Background to this AIO Research & Methodology
2024 marked a transformative year for AI Overviews in search, but we anticipate 2025 will reveal their true impact on the search landscape.
So, in 2025, we’ll be stepping up our programme of generative AI research studies and launching new generative AI tools to help you optimise across all AI Search Engines, be they traditional search engines, chatbots, LLMs or any combination thereof.
In early 2024, we started exploring Google’s Search Generative beta launch by researching the potential impact on ecommerce and of generative AI on branded and product search.
Later in the year, we started a series of collaborative research studies on the full rollout of AI Overviews in partnership with our esteemed friend and SEO veteran Rich Sanger. These studies covered many aspects of AI Overviews and there are multiple studies on his site, he’s also taken time to breakdown the Google SGE patent which is well worth a read. There’s too many resources to list them all here, fortunately he has wrapped them all nicely into an AI Overview library.
If like me, you prefer facts and data over conjecture, then you’ll be pleased to hear that Rich and I will be continuing our collaboration in 2025 and will shortly be publishing our next collaborative research study on the volatility of AI Overviews.
But what better way to start a new year, than to introduce a new collaborator, Dave Cousin, (AKA “Dave the SEO”), an experienced SEO speaker and consultant, who has been working with major brands and agencies for the past 20 years.
He’s researched a unique categorised list of 10,000 search terms and has dived into the depths of our comprehensive AIO rank tracking and SERPs reporting data to pull out some insights into a whitepaper that gives you a real feel for how things stand at the start of what will be another fast-moving year for everyone in SEO.
If you like understanding all SEO topics inside out, then you’ll appreciate this whitepaper and Dave’s commentary which goes into great depth about the composition and structure of the AIOs. It’s about a 4/5 on my personal SEO geek scale, providing in-depth analysis and actionable insights – but I appreciate not everyone likes lots of charts and stats, so alternatively please enjoy this brief summary of our findings below.
SERP data was collected for 10,000 keywords across 7 industries in the US on desktop devices in December 2024.
As always, we have shared the keywords we used on Google drive. For any SEO developers or data scientists, I have also shared the detailed AI Overview export files and the AIO PowerPoint slides we usually share with all customers (feel free to get in touch and ask me any questions about the data).
How Often Do AIOs Appear?
Our research shows that:
· AIOs still don't show up for the majority of searches.
· AIOs appear for a relatively low number of searches (29.9% of Keywords and 11.5% of Total Search Volume).
· AIOs don't appear more frequently for terms with higher search volumes.
· AIOs do show for less popular terms on a regular basis, though most frequently for medium search volume terms (501-2,400 searches/month).
This may seem low, but it's important to remember that the percentage of searches that trigger an AIO is even lower when you consider that AIOs don't appear more frequently for terms with higher search volumes.
The research did find that AIOs are most likely to appear for keywords with mid-range search volume. 42% of keywords with a monthly search volume between 501 and 2,400 have an AIO, and 36% for those with a search volume between 101 and 500. This suggests that there are still opportunities to rank well in traditional search results and you shouldn't only focus on high-volume keywords when optimising for AIOs.
How Often Do AIOs Show by Industry and User Intent?
AIOs don't appear equally across all topics or sectors and the opportunities in your sector will depend on what you are selling. Of the seven sectors the research looked at, Telecoms had the highest percentage of keywords with AIOs (56%), while Beauty/Cosmetics had the lowest (14%).
This might be because the services or products in the Telecoms and Business Services sectors (which also had a high percentage of keywords with AIOs at 41%) tend to be more complex, so users need more information to make informed decisions. Even within sectors such as Beauty or Clothes, there will be some areas where people have more complex problems, or there is more risk involved in the product so more research is likely, e.g. Premium trainers are likely to be researched in more detail than a t-shirt.
AIOs do show more often for some industries/sectors than others. The appearance of AIOs also varies significantly by user intent. This makes sense if you consider that Google is likely basing the appearance of AIOs on an understanding of when users find them useful. The research found that:
- AIOs do appear more often for Problem Solving and Specific Question intents. These are also the intents where AIOs are most able to give users a useful answer and where they show the most (74% and 69% respectively). This may be because these types of searches are more likely to have a definitive answer that can be easily summarised by an AIO.
- AIOs don't show for any navigational queries. This isn't surprising, as AIOs aim to provide users with an overview of a topic, not help them navigate to a specific website.
- Topic Research searches are much less likely to have AIOs. Only 3.37% of keywords in this intent category show an AIO. This is probably because these terms tend to be very broad.
Problem Solving and Specific Question intent questions are key areas to focus on. While questions fully answered by AIOs won't always lead to traffic, there are opportunities to attract clicks with compelling titles for links by finding terms related to bigger research journeys, high-value or high-risk decisions, or search terms where people want to research more for pleasure. For Range Info / Product Service intent terms, AIOs only show 10.8% of the time, but these and other mid-funnel, commercial, non-brand terms can be great for brand awareness as it is possible to optimise for your brand to appear and be recommended within an AIO.
Brand vs. Non-Brand
Non-brand keywords are more likely to trigger AIOs than brand terms. 33.3% of non-brand keywords and 19.6% of brand terms show an AIO.
This is unsurprising as brand terms tend to be lower in the funnel, indicating that someone is close to making a purchase and an AIO may slow down the buying process. However, AIOs can still be relevant for brand terms, particularly for Informational intent keywords, with 69.8% of Informational brand keywords showing an AIO. This presents a great opportunity to influence what Google writes about your brand in the SERPs, particularly for post-purchase questions and problem-solving searches, which can lead to increased customer satisfaction and improved repurchase/renew rates.
Impact on Traditional Search Results
With the "Show more" button clicked, AIOs can push traditional organic results further down the page, potentially even below the first fold.
Before the AIO is expanded, two organic results are typically visible on mobile and three on desktop. However, once the AIO is expanded, the median pixel drop is 220 pixels (the equivalent of about two organic results) on desktop. This means that only the first organic result might still be visible without scrolling. On mobile, even the 25th percentile pixel drop value is enough to push both initially visible organic results off the screen.
This covers some of the main highlights. If you would like to get into more depth then please do download the Full Whitepaper to discover more details about:
- The different types of AIOs and how they're laid out
- How many links AIOs typically include and which websites are most likely to be featured
- The impact of AIOs on Paid Search ads
- How to optimise your website content for AIOs
I'm currently working on research into AI Overview Volatility and whether this is driven by changes in the SERPs and organic rank changes or whether AI Overviews fluctuate independently of other SERP changes.
Stay tuned.
To the SEO community:
I'm actively seeking research partners for future studies in this rapidly evolving space. If you're interested in contributing to cutting-edge AIO research, please reach out to me to discuss potential collaboration opportunities.
