Over the past couple of years, Google has steadily woven generative AI into its search experience. At 1827 Marketing, we’ve tracked these changes closely–from the 2023 announcement of the Search Generative Experience (SGE) to the rollout of AI Overviews in 2024.

At each step, Google has moved closer to becoming an “answer engine” and the ripple effects are significant. Traditional search strategies are being up-ended. A year’s worth of data now confirms what many feared. AI-powered search is reshaping how people find information—and how often they click. Impressions might be up, but click-through rates have dropped dramatically. Most searches now end without a click at all.

That’s a serious challenge—existential for some. For B2B marketers, it raises urgent questions about discovery, traffic, and content strategy in a world where Google is the destination, not the start of the customer journey.

What Are AI Overviews?

Rolled out as a default feature in 2024, AI Overviews brought the core promise of SGE to the mainstream. For an increasing number of queries, users now see a concise, AI-generated answer at the top of the search results. Google synthesises and summarises information from multiple sources, with a sidebar of clickable citations.

By May 2024, AI Overviews were live for all U.S. users, with global expansion quickly following. By year-end, Google expected over a billion people to be using them. In March 2025, it upgraded to Gemini 2.0 to support more complex queries.

The result? A more streamlined experience for users, but a new challenge for brands and publishers seeking to maintain organic visibility and traffic.

Have AI Overviews Changed Search Behaviour?

Yes–measurably. We now have a year of data on AI Overviews, and several studies have shed light on how generative summaries affect traffic, click-through rates (CTR), and user habits.

The data paints a picture of search where:

AI summaries dominate informational queries, and are growing more common by the month.
Organic clicks are down, in many cases by roughly half when AI is present.
Zero-click searches are now the norm.
Brand searches (i.e. users explicitly seeking a known source) is one of the few things insulating queries from AI cannibalisation.

Publishers and marketers need to rethink SEO strategy. You can’t just think about ranking for keywords anymore. You need to be thinking about being cited in AI answers and providing something beyond what the AI can generate.

AI Overviews Are Expanding

Semrush data shows that 13% of Google searches an AI Overview by March 2025–up from 6.5% in January. Other analyses suggest penetration is now far higher. A study by Botify/DemandSherpa at the end of 2024 found that, when conditions are right, Google’s AI Overviews can appear in up to 47% of searches.

This means that for up to half of searches, the top ranking organic result appears below the fold because the AI answer box comes first.

Informational Queries Are the Primary Target (For Now)

Around 88% of queries that trigger AI Overviews are informational. Questions like “how to…”, “what is…”, or “why does…” are prime candidates.

But they caution this is Google in their testing phase. They point out that “this is consistent with Google's previous rollout strategies for features like Featured Snippets”. Google begins with fact-based low-difficulty queries before expanding into other areas, like more commercial recommendations (e.g. “the best X”)..

Navigational or branded queries currently show an AI summary more rarely. However, the percentage of branded queries with AI Overviews has doubled from 0.7% to 1.4% between Jan and Mar 2025, suggesting Google is testing AI answers even for navigational queries.

This means that your top-of-funnel, educational content is most at risk of being summarised. Mid- and bottom-funnel remain relatively safe. For now.

Click-Through Rates Have Dropped

When Google gives the answer to a query directly, fewer users click through. The evidence is stark:

Based on an analysis of 5000 keyword searches, Mail Online saw CTR drop 56% on desktop, and 48% mobile. Even when cited in the AI Overview, users still clicked far less. Being featured as a source still led to a 44% lower CTR on desktop, and 32.5% lower on mobile compared to a regular #1 ranking.

Even when you “win” by getting your brand into the AI answer, nearly half the people who would have clicked your link might not do so.

Those results are being echoed elsewhere. An Ahrefs study found a 34.5% decrease in clicks for top-ranking pages across 300,000 keywords. SEO agency Amsive reported a 20% decline in non-branded search CTR when an AI Overview was present.

Zero-Click Search Is Prevalent

According to SparkToro, as of the end of 2024 59.7% of all European searches and 58.5% of U.S. searches ended in zero clicks (i.e. the user did not visit a website).

Semrush’s data shows a slightly different picture. Zero-click searches range between 34-47%, with the higher percentages for results with AI Overviews. However, they surmise that this was because of the informational nature of the queries. The data actually showed a slight decrease in zero-click behaviour over time, leading them to challenge the belief that AI Overviews cause an increase in zero-click behaviours. Instead, they see “a more complex interaction between query type, user intent, and how AI answers are delivered.”

Brand Searches Are More Resilient (For Now)

Branded queries–where users search for a company by name–are currently relatively safe from AI disruption. Echoing Semrush’s findings, Amsive found that only 4.79% of branded keywords triggered AI Overviews.

They also found that when branded terms trigger an AI Overview, CTR actually rises by 18.6%. Why? Probably because if someone searches “Brand X services”, sees an AI box that mentions Brand X with a link, it reinforces that they should click it.

For marketers, this highlights the value of brand demand. If users come looking specifically for you, you’re more likely to get the click.

Google’s AI Mode

These are just the effects we’ve seen because of AI Overviews. If a simple one-summary at the top of the SERPs can cause this much disruption, what happens when Google offers its next iteration of AI powered search?

Well, we’re about to find out. At I/O 2025, Google launched AI Mode: a deeper, chat-based search experience powered by Gemini 2.0 that Google’s CEO Sundar Pichai pitched as a “total reimagining of Search”.

It is Google’s answer to the question: How can we make search do all the work for the user?

AI Mode offers:

A Chat-Based Interface: AI Mode is an immersive Q&A environment. Users might not even see a traditional list of blue links unless they actively switch views. Google says that helping people to discover content on the web remains a goal, and that it will still be easy to click on sources and explore further. But when AI gives a comprehensive answer, the incentive to click those sources is likely to be low.
Query Fan-Out: Behind-the-scenes, AI breaks your complex query into multiple sub-queries and searches for each of them simultaneously. It then aggregates the results into one coherent answer. It’s not just reading and regurgitating existing content. It is actively researching on your behalf, doing dozens of Google searches for you, and merging the findings. The goal is to cover more ground and capture nuanced details, so the answer is thorough.
Multimodal Reasoning: AI Mode isn’t limited to text. Thanks to Gemini 2.5, It leverages multimodal capabilities to incorporate images, charts, or other media into responses when useful. And it can handle queries about images. You might upload a photo and ask the AI about it – think Google Lens on steroids.
Real-Time and Structured Data Integration: One limitation of early generative AI was the model training data cut-off dates. AI Mode mitigates this by integrating live data sources. It pulls from the current index, uses Knowledge Graph facts, and shopping data. Google also hints they’re working on factuality improvements and using ranking signals even within the AI answers, so AI Mode will quote trusted sources as well as providing current, accurate answers.

While still in limited release (it is an experimental opt-in limited to U.S. users as of mid-June 2025), AI Mode signals where Google wants search to go. If user feedback is good, we could see it rolled out to everyone as the new default search experience sooner rather than later.

What Will AI Mode Do To Discoverability and Traffic?

If – or more likely when – AI Mode becomes the default search experience, it marks a potential end to the old SEO contract: publish content, earn traffic. It will shift us to what some are calling the “machine web”.

Currently, content is created by humans for humans to find, click, read, and engage with. In the machine web, content is digested by AI systems, which then relay answers to users.

Google will still slurp up your content, but if users no longer need to visit your site to get the information, how do you get the value?

Google, meanwhile, claims it could rejuvenate the web by forcing new business models, surfacing diverse content, and referring higher-quality, better informed, and more qualified visitors.

From a marketing standpoint, as AI search becomes widely adopted we can expect to see the following impacts:

Organic search traffic will drop further: If AI Overviews nibbled away at referral traffic, AI Mode could gobble it entirely. The percentage of zero-click searches could creep higher as more complex queries are resolved on-page. AI can only get better at answering and users will come to trust its answers more, especially for mobile, voice, or quick tasks.
SEO metrics will need to shift: As marketers, we’ll need to capture and show value to avoid knee-jerk reactions to declining raw traffic. Measuring the success of our strategies by clicks or position in classic SERPs will be meaningless. Instead, we will need to focus more on metrics like visibility within AI answers, brand mentions, impressions, conversions, and lead quality etc.
Discovery will rely on AI inclusion: With AI as the new gatekeeper, reading and curating most content for users, marketers will need to consider optimising for AI consumption and humans. Google’s advice suggests it will use the same signals of authority, relevance, and freshness it uses in ranking.
Integration of ads and monetisation of AI results: Shortly after launching AI Mode, Google started testing ads inside AI summaries. Even AI answers will have sponsored content. That could be an opportunity for paid visibility if organic gets tough.
Pushback from publishers: Some publishers and content creators reliant on referral traffic for ad revenues are exploring using nosnippet or even noindex tags to prevent Google from using their content in AI summaries. There’s talk of new industry standards or even legal frameworks to ensure content creators are compensated or at least attributed properly in the AI age. As a B2B marketer, you likely won’t spearhead those battles, but it’s a trend to watch.

So, What Should B2B Marketers Do Now?

You might be asking: “So, do I still bother creating content? How can our brand be found if people aren’t clicking links? Is SEO dead?” All natural questions right now. Here are our answers.

Keep Creating Content – But Make It Count

Churning out endless low-value blog posts just to rank for every long-tail keyword relevant to your industry should never have been a thing. But now, that era is definitely over. Commodity content is being eclipsed by AI. Quality over quantity has never been more important.

Instead, focus on content that is uniquely yours, valuable, insightful, or novel – content the AI can’t easily replace. This could be:

Move down the funnel: In our earlier piece about AI-powered search strategy, we advised gearing content toward more commercial intents since chatbots handle basic info queries. That holds true – think case studies, product comparisons, decision guides that target users who are closer to converting. Those queries are currently less likely to have AI answers, but when that changes, they will be ready to drive a high-intent user to click your link for details.
But don’t abandon informational content entirely: While AI might “steal” some of your thunder, top-of-funnel content is still valuable for brand awareness and for feeding the AI for citations, especially when you do it brilliantly.
In-depth, comprehensive guides: One way to win is to create the content that the AI might draw from heavily. Long-form, well-structured guides that answer a question thoroughly could get cited multiple times. Think of it as writing the source material that the AI needs. If it pulls 5 facts and tips from your guide, that’s 5 citations – and a higher chance the user eventually clicks to see more.
Original research or data: New data gives you an edge. For example, a survey your firm conducted, or a comprehensive analysis you’ve done. Becoming the source that other people cite is a longer game, but aim to create content that others in your industry will reference. If your facts and insights permeate the web, AI systems picking up those facts will trace back to your site. This is akin to classic link-building but in an “information building” sense.
Expert opinions and thought leadership: Build your brand by leaning into your team’s expertise. Opinion pieces, trend forecasts, commentary on news – these can stand out because they carry your story and perspective.
Up-to-date content: If you’re in an industry where things change fast (e.g. tax law updates, new software versions, market trends), being able to publish authoritative content on new developments fast makes you the go-to source.
Multimedia and interactive content: A useful video, infographic, or tool/calculator, could be something AI points to to enrich its results. And if not, those formats give users a solid reason to click through. Broadening your formats also plays to a fragmenting search experience, where people are searching for information on multiple channels - not just traditional search engines.

Google has said their core goal remains to reward “outstanding, original content that adds unique value” – that hasn’t changed.

Google’s algorithms (AI or not) favour content that checks its EEAT (Experience, Expertise, Authoritativeness, Trustworthiness) boxes. And frankly, your human audience will also prefer content that isn’t shallow. So optimising for users first is still the mantra. If you wouldn’t read or share your own blog post, don’t publish it just to hit a quota.

Design Content for AI Visibility

Traditional SEO (title tags, meta descriptions, keyword use) remains necessary. To get picked up in AI Overviews or AI Mode answers, you need to make your content as machine-readable and context-rich as possible, and make it easy for AI to give you credit.

Don’t neglect classic SEO fundamentals: AI search still needs a robust web to pull from. Good on-page SEO (clear titles, headings, fast site speed, mobile-friendly design, etc.) still matters for making your content accessible and high-ranking. It’s also non-negotiable for a satisfying customer experience.
Use Structured Data (Schema Markup): Google’s documentation encourages using structured data, as long as it matches the visible content. It won’t guarantee inclusion, but incorporating structured data on your pages helps Google understand the content and potentially use specific snippets.
Write in “chunks” and clear sections: AI tends to grab chunks of text that answer a specific sub-question. Think of subsections as potential standalone answers to questions. Use clear headings, bullet lists, and concise paragraphs that directly answer common questions.
Provide direct answers: It can help to pose and answer questions within your content. This Q&A style, which many SEO writers already use for featured snippet targeting, can work for AI too. Elaborate and provide depth – you want to provide both the quick answer and the deeper insight if they click.
Clearly identify quotes or key points: For example, in an article, if you quote an expert or have a notable line, consider using blockquote formatting or pulling it out as a callout. An AI might detect that formatting and attribute the quote to you or the person in question.
Ensure accuracy and cite your sources: In-depth pieces that reference reputable data or external expert insights can bolster your credibility. Accuracy matters beyond its impact on your professional reputation. Google is using techniques to improve factuality of AI answers – if your content contains inaccuracies, poor citations, or hyperbole, it may be bypassed.
Keep content up to date: Freshness is a ranking factor and likely an AI inclusion factor, particularly for queries where up-to-date info is expected. Use your publish dates and update dates visibly. Google’s systems often display those; an AI might prefer citing a 2025 article versus a 2019 one for a query about “current trends”.

Importantly, don’t block Google’s AI unless you have a very strategic reason. AI is the present and unavoidable future of search. We believe it is better to be part of the AI web and figure out how to get the most value from it. Most B2B organisations will want to be included in the conversation. The upsides of being cited (visibility, trust) outweigh the downside of losing that immediate click.

Expand Your Concept of Search

Search isn’t confined to Google. It’s fragmented, multimodal, and happens across a range of platforms. That means your traffic can come from anywhere–as long as you know where your audience is and you’re there to capture it.

Search is a behaviour, not a channel: People search on Reddit, YouTube, LinkedIn, TikTok, ChatGPT, and across networks of influence—not just Google. Your brand needs to be discoverable across this wider landscape.
Create for cross-platform discoverability: Wherever your potential buyers are, when they ask for answers will your brand be part of the conversation? Think beyond blog posts. Useful YouTube explainers, clear Slideshare decks, and standout contributions to niche forums or industry conversations all help you become “the answer” across a distributed search ecosystem.
Be intentional about where and how you show up: This isn’t about chasing every trend and every opportunity. Resources are finite. Focus on the channels your buyers actually use when researching, evaluating, and asking peers for recommendations.
AI Overviews pull from diverse formats: Google increasingly surfaces video, podcast snippets, and social content in AI answers. That means YouTube videos, infographics, and well-tagged multimedia can improve visibility—even if they’re not hosted on your site.
Zero-click doesn’t mean zero impact: Think of it as planting seeds. An individual mention in an AI search answer, a contribution to a Reddit thread, or an article shared on LinkedIn might not drive immediate traffic, but cumulatively they shape perception, build trust, and influence future branded searches.

Visibility across platforms builds recognition—but recognition alone doesn’t close deals. So while you won’t win every click, the goal is still to earn one eventually. Treat content and mentions as stepping stones as part of a longer journey: building enough familiarity and trust that, when the time comes, your website is the one they seek out.

Double Down on Brand

As noted, branded searches and strong brand affinity can be a shield against declining traffic. If someone specifically wants your content, Google’s AI is less likely to divert them elsewhere. So work on strategies that increase branded search and direct visits:

Thought Leadership & PR: If your company’s ideas are being discussed in the industry, people will search your brand. Publish bold POV pieces, get your experts on podcasts or webinars and in the press, contribute guest articles. When someone hears your CEO speak at a conference and then searches your company – that’s a branded search that you’ll likely get the click for.
Create content around your experience: Case studies, client stories, webinars with live Q&A. AI is good at summarising the web’s text, but it can’t replicate your insights or the vibe of a live discussion. Emphasise your human expertise and proprietary knowledge. This not only differentiates you but means that if someone really needs that depth, they have to come to you.
Community and Networking: Build communities (LinkedIn groups, Slack communities, etc.) or leverage existing ones. If prospects encounter your brand in a community context, they might bypass search and come directly to your site next time they need information.
Branded Campaigns: Old-fashioned brand advertising can drive users to seek you out by name. The goal is to be top-of-mind so that when they have a problem; they think of you, not just a generic query.
Optimise your site for navigational searches: Make sure that when someone searches your brand or product names, the search results show your best pages and not just aggregator or review sites. This is basic brand SEO. Have properly structured homepage/title, use schema like Organisation markup, and consider running Google Ads on your brand terms. Branded queries might be resilient, but you still want to control the narrative on the results page.

A powerful brand won’t make you immune to the changes in search—but it gives you leverage. If someone’s looking specifically for you, they’re more likely to find you, click, and convert. In a zero-click world, building that kind of intent is one of the few things still in your control.

Focus on Conversion and Experience to Make Every Click Count

Since you may get fewer clicks, you want to maximise the value of each visitor you get. This means:

Analyse your analytics: Look for patterns around queries that are losing traffic. Do they coincide with AI answers? And which pages are still doing well? This can inform your content strategy. If you find your “opinion” blogs aren’t losing traffic but your FAQ pages are – that tells you something about where to put your effort.
An excellent page experience: Google has noted that even with AI in the mix, if a user clicks through, they expect a great experience. That means a fast page load, mobile-friendly, and no annoying pop-ups that drive them away. If your page is slow or hard to navigate, you not only lose that prospect, but Google’s systems notice and that could hurt you overall.
Optimise for conversion and engagement: If AI search visitors are more qualified, then make sure you capture that value. Make the limited copy that appears in citation as enticing as possible. Have clear CTAs, offer next-step content, maybe even custom messaging like “Get the full report” if they likely came from an AI summary of your report. Since they may have got some answers already, think: what can I offer on the page that extends beyond the summary?
Measure beyond clicks: Start tracking metrics that go beyond traffic referrals. For example, maybe your overall organic traffic drops 20%, but you find that form fills or demo requests from organic stayed flat. That might be because the people still coming are the serious buyers. You need to uncover and communicate that nuance. This will help in justifying content efforts to stakeholders who might otherwise panic seeing your traffic dip.
Diversify your traffic sources: We’ve already discussed why this is more than a safety net. However, now you know you’re going to experience a sharp drop in search traffic, it’s time to ramp up other channels. That could be referrals from partnering with industry sites, social, nurturing the audience you already have through email marketing, and paid search or display for key campaigns. In other words, don’t put all your eggs in Google’s basket–or in any other single platform’s either.

Above all, don’t panic – strategise. It’s easy to get caught up in the doom and gloom. But the steps we’ve outlined aren’t rocket science. Unless you’re the Daily Mail relying on clicks to drive ad revenues, there are tangible steps to take to adapt.

In marketing, change is the only constant. SEO has had seismic shifts before. Remember the mobile revolution, or the rise of social, or Google algorithm updates that upended your SEO tactics? Each time, marketers re-focused on the core mission: providing value to users and finding creative ways to get that value in front of the audience. This time is no different.

The Future is Still Bright for Adaptable B2B Marketing

There’s no point denying it. AI powered search is disruptive. It’s altering how people find information and how traffic flows on the web. And the pace of change – from SGE to AI Overviews to AI Mode in just two years – is enough to make your head spin.

But remember: Google’s goal–stated repeatedly–is to serve the user the best, most relevant information. If you align your strategy to genuinely deliver that rather than trying for quick fixes or to game the system, you have a north star.

As we’ve said before and will no doubt say again–much of the advice we’re giving are best practices you should already be following.

The context has shifted but the fundamentals remain the same. Despite the disruption, people still need to buy goods and services. They still need to find and evaluate potential suppliers. They still crave insight, expertise, and human connection.

Potential customers might not click immediately, but if they repeatedly see your name and content cited in answers–as well as elsewhere–it will build recognition and trust. When they need more detail, you’ll already be part of their consideration set.

As for your marketing– it still needs to do what it has always done: pave the road for people to find you. Content will still play a huge role in getting people to know, like, and trust your brand. AI might be able to answer some of their questions, but it can’t do what you do for your customers. It’s your job to understand your audience’s needs and meet them where they are. Even if where they are is now an AI chat box.

Change is challenging, but it drives innovation. So, keep calm and carry on. Plan, evolve, and see this moment as a catalyst. The machine web might be rising, but smart human marketers will rise right along with it and find ways to get the machines working for us, not against us.

Work With Us

How to Adapt B2B Marketing Strategies for Google's AI-Powered Search Era