AI Overviews: How it Works & How to Optimize for It

How AI Overviews Works

AI Overviews is a new, and ever-changing feature that was released by Google en masse to U.S. searchers on May 14. A lot has changed even since its release with more changes likely to come, but here are some common questions and answers about the latest search feature.

When Do AI Overviews Appear?

• AI Overviews does not display for all Google searches. In fact, AI Overviews is triggered on fewer than 10% of all search queries.
• It is designed to answer more complex, multi-pronged questions, but it also seems to trigger often on simpler, “What Is” questions and comparison queries, as well.
• AI Overviews often appear alongside other informational SERP features, including “people also ask”, featured snippets, and discussion and forum modules.

Types of AI Overviews

There are three types of AI Overviews that appear in search results:

1. Infomational: Informational AI Overviews display in the form of paragraphs, breakouts, bullet points, citations, images, and videos.
2. Shopping: Shopping AI Overviews provide actual products — including in individual, carousel, and grid formats — and product recommendations from review sites.
3. Local: Local AI Overviews include citations, local business listings, and map results.

How Does AI Overviews Generate Its Result?

At its core, AI Overviews leverages Google’s baseline search algorithm as its source of information.

“While AI Overviews are powered by a customized language model, the model is integrated with our core web ranking systems and designed to carry out traditional ‘search’ tasks, like identifying relevant, high-quality results from our index,” explained Liz Reid, VP and Head of Google Search.

Although the output may look different than other organic search results, the way AI Overviews sources its information is the same way Google sources of information for all other organic search results.

One difference between AI Overviews and organic search results, however, is how heavily AI Overviews weighs relevancy over domain authority when it comes to producing its results. While organic search results are heavily weighted in page authority, which is measured by metrics like backlinks, site traffic, and other “popularity contest”-type factors, AI Overviews seems to really only care about relevancy of the result to the searcher’s query.

“It seems to only consider relevance,” explained Garrett Sussman on a recent Moz Whiteboard Friday session. “The things that we’re used to when it comes to AI or to any sort of SERP result is authority and links and backlinks… it’s just relevance in reality.”

How to Optimize for AI Overviews

Tips for Optimizing Your Content for AI Overviews

Really, we wouldn’t suggest changing your overall SEO strategy, but there are a few tweaks we recommend in the way you write and publish new or under-performing content if you want to rank in AI Overviews results:

• Ensure your content is relevant to searches that appear most in AI Overviews. TIP: Use a tool like this AI Overviews Visualizer to measure your content’s relevancy to your target searches.
• Use authoritative tone of voice, quotes, and statistics, which are more likely to appear in large language model outputs.
• Break up your content into approximately 700-character segments. Smaller segments of copy are more likely to get picked up by large language models (LLMs).

Tips for Optimizing for Shopping AI Overviews

Shopping AI Overviews pulls information largely from Google’s Shopping Graph, which is a real-time data set of more than 35 billion product listings. The Shopping Graph pulls detailed information about each product, including product availability, ratings and reviews, sizes, colors, materials, and more, helping online shoppers to find products down to every detailed search criteria.

• Include complete, detailed information for every product and share those on your Google Merchant feed.
• Solicit quality product ratings and reviews
• Generate “buzz” about your product on third-party sites (think: subreddits, and influencer or affiliate sites).

Tips for Optimizing for Local AI Overviews

Local AI Overviews pulls from much of the same sources of information as the Google Map does, with a few exceptions. Those sources include directories, but also other sources, like local subreddits and local “best of” lists.

• Ensure you are listed on major and niche industry directory listings and map results.
• Consider a local PR push to land local “best of” awards and generate buzz about your location locally.

Tools to Better Optimize for AI Overviews

There are several tools to help measure and improve your results using AI Overviews. Here are a few:

• Ziptie syncs directly with your Google Search Console account to generate actionable data to improve your website rank and monitor your AI Overviews performance.
• MarketBrew’s AI Overviews Visualizer analyzes meta titles to help you better understand how Google perceives them. It can also help you identify gaps in your content’s relevance to a particular topic and make necessary adjustments to improve its chances of being featured in AI Overviews.
• The popular SEO tool, SEMRush has built in tools to help you monitor your website’s visibility within AI Overviews. 

Should You Shift Your SEO Strategy for AI Overviews?

Shifting your SEO strategy to prioritize AI Overviews can offer enhanced visibility and engagement with AI-powered search engines. However, it’s crucial to consider potential risks before making major changes to your SEO strategy — especially if things are working well.

The Risks

Algorithm Changes

Search engine algorithms are constantly evolving. Over-reliance on AI Overviews might leave your content vulnerable to algorithm updates, potentially leading to a drop in rankings if the optimization strategy is not adaptable.

“They’ve been rolling it back. There has been no reliability in terms of how Google is playing this out,” said Sussman. “Initially, they rolled it out all over the place. They got all sorts of controversial AI-generated results that got them a lot of bad press, and they rolled it way back.”

Limited Reach

While AI Overviews are gaining prominence, they may not be the primary mode of search for all users. At the time of writing, Google’s AI Overviews is only appearing in fewer than 10% of all search queries. Focusing solely on this format might limit your reach to a narrower audience segment, so spending too much time catering to this new and evolving feature, may be more risky than beneficial.

Changing Features

Just in the limited public lifespan of AI Overviews, they have deployed features, like a follow-up questions feature, that they have since removed. Spinning your wheels trying to optimize for each new feature, only for it to be taken away within the month, may be valuable strategy time wasted until the tool’s features become more settled.

Potential Keyword Ranking Trade-offs

There is the risk that if you’re modifying existing content that is ranking organically for AI Overviews that you trade off your organic rank for AI Overviews, which may generate more clicks.

If your content is ranking organically in SERPs for high priority keywords, we do not advise you modify that ranking content to also try to capture AI Overviews, as you may tank your rankings in organic results without actually gaining any benefit in the AI Overviews snippet. If you have non-ranking content, by all means, take that risk as there’s nothing to lose.

The Importance of a Holistic SEO Strategy

Whether you decide to incorporate an AI Overviews targeted approach into your SEO strategy or not, we always recommend the adoption of a holistic SEO approach that combines AI optimization with strategies that cater to tried and true SEO methods that are proven to be successful for your brand.

We also advise our clients to take measured risks. For example, let’s not fix what isn’t broken. If something is working (i.e. ranking in the top spots in Google’s SERP), let’s leave that part of the strategy alone. However, if a particular keyword is not gaining traction using traditional methods, why not give optimization for AI Overviews, specifically, a go? 

Regularly monitoring algorithm changes and adapting your strategy accordingly is also crucial for long-term success.

How Can You Measure the Impact AI Overviews is Having on Your Site?

We also haven’t found a single tool that can flag which keywords appear in AI Overviews (let us know if you know of any and we’ll test it out and link to it here!). However, there are still ways to track where AI Overviews appears on your keywords and measure its impact.

Ways to Track Which Keywords Appear in AI Overviews

Use Google Search Console

Unfortunately, Google Search Console does not distinguish which search impressions and clicks came from AI Overviews versus organic results in Google Search Console. That’d be just too easy. (sigh..)

However, with some elbow grease, you can pinpoint which keywords are triggering AI Overviews in Google Search Conosle. This takes some trial and error, but keep in mind the types of queries — typically long-tailed “What Is” searches — that often trigger AI Overviews results as a starting point in your deep dive.

Use Third-Party SEO Tools

You can also use SEO tools, like Moz or Semrush, to track fluctuations in your organic search visibility, both impressions and click-through rates before and after the keyword’s appearance in AI Overviews.

Some SEO tools, including Semrush and SE Ranking, have begun incorporating AI Overview tracking features in their keyword tracking tools. While it doesn’t flag all instances of search queries that led to AI Overviews appearances, it does show when your website shows in AI Overviews on keywords you manually track in the Position Tracking tool. You can also identify new opportunities within AI Overviews and view the prevalence of AI Overviews across different industries in its “Semrush Sensor” tool as well.

Using Google Analytics

In GA4, you can create a custom report that segments out URLs containing the “#:~:text=” parameter in the URL. This specific string of characters is a parameter that appears in the URL of links within AI Overviews and indicates that users clicked on a link within an AI Overview.