your content is strong, they will mention your brand, quote your frameworks or include your product in shortlists. 

The question is not simply “how much non-brand traffic did we get,” but “how healthy is the bridge from AI-mediated discovery to branded intent.”

Practically, that means:

• Tracking trends in pure non-branded queries, branded queries and mixed queries (“[brand] for [use case]”, “[brand] review”, “[brand] vs [competitor]”) 

Note –  Your SEO team can do that with any SEO tool, and even with Google Search Console by using filters smartly.

• Noticing when non-brand clicks are flat or declining, but branded and mixed queries are growing. That can be a sign your brand is being injected into AI answers even if you do not see all the upstream impressions.
  
• Qualitatively reviewing which branded query variants are growing.

Speaking of B2B SEO use cases for the SaaS or IT industries, “reviews,” “pricing,” “implementation” and “integration” terms can signal high purchase intent, other than just generic “[brand] tool.”

The job of a SEO specialist here is to ensure that your category content, comparison guides and thought leadership are present in the sources AI systems trust, so that this bridge keeps getting stronger.

Lead quality and pipeline contribution: does AI move the needle?

At some point, every AI search conversation comes back to the same question: does this help us close better deals or not?

If AI search behavior means fewer but more qualified clicks, the way to prove it is through lead quality and pipeline contribution.

Start with simple comparisons:

• Segment opportunities in your CRM by primary acquisition channel or first-touch channel.

• Within organic, isolate referral traffic that comes from AI tools and bring down to where that generated leads.

You can do that in several ways in GA4, for instance:

1. Use the “Page referrer” and “Landing page + query string” as dimensions, and “Sessions” or “Users” in an Explore chart
2. Set a regex to filter out page referrers like: 

.*(chatgpt\.com|openai\.com|claude\.ai|perplexity\.ai|mistral\.ai|gemini\.google\.com|copilot\.microsoft\.com|copy\.ai|meta\.ai|meta\.com/ai|x\.com/i/grok).*

3. Isolate landing pages that are meaningful for your conversions/ lead generation in the same or in another chart and check if they generated an applicable key event or transaction

• Compare metrics like opportunity-to-win rate, average deal size, sales cycle length and churn or expansion rates for organic-origin deals vs other channels.
  

If you see that organic deals from AI-affected topics close at a higher rate or carry larger amounts, you have concrete evidence that AI search is filtering for more serious buyers.

That is the kind of data that wins budget conversations.

How to adapt your content strategy for AI search behavior

If AI search behavior is compressing the funnel, your content has to do two jobs at once.

1. It needs to be quotable and trustworthy enough for AI systems to use as a source
2. It must be clear and persuasive enough for human buyers arriving later in the journey. 

That means moving from blog posts to structured, opinionated resources that map directly to how buyers research and buy.

Let’s translate that into what stays and what changes in terms of content production.

SEO content plans, topical and semantic authority, and best practices for B2B

To create successful content in the AI era, you still need a proper SEO content plan, a strategy to tackle topical gaps of your industry, and a series of updated, technical best practices that can both help your organic results and your generative engine optimization (GEO) efforts.

SEO content plans: topical pillars and more weight to BoFU 

When building the editorial plan, we don’t want to change the premises: building content pillars, subtopics and children articles that respond to actual user search data and continuously fortify our brand signals towards the algorithms. 

We say “continuously” because, assuming quality and technical cleanliness are there, recency plays a big role: about 65% of AI bots visits were for content published in 2025.

In other words, to maintain frequency and recency, keep creating content.

When it comes to organizing topics and keyword clustersby intent, the new B2B scenario sees “how to choose”, “implementation problems”, “X vs Y”, “success stories”, and “lists” taking more space in the plan because, if the web user is at a later stage of the funnel, it’s only reasonable to have more content that addresses decision-making points.

To do that, we want to audit content not just against product representation, but also and mostly against the real questions our customers ask (in sales calls, in search, across social communities, etc.).

Informational topics cannot disappear either. True, we are observing zero-click phenomena mainly around ToFU content (68% of LLM users use the platforms to conduct research and summarize information ⁵), but this doesn’t mean stop producing it.

In fact, if anything, we want to evolve informational content crafting for it to focus it more on AI-readability, to increase the chances of our brand getting cited in the LLM’s answer.

Keep creating content.

Remember, you can use an SEO AI content tool to accelerate 7x editorial production by creating templates that automatically fit best practices.

For that, exploit SEO formats across specific types of blog posts, like VS articles, editorial pieces, and listicles directly in your CMS.

Authority, trust and other signals that AI considers

Search algorithms rely heavily on signals of expertise, experience and trust, that didn’t change. 

AI systems do the same.

As iPullRank beautifully puts it: 

“Your visibility depends on how each platform curates and filters sources; you may be excluded from the answer entirely if your content doesn’t match the platform’s editorial signals.” ⁶

In practice, that means you want to surface the same things that would convince a cynical buyer: who is behind the content, what proof you bring and where you have been referenced before.

The following are not your typical bland recommendations we snatched off of whatever AI output. They are real, tested and – we venture to say – evergreen best practices that have demonstrated impact on organic AI results and search in general.

Alone, they mean nothing.

Together with valuable content and other SEO elements in place, they do what they have to do: make your content awarded by search engines.

E-E-A-T still rules

AI systems lean on E-E-A-T signals to decide which sources to trust and quote.

That means showing :

• Real experience (who actually did the work)
• Clear expertise (who is writing and why they are qualified)
• Visible authority (recognition, case studies, third-party mentions)
• Strong trust markers (transparent policies, up-to-date content, secure site).

The more your pages reflect to be coming from real field experts, the more likely they are to influence AI search behavior.

Disclaimers are not just legal components

When publishing content like “best of” lists or vendor comparisons, disclaimers do more than keep legal teams happy.

They signal that your methodology is transparent (what you evaluated, what you did not, any commercial relationships) and that your content is opinionated but fair.

This helps users and AI systems read your content as guided judgment, not hidden advertising.

Meta tags and structured data (without the heavy lifting)

Meta elements (titles and descriptions) might be one of the most simple, and yet overlooked, SEO asset companies have. Without them, one will miss critical SERP conversions and distract LLMs’ attention.

In a moment of our story where AI is on everyone’s lips, we should be allowed to generate perfect meta tags at scale and in bulk.

To achieve that, you can use technical SEO tools that not just generate descriptive elements, but also consider keywords and search intent.

Adding a schema markup where it makes sense, especially for FAQs, how-to content, product details and reviews, is another best practice.

Schema is not a magic AI ranking button, but it increases the chances that search systems can interpret your content correctly and match it to relevant questions. 

Internal and external links are the cardiovascular system of your website

Internal links help AI (and users) understand how your expertise is structured: which pages are foundational, which ones support them and how a topic cluster fits.

When pointing to credible, selected sources, external links reinforce that your arguments are grounded in more than self-promotion.

Together, smart internal and external linking creates a web of meaning that AI systems can crawl, interpret and reuse..

If you have third-party recognition, guest posts or quotes in reputable outlets, make sure those are easy to find on your site and tied to your core topics.

Getting your team aligned on AI search behavior

You can have the best AI SEO insights in the world and still lose if your organization is stuck in a pre-AI mindset. Alignment is as much about language as it is about metrics. 

So, the point is to bring adjacent teams on the same team.

When everyone sees AI search as part of the journey, not a black box, it becomes easier to prioritize and fund the right experiments.

• Sales needs to know how prospects are discovering you through AI and which content those systems are likely to surface.
• Product marketing needs to understand which narratives and use cases resonate in AI-driven research. 

Fewer clicks, better conversations, clearer stories

AI search behavior has changed the way users interact with the digital world.

For B2B marketers, the implication is clear: stop obsessing about the volume of clicks and start obsessing about the quality of the conversations and opportunities those clicks create.

Yes, there are fewer visits for many topics. But the buyers who do land on your site are farther along, better informed and more likely to become high-quality opportunities. 

SEO teams now need to align content, analytics and storytelling with this new reality, building resources that AI systems want to quote and that serious buyers want to read. 

The following might seem contradictory but, yes, getting clicks is not inherently wrong, and we like to keep our clients’ charts nice and up.

This is a small example showcasing what SEOritmo achieved with just content optimization and technical adjustments in 4 months for a SaaS in a YMYL, highly-regulated industry.

Thus, an AI SEO software should feel like an analyst embedded in your team.

At a minimum, it should:

• Ingest your data and help you find the patterns your marketing team needs.
• Help you combine insights with tactics that respond to hyper-personalized scenarios, for your company, not just your industry 
• Help you work 70% faster with your SEO content production and technical SEO tasks
• Highlight shifts in the SEO game and in LLMs responses so you can kill two (and more) birds with one stone

Learn more about how to get all of that by contacting us.

External sources, references and quotes

Cited sources

1: Behavior & click-through data – Pew Research – link behavior with AI summaries: https://www.pewresearch.org/short-reads/2025/07/22/google-users-are-less-likely-to-click-on-links-when-an-ai-summary-app

2: SparkToro – zero-click study original: https://sparktoro.com/blog/2024-zero-click-search-study-for-every-1000-us-google-searches-only-374-clicks-go-to-the-open-w

3: Search Engine Journal – Inside ChatGPT’s Confidential Report Visibility Metrics: https://www.searchenginejournal.com/inside-chatgpts-confidential-report-visibility-metrics-part-1/561608

4: Seer Interactive – AI Brand Visibility and Content Recency: https://www.seerinteractive.com/insights/study-ai-brand-visibility-and-content-recency

5: Bain & Co. – Goodbye Clicks, Hello AI: https://www.bain.com/insights/goodbye-clicks-hello-ai-zero-click-search-redefines-marketing/ 

6: Ipullrank – User Behavior in the Generative Era:  https://ipullrank.com/ai-search-manual/search-behavior

Other sources

Advanced Web Ranking – Google AI Overview Tool: https://www.advancedwebranking.com/free-seo-tools/google-ai-overview 

Search Engine Land – coverage of the Pew study: https://searchengineland.com/google-ai-overviews-hurting-clicks-study-459434

Whole Whale – commentary on Pew findings: https://wholewhale.com/tips/googles-gaslighting-pew-research-confirms-what-seos-already-know-about-ai-overviews/

Search Engine Land – zero-click study: https://searchengineland.com/google-search-zero-click-study-2024-443869

Search Engine Land – CTR drops with AI Overviews: https://searchengineland.com/google-ai-overviews-drive-drop-organic-paid-ctr-464212

AI search & strategy – Google Search Central – AI features and your website: https://developers.google.com/search/docs/appearance/ai-features

Google Search Central blog – succeeding in AI search: https://developers.google.com/search/blog/2025/05/succeeding-in-ai-search

Search Engine Journal – tracking AI Overviews visibility and AIO impact: https://www.searchenginejournal.com/google-aio-track-visibility-serpapi-spcs/560470/

Search Engine Journal – impact of AI Overviews and how publishers need to adapt: https://www.searchenginejournal.com/impact-of-ai-overviews-how-publishers-need-to-adapt/556843/

Relevantaudience – practical guide to optimizing for Google’s AI search: https://www.relevantaudience.com/seo/optimizing-for-googles-ai-search-a-guide-for-website-owners/

Growth Marketing Pro – list of AI search monitoring tools: https://www.growthmarketingpro.com/best-ai-search-monitoring-tools/  

B2B & AI search conversion insights – Passionfruit – AI search referrals vs classic clicks: https://www.getpassionfruit.com/blog/are-ai-search-referrals-the-new-clicks

Goodie – AI search statistics and funnel implications: https://higoodie.com/blog/ai-search-statistics