What Do AI Overviews Mean for SEO in 2025?

Since their launch in May 2024, AI Overviews have continued to have a dramatic impact on search traffic.

In this article, I cover new 2025 research on the impact of AI Overviews, plus how to adapt your content to improve your chances of appearing in an AIO.

What Are AI Overviews?

AI Overviews are a Google SERP feature that uses generative AI to answer user queries.

These overviews pull information from multiple sources to synthesize an answer that includes links to multiple web pages.

AI Overviews were originally called Search Generative Experience (SGE).

Why Are AI Overviews Important?

Google’s AI Overviews are part of the rapid transformation in how information is stored, retrieved, and served online to users.

AI Overview placement: Because AI Overviews appear at or near the top of Google search results, “blue links” are typically pushed further down the page, resulting in reduced clicks to ranked websites.

Declining traffic: For many sites, the increase in AI Overviews and zero clicks means reduced traffic and revenue.

Why is traffic declining?

In our research into searchers’ trust in AI, we found that only 18.6% of people who have seen AI Overviews “always or usually” click links to read the source material.

How Do I Measure the Impact of AI Overviews?

New Semrush/Datos research on 10+ million keywords that trigger AI overviews reveals that AI Overviews appeared on 13.14% of queries as of March 2025. This represents a doubling from 6.49% in January 2025.

Industries experiencing the fastest growth of AI Overviews were:

• Science (+22.27%),
• Health (+20.33%)
• People & Society (+18.83%)
• Law & Government (+15.18%)
• Travel (+14.3%)

Search intent: 88% of AI Overviews occur on informational keywords. Low-competition, low-difficulty, low-CPC keywords predominate.

However, AIOs on navigational, commercial, and transactional queries, while in the single-digit percentages, are growing too.

Keyword Intent	AIO % Jan 2025	AIO % Mar 2025	% increase
Navigational	0.74%	1.43%	+ 93%
Commercial	6.28%	8.69%	+38%
Transactional	1.69%	1.76%	+4%

User Behavior

Our research shows that those most likely to click through to AI Overview sources earn from $150k to $200k.

We found that over 42% of users say they’ve seen “inaccurate or misleading content” in AI Overviews. Over 35% say the overviews are “missing important context.” And 31.5% report “biased or one-sided answers.”

• 30-45 seconds: The average time users spend on AI Overviews, per Indig’s study.
• Reduced visibility: 70% of searchers “never read past the first third of an AIO”
• 88% click the “Read more” option.

Recency Bias

Seer Interactive found that 85% of AI Overviews cite content published in the last 2 years (2023-2025).

Financial guidance, especially content related to payroll and taxes, shows the most extreme recency bias, followed by the travel and energy industries.

However, ranking data suggests that high-quality evergreen content, including how-to guides and other instructional materials, has a long shelf life.

How Do I Measure the Impact of AI Overviews?

You can track your presence in AI Overviews by using Semrush. This is my tool of choice for AI tracking, brand monitoring, and SEO.

AI Overview data is available in multiple areas of Semrush, including these tools:

• Organic Research
• Position Tracker
• Domain Overview
• Keyword Overview
• Semrush Sensor

In the Semrush display pictured below, we see that in September 2025, Exploding Topics pages were cited in almost 4,000 Google AI Overviews.

This SERP Features Trend chart appears on the Organic Research page in Semrush. Scroll down below this chart to see how many of your AI Overview citations link back to your site.

Click on either of these AI Overview icons to see a list of the specific keywords along with related metrics, such as:

• Traffic
• Volume
• Keyword Difficulty

Note the AIO icon used in the Position column. Hover over the icon to see whether the related AIO links back to your domain.

How to Optimize for AI Overviews

Accuracy, clear language, and logical structure help Gemini surface your content in AI Overviews.

To facilitate chunk-level retrieval (also called snippet extraction), structure content so that Gemini can easily identify fact-rich, concise passages to answer specific queries. The abstracted passage must stand on its own, requiring no other information.

Organize content logically, use descriptive headings, and nest heading tags properly.

For example, instead of this....

<h3>List of Good Cat Foods</h3>
<h2>Blue Buffalo</h2>
<h2>Purina One</h2>
<h2>IAMS</h2>

....be specific and maintain htag hierarchies:

<h2>Best Cat Foods Recommended by Vets</h2>
<h3>Blue Buffalo Tastefuls</h3>
<h3>Purina One Indoor Advantage</h3>
<h3>IAMS Protective Health</h3>

Group closely related topics together within a page section. Use conceptual boundaries to identify sections needed and avoid arbitrary rules, such as making each section of equal length.

For example, instead of setting a word count goal of 500 words per section, you organize your sections around complete ideas, topics, or logical breaks. This might mean one section covers an entire concept in 400 words, while another covers a complex topic in 800 words.

Lead sections with concise summaries: add a “Key takeaway” to page sections to help Gemini summarize the content accurately.

Example: Each article on the Nielsen Norman Group website features a short summary at the beginning.

Summaries help readers scan pages, too, so they’re helpful to your audience.

Next, make content scannable. Use comparison tables and bulleted or numbered lists where appropriate.

Keep in mind that AI models like Gemini process structured content more efficiently than dense paragraphs.

Support clarity by reducing fluff (e.g., unnecessary adjectives and adverbs), which can make content hard to read.

Use short introductions and paragraphs. Use short sentences where possible, but allow for sentence length variation, which is a feature of compelling writing.

One idea per paragraph: To optimize for readability and AI retrieval, keep text blocks focused on a single idea. Otherwise, the AI Overview may have difficulty finding the answer it’s looking for.

Avoid leading with ambiguous pronouns: Instead of beginning a paragraph with “It,” “That,” or “This,” name the subject. Otherwise, the AI Overview may make a wrong inference, resulting in an inaccurate citation.

Case in point: A colleague noticed an AI Overview, which claimed that Semrush offers a 50% discount to nonprofits. That is inaccurate. The AIO pulled the information from a software review. Specifically, it was from a paragraph about Asana, which began with “The software.”

While the meaning of “The software” would be clear to any human reader, it was not clear to Gemini.

Subject-predicate-object: Consider defining technical terms with a subject-predicate-object (SPO) construction.

AI models, including Gemini, seek to discern semantic relationships between entities (things). The subject-predicate-object structure mirrors how these models internally represent knowledge:

• Subject: The entity being discussed
• Predicate: The relationship or action
• Object: What the subject relates to or acts upon

This structure makes it easier for Gemini to extract clear, factual statements for AI Overviews.

Consistent terminology: Use consistent terminology and define acronyms.

Shifting between different terms for the same concept may confuse Gemini. That said, including a paragraph that informs the reader of alternative terminology is fine.

Here’s an example of including alternative terms:

“Other terms being used by marketers for GEO include AIO (Artificial Intelligence Optimization) and AEO (Answer Engine Optimization).”

Authority and Trustworthiness

Brand mentions across the web and original research can boost AI Overview citations.

The top 3 correlation factors for a brand showing up in AI Overviews are:

• Brand web mentions
• Brand anchors
• Brand search volume

Cite sources: Back claims with citation of trustworthy sources. Where appropriate, cite data or statistics.

Also consider publishing quotes from experts. Insider quotes can be particularly helpful when discussing how-tos or trends.

For instance, Emily Gertenbach’s Exploding Topics article on programmatic SEO taps the expertise of innovators like Sai Deshmukh, Director of Organic Growth at Amber.

The inclusion of original quotes makes for an engaging read. If cited in AI Overviews, these quotes would require a link or brand mention.

Publish original research: Consider commissioning original research to increase brand mentions across the web.

At Exploding Topics, we’ve run a number of original research studies and produced helpful and insightful content about the results.

Most recently, we surveyed users on their use of AI at home and at work.

Monitor brand mentions: There are several Semrush apps that can provide monitoring and give you ideas for increasing web mentions.

• Brand monitoring: Use this app to track social and press mentions plus backlinks.
• Media monitoring: Analyze sentiment, access brand mentions, and turn them into backlinks
• Mentioned - Outreach Wizard: Simplifies building connections with people and brands mentioned in your content.
• Influencer Analytics: Find the right influencers for your clients and manage everything from outreach to campaigns and reporting.
• Trend Analytics for Reddit: Makes it easy to find communities and trends relevant to your niche.

Focusing on GEO vs SEO

When you optimize textual content for AI Overviews, is that SEO (Search Engine Optimization) or GEO (Generative Engine Optimization)?

Ranking in the top 10 (one outcome of good SEO) only has a “positive moderate correlation” to getting cited.

So while ranking high may increase your chances of being cited, clearly, other factors are at work, too.

Public comments by top analysts and consultants show AI Overviews are continuing to have a significant impact across industries, particularly those focused on informational content.

Marketing innovator Avinash Kaushik laments that most businesses are doing nothing to optimize for AI. He estimates that heading into 2026, brands stand to lose as much as 30% of their organic traffic.

The time to take action is now, whether you call this an extension of SEO, or a whole new process: GEO.

• Commit to learning. Identify resources, such as blogs and newsletters, that will help you track the impact of AI Overviews.
• Maintain your commitment to SEO, particularly to creating authoritative, helpful, and trustworthy content.

Agencies that act now will capture a major competitive advantage. Start tracking AI Overview visibility with Semrush and implement GEO strategies today, before your competitors do.