AI Search
With Google's AI Overviews and 'AI Mode' set to transform the search landscape, publishers and agencies are grappling with declining click-through rates and the need to adapt their SEO strategies.
by | 05/21/2025
Google has just introduced ‘AI Mode’ to its search engine, initially available to all users in the US, with a global rollout planned for later. Powered by its Gemini 2.5 model, this feature enables more conversational interactions and adds new capabilities such as automatic concert ticket purchases and live video search.
In addition, the tech giant recently announced that its AI Overviews in search now reach 1.5 billion users monthly. Unlike traditional search results, AI Overviews summarises information from multiple web pages with the idea of helping users quickly grasp complex topics. Launched last May in the US and in other countries last August, AI Overviews have helped the tech giant's search revenue grow 10% year-over-year to US$50.7 billion.
However, while AI Overviews may have benefited Google, analysis from Ahrefs indicates they significantly impact organic search traffic, with a 34.5% reduction in click-through rates (CTR) for top-ranking pages. This decline challenges Google’s claims that AI Overviews improve user engagement. Furthermore, AI-generated overviews are placed above traditional web links, which are vital for online publishers dependent on traffic referrals from Google’s search engine, raising concerns about Google taking too much control.
"With AI Overviews, Google is making it super easy for users to get quick info, which is awesome," says Gary Cheung, managing director, NP Digital. "But on the flip side, it seems like they’re sidelining publishers by summarising content right in search results. It’s great for users who want fast answers, but it raises concerns about how much visibility the original creators get for their hard work and the lower click-through-rates."
At least for now, Google is keeping publishers in the dark, offering little transparency around how publishers are supposed to adapt their SEO strategies, especially as Overviews prioritise summaries over direct links.
"Historically, SEO algorithms have never been 100% transparent," says Bob Du, managing director at Jellyfish Singapore. "With AI Overviews, publishers are not yet able to understand referral traffic contribution from embedded in-line links versus organic search results, and this does escalate the level of complexity, admittedly leaving a little more to be desired from a transparency standpoint."
Given the significant long-term impact that AI Overviews will likely have on the search landscape, some media agencies are already rethinking their SEO strategies.
"Instead of just chasing keywords, we’re focusing on crafting richer, more engaging content that really grabs attention," says Cheung. "We’re looking to incorporate more visuals, like videos and infographics, to draw people in. We’re fine-tuning our content to answer specific questions, so even with those summaries popping up, our pieces still stand out."
In this new era, audiences are searching on different platforms, therefore there’s a pressing need to optimise not only for traditional search engines but also for AI-driven platforms. Du at Jellyfish says that publishers now have more work to do, which is to learn how they are earning their share of model, specifically from Gemini (Google's AI chatbot) to gain visibility in AI Overviews and Gemini’s citations.
"We have built a tool called Share of Model to help us understand what Gemini thinks about a brand, taking into account the category, products and key audience personas," says Du. "This helps our clients identify the top cited domains captured by Gemini. It informs briefs across paid, owned and earned media. Specifically, you can fine tune content towards specific attributes that you would like to shape Gemini’s perceptions on to gain visibility. From there, you can continue to track your share of voice and measure performance."
However, Stephanie Wong, SEO director at Assembly APAC, says that SEO fundamentals don't change overnight.
"What's worked historically is still going to be critical, and there's no need to do a big strategy pivot just because there’s a new product," says Wong. "Studies have shown that over 50% of sources cited in AI Overviews are from first-page results. That said, AI Overviews reward content structured for quick consumption. We're refining content formats to align with this, incorporating more Q&A structures and conversational tones, and ensuring schema markups are airtight. The goal isn't to fight AI Overviews—it's to work with them and stay ahead of the curve."
Despite these optimisation efforts, Google has always dominated search, and AI Overviews is just another evolution of that dominance. SEO has never been a level playing field, and results are never guaranteed. While publishers, agencies, and SEO experts can try to optimise their content, Google ultimately sets the rules, which can be vague.
Recent studies suggest that AI Overviews are routinely taking a chunk of visibility away from organic results, leading to a notable decline in referral traffic click-through rates. Amsive completed a large study based on 700,000 keywords across 10 websites in five key industries—finance, education, SaaS, healthcare, and pets—focusing on 10,000 keywords that triggered AI Overviews while already ranking organically. According to them, keywords that now trigger AI Overviews saw a CTR decline of 15.5% on average.
"We’ve definitely seen impact on referral traffic, in some instances we have seen reductions of 30%," says Sebastien Kriegel, digital partner, transformation & activations, UM Australia. "We are closely monitoring this impact via CTR trend analysis looking at the linear regression of data over the last 12 months."
While studies suggest a significant impact on referral traffic, particularly click-through rates, experts say that measuring the effectiveness of Google AI Overviews on referral traffic should be done differently.
"Since AI Overviews lean more towards research-based queries rather than conversion-driven keywords, success should be measured differently," says Wong. "Time on page, engaged sessions, and interaction rates should be what we look at to determine the value of this traffic."
Du concurs, stating that "instead of focusing on just referral traffic volumes, engaged sessions and conversion rates might provide a better consideration of impact post generating strategies to address your Share of Model on Gemini and AI Overviews mentions."
With nearly 19 million keywords now featuring AI Overviews, leveraging keyword data to inform content strategy is crucial. As a result, many are finding new approaches and best practices to better optimise content.
19 million keywords now feature AI Overviews.
"AI Overviews thrives on top-of-funnel, research-based queries—'best coffee brands' is more likely to trigger AI Overviews than 'buy coffee now'. Understanding this distinction is crucial," says Wong. "Instead of chasing every AI Overview opportunity, we focus on the right ones, where brand exposure leads to real audience engagement. That means optimising for questions, enhancing authority in research-based topics, and ensuring our content is the go-to source that Google wants to feature."
Cheung at NP Digital says that his agency is diving into keyword data to inform its content strategy.
"Our goal is to create content that not only ranks well but is also genuinely helpful—so when AI gives a summary, our work doesn’t get lost in the shuffle," says Cheung. "More brands are showing up in featured results, and that’s where we need to aim to be. Conversational search is leading to new, longer, and more specific queries, giving us opportunities to target niche audiences with tailored content."
Meanwhile, Kriegel at UM says they are leveraging tools like SEMRush and Ahrefs to identify AI Overview keywords and compare against Google Search Console to identify zero click terms. "We are monitoring this changing space closely to understand keyword strategy implications and how this is being affected by content," says Kriegel. "Once we have more confidence in the data from a longer period of user search behaviours, we will look to implement a zero-click keyword SEO strategy."
Research from Jellyfish and YouGov finds that 66% of 18- to 24-year-olds in the US ask AI services such as Google Gemini, ChatGPT and Meta AI to make a recommendation about a brand, product or service. It's clear that the way people are searching is changing.
"This is beyond AI Overviews with people switching to other platforms like AI chatbots (Perplexity, Copilot) and social platforms (TikTok, Pinterest, Reddit) to get answers to their questions," says Kriegel. "This evolving space is a catalyst for us to rethink search holistically and focus on search everywhere."
While Google's AI Overviews are likely to significantly impact the search landscape, Wong stresses that the only certainty is constant change.
"The product is new and evolving fast. It's hard to say if what works now will work even two to three months down the line," says Wong. "However, one thing is certain: diversification matters. Brands relying on organic search for the lion's share of their traffic need to rethink their approach. AI-driven search is growing, and marketers need to invest in multi-channel strategies to maintain visibility—because in digital, agility wins."
Most read
Dentsu drops sale of international arm but is open to deals in local markets
Dentsu to appoint Takeshi Sano as its new global CEO
WPP launches new Elevate28 strategy with four core divisions and $676 million savings
Omnicom to axe and relocate more jobs as it doubles cost savings target to $1.5 billion
How a mistreated monkey led to a marketing opportunity for IKEA
Dentsu reports record loss ahead of CEO transition, suspends dividends
Lunar New Year campaigns worth watching in 2026
Publicis hits record profit margin, adds 5800 staff and hikes cash bonuses 8%
WPP plans to launch WPP Creative
Publicis creates Japan and South Korea sub-region
Related Articles
