Is it worth optimising for AI Overviews?

Our previous AI Overviews analysis was carried out when this feature was still called Search Generative Engine (SGE), in January 2024. At that moment, across a smaller dataset, we found that overall 55% of the results overlapped with the organic listing results, which meant that over half of the AI Overviews citations were drawing from content already ranking on page one. This gave us an early indication that traditional optimisation efforts were still playing a role in AI-powered visibility.

However, since then, Google has rebranded and expanded AIO significantly, both in terms of geographical reach and algorithm complexity. Overall, the AI industry is evolving at an unprecedented pace, where advancements in AI technologies are impacting industries globally even beyond SEO:

• In March 2024, Google began testing AI Overviews directly in the main search results, even for users not enrolled in Search Labs.
• In May 2024, SGE was officially rebranded as AI Overviews and launched in the U.S., with plans to expand globally.
• Google rolled out AI Overviews for signed-out users in the US. This marked a significant milestone, as AIOs are no longer restricted to logged-in users, opening the feature to a broader audience.
• By August 2024, AI Overviews were rolled out to several countries, including the UK. 
• In October 2024, the feature expanded to over 100 countries, marking a significant global presence. With Google expanding its linguistic and geographical coverage, AIOs were quickly becoming a standard part of search experiences worldwide.
• In March 2025, Google started testing AI Mode, an experimental feature offering a fully AI-generated search experience, available to Google One AI Premium subscribers. Find out more about AI Mode.
• Google has now integrated Gemini 2.0 into AI Overviews, enhancing the system’s ability to handle complex queries involving math, coding, and multi-step reasoning.
• According to BrightEdge, following the March 2025 core update, there was a noted decrease in the overlap between AI Overview citations and the top 10 organic search results – 15% overlap.
• On the other hand, Google Search Console (GSC) still lacks AIO-specific reporting tools. While AIO clicks are being recorded, there’s no dedicated visibility into their performance. This lack of transparency adds complexity to tracking and optimisation efforts.

With the introduction of Gemini 2.0, changes in citation behaviour, and the reduced overlap between AI and organic results reported after the March 2025 core update, it seems clear that the landscape has quickly shifted. As such, a fresh round of research was needed to assess the current state of AI Overviews, further emphasising the need for agile strategies in this fast-changing space.

At Varn, we have gathered a keyword set of around 7K keywords split into different industries: Ecommerce, Health & Pharma, Travel & Tourism, Insurance, Finance, SEO & UX and B2B. These have been run through google.com in April 2025, with a VPN set in California, USA, to ensure a fair comparison with our previous test. 

The new research includes a few different angles. We will be answering the following questions:

• How frequently do AI Overviews appear across the search results?
• How many sources do AIOs link to?
• What percentage of AIOs content comes from organic search results?
• How clickable are AIOs?

Our research explored several key dimensions. These were our findings.

• Interestingly, Health & Pharma leads the pack with 47.49% of searches showing AIOs. This suggests Google is highly confident in surfacing related AI-generated summaries, likely due to the abundance of authoritative, structured content. This reinforces the need to publish well-structured, medically reviewed content aligned with EEAT guidelines.

• SEO & UX and B2B follow closely, at 44.31% and 39.96% respectively. These sectors often involve informational queries where AI can summarise answers effectively. Brands in these spaces should focus on publishing expert-led content that directly addresses user intent in a concise manner.

• Finance & Investment and Travel & Tourism sit mid-table, which may reflect Google’s caution around surfacing summarised information in Your Money or Your Life (YMYL) areas with highly dynamic content. Still, with over 1 in 4 finance searches triggering AIOs, it’s vital for financial brands to ensure content clarity, compliance, and trust signals are embedded across key pages.

• Insurance and especially Retail & Ecommerce show minimal AIO activity. This is expected in conversion-driven sectors where transactional queries are less suited to summarised answers and more reliant on product feeds, PPC, and traditional organic listings. However, ecommerce brands should still monitor for informational AI Overviews on top-of-funnel keywords.

Different industries exhibit varying patterns when it comes to AI Overview triggers and understanding these can provide businesses with a competitive edge. By tailoring content to address these unique triggers, companies can optimise their chances of appearing in AI Overviews, thereby enhancing their overall search visibility.

With this angle, we wanted to provide a valuable insight into how Google’s AI Overviews compile and attribute information, particularly how many distinct URLs are referenced in each AI-generated summary. The findings shed light on the complexity, diversity, and trust signals that may be influencing AIO composition.

AIOs citing only 1 or 2 sources are extremely rare, implying that Google intentionally avoids overly narrow sourcing, likely to mitigate incomplete information.

On the opposite end, AIOs citing 10+ sources may correspond to more complex or ambiguous queries, where a broader range of inputs is needed to form a satisfactory summary, though they still represent a relatively small share.

This data reinforces the idea that AI Overviews are not simply pulling from top search results, but instead curating from a contextually relevant pool of sources. With limited citation slots available (typically between 5 and 8), not all “Page 1” results will be included. Sites that demonstrate clear expertise and trustworthiness are more likely to be chosen. And given that Google pulls from diverse URLs or domains, being cited in third-party content such as industry blogs, forums or thought leadership pieces, may increase your chances of an indirect inclusion in AIOs.

The most critical aspect of the new research is again the overlap between URLs featured in AI Overviews and those appearing in traditional organic search results. Below is a graph that demonstrates how often a URL cited in AIOs has appeared within traditional Google rankings, for the same keyword(s). Overall, we see that the URLs cited in AI Overviews are not ranking in traditional results 78.6% of the time.

The graph below demonstrates similar data, but also includes search results included within People Also Ask questions, and Featured Snippets – both features included within organic Google SERPs. Interestingly, we see the percentage of AIO cited URLs that don’t rank in the first 2 pages of organic SERPs decrease when we start to include these additional features – showing that Featured Snippets and PAA results are more closely linked to URLs being cited within AIOs than results shown within the standard search results are. The percentage of AIO URLs not showing within search drops from approximately 78.6% to 70.4%. 

However, the substantial share of cited URLs that do not rank at all across all industries indicates that AIOs are pulling in sources beyond the top organic results. This fragmentation creates uncertainty for SEO visibility and makes AIOs a critical new layer of SERP analysis.

In the Insurance, Travel and Retail industries, the picture is even more extreme. Only a few URLs cited are on Page 1, while a sizable 70%+ are not ranking at all. This suggests that in these commercial spaces, Google is pulling from alternative sources such as community content, or niche publishers that provide unique value or context not captured in traditional SERPs.

This points to a broader trend where ranking well no longer guarantees visibility in AI-generated summaries.

Obviously, as AI Overviews are rolled out, the general concerns are whether these will reduce website traffic from SERPs, or more importantly, whether people are actually interacting with the sources cited in AIOs. This brings us to the question: is it worth optimising your content or strategy to target AI Overviews?

As mentioned before, Google is currently tracking clicks and impressions in AI Overviews, however, there is no way to isolate these within Google Search Console reports. This means that producing insights into visibility and click through rates has become a real challenge. While data is still emerging, early studies and industry experiments suggest:

• AI Overviews reduce clicks by 34.5%. A study by Ahrefs of 300K keywords has seen a massive drop in clicks on their position #1 year on year, when AIOs are now present.

From our dataset, we have also reviewed the average CTR for the keywords where AI Overviews are present in March 2025 vs March 2024.

The table above shows a positive increase in the click through rates associated with URLs featured in position #1 of Google. However, this dataset is limited to keywords associated with our existing clients – clients who each have an optimised, dynamic search strategy in place. Therefore, this data is somewhat biased, as we purposefully work to drive more position #1 rankings, and to improve organic CTRs as part of our ongoing work. It is also worth noting that AIOs have changed a lot in the past 12 months, which will also likely be impacting this data. Our data is also limited in size and so much smaller than other studies available online such as the Ahrefs study mentioned above.

One conclusion we can draw from this table is that the lower your organic Google ranking, the more your organic CTR is likely to drop since the introduction of AIOs. Whilst CTR typically drops in line with your ranking position anyway as a general rule of thumb, this drop appears to have increased over the past 12 months, potentially due to the impact of AIO search results. 

In short, here at Varn we believe AI Overviews are reshaping the click landscape, not eliminating it. The sites that reshape their content strategy to be both informative and AI-digestible will be best positioned to retain, and even grow, visibility in the evolving SERP environment. Contact us if you would like us to support you with your AI visibility.

1. AI Overviews are becoming mainstream.

Since their rebrand and rollout in 2024, AI Overviews (AIOs) have rapidly expanded globally, now appearing in 36.3% of searches, with particularly high prevalence in Health & Pharma (47.49%) and SEO & UX (44.31%) sectors.

2. Traditional SEO still plays a role, but it’s not everything.

URLs cited in AI Overviews are not ranking in traditional results 78.6% of the time. This confirms that while SEO best practices remain valuable, AIOs also pull from a wider, more diverse set of sources, often bypassing traditional Page 1 rankings entirely.

3. Being cited doesn’t require ranking, but it helps.

AIOs frequently source from Featured Snippets, People Also Ask, and high-EEAT domains, even if they aren’t ranking on page 1. Getting cited indirectly via trusted third-party domains is a still viable off-page strategy.

4. Google’s preference is for breadth, not depth, in citations.

Most AI Overviews reference between 5 and 8 unique sources, suggesting a need for content that is both authoritative and concise. Overly narrow or overly complex content appears less likely to be featured.

5. Industry visibility varies widely.

• Health & Pharma benefits from high AIO inclusion, likely due to structured, credible content.
• Retail & Ecommerce has the lowest appearance, pointing to limited AIO presence for transactional queries.
• Insurance and Finance sectors also show low overlap with organic rankings, indicating Google may prefer third-party or niche content here, enhancing user-generated content and reassuring the importance of EEAT signals.

6. Tracking remains a challenge.

Despite Google logging AIO clicks and impressions, Search Console lacks dedicated AIO reporting, making it difficult for brands to isolate and optimise for this feature. This limits clear attribution and necessitates more agile analytics solutions.

7. AIOs are influencing CTRs, but not uniformly.

While broader studies suggest AIOs reduce clicks, our own data shows CTR improvements at position #1 year-on-year for sites already implementing strong SEO and content strategies.

8. AI Mode is on the horizon.

With the launch of Gemini 2.0 and Google AI Mode, the shift towards conversational and summarised results is accelerating. Early adaptation will be key to long-term visibility.

So, to answer the main question of this article – is it worth optimising for AI Overviews? We’d say yes, but with a strategic mindset. Our research suggests that AIO visibility does not come from high-ranking pages alone but from contextually relevant, authoritative, and diverse sources. Therefore, there are many ways to approach an AI-driven strategy, such as refining on-page content to better match informational intent, or strengthening off-page signals through authoritative backlinks and citations. As AI Overviews continue to reshape the SERP, brands that align their content with these principles will be best positioned to maintain and even grow their visibility. If you are struggling with the new AI advancements, get in touch with our innovation team – we’d be happy to support your strategy for maximum AI visibility.