Showing Up in AI Overview: 9 Things We've Learned

We research digital businesses every day, if you're building in this space, get our market clarity reports

Google AI Overviews changed how search results work, and most people are still figuring out what actually matters.

The traditional SEO playbook doesn't apply here because AI Overviews pull citations based on different signals than regular rankings.

Practitioners who tracked thousands of AI Overview appearances discovered patterns that contradict what we thought we knew about search visibility.

We've analyzed these findings to extract what actually works when you want your content cited in AI Overviews, and some of these insights will surprise you (especially the one about brand mentions beating backlinks by 3x, or how Reddit holds 22% influence despite being only 4% of training data), which is exactly the type of research we compile in our market clarity reports.

Learning	Details
We learned that winning traffic across topic clusters beats individual keyword rankings	AI Overviews don't just look at your rankings for individual keywords. They pick sites that dominate traffic across entire topic clusters using a "query fan-out" technique that splits queries into multiple sub-queries exploring different subtopics simultaneously.
We learned that brand mentions correlate 3x stronger than backlinks for AI visibility	Ahrefs studied 75,000 brands and found branded web mentions correlate at 0.664 with AI Overview appearances while backlinks only correlate at 0.218. LLMs are trained on text, not link graphs, so brand discussions matter more than hyperlinks.
We learned that vendors creating comparison content get cited at 7% rate	Software vendors creating honest comparison guides (featuring their product first, then competitors) get cited by AI Overviews at ~7% for "best X" queries. The AI doesn't distinguish vendor bias when the content ranks well and provides comprehensive coverage.
We learned that schema quality determines AI inclusion, not just presence	Controlled experiments showed pages with complete, well-implemented schema appeared in AI Overviews while pages with poor schema never did, even when ranking for more keywords. The AI evaluates schema completeness and proper formatting, not just existence.
We learned that 44% of AI citations come from pages outside top 10	AI Overviews operate on dual architecture where the LLM generates text and Google's index separately provides citations. This means pages ranking #15 can get cited if they contain the exact information the AI summary discusses, regardless of traditional ranking position.
We learned that Reddit holds 22% influence despite being 4% of training data	Reddit became the #1 most cited domain in AI Overviews, growing from 1.30% citations (March 2024) to 7.15% (June 2024). LLMs weight authentic, conversational content more heavily, and Reddit's discussion format provides the natural language patterns AI models prioritize.
We learned that impressions rise 49% while CTR drops 30% with AI Overviews	Google counts showing your URL in AI Overview sources as an "impression" even when users don't click. This creates a shift from "traffic economy" to "visibility economy" where content gets seen and attributed but not necessarily visited.
We learned that 81% of AI citations originate from mobile searches	Google's mobile-first indexing means your mobile version determines AI Overview eligibility. Combined with 67% of searches happening on mobile, the algorithm heavily prioritizes mobile-optimized content for citations.
We learned that conversions stay stable despite 60% traffic drops	AI Overviews filter out low-intent browsers by answering basic queries directly. Users who still click through after reading the AI summary show higher purchase intent, acting as a pre-qualification mechanism that maintains conversion rates despite lower volume.

1/ You need traffic share dominance across topic clusters for Google AI Overviews

We have learned that Google's AI Overview doesn't just look at your rankings for individual keywords when deciding which sites to cite.

Field testing shows the algorithm picks sites that dominate traffic across entire topic clusters, basically sites that win traffic for multiple related searches in the same area. This happens because AI Overviews use a technique called "query fan-out" where a single query splits into multiple sub-queries exploring different subtopics at once, then pulls from sources that cover the whole information ecosystem.

The algorithm looks for sites with broad topical authority because they show genuine expertise, not just keyword tricks. The AI needs comprehensive, connected information to generate quality summaries, so sites that only rank for isolated keywords get skipped.

Map out the complete semantic territory for your niche topics and create content hubs covering all angles of a topic cluster instead of just optimizing individual pages for specific keywords.

We have market clarity reports for more than 100 products — find yours now.

2/ You'll see brand mentions outperform backlinks 3x for AI Overview visibility

We have learned something that contradicts everything traditional SEO taught us about link building.

The largest study analyzing 75,000 brands by Ahrefs discovered that branded web mentions correlate at 0.664 with AI Overview appearances, while traditional backlink volume only correlates at 0.218. Branded anchors correlate at 0.527, and even Domain Rating only reaches 0.326, so your brand being discussed across the internet (even without links) matters way more than your backlink profile.

This pattern exists because Large Language Models train on text content, not link graphs. When AI systems see your brand mentioned frequently across different contexts, they weight it as more authoritative than isolated backlink profiles.

Shift 30-40% of your link-building budget toward "mention-building" campaigns focused on getting brand discussions in Reddit threads, Quora answers, industry forums, and guest contributions (even without backlinks).

3/ You can get cited in AI Overviews by creating vendor comparison content

We have seen people mentioning that software vendors who create their own comparison content get cited by AI Overviews at surprisingly high rates.

Analysis of 8,000+ AI citations discovered something counterintuitive: vendors creating comprehensive comparison content (featuring their own product first, then honest competitor analysis) get cited at approximately 7% for "best X" and "top Y" queries. Companies like Thinkific, LearnWorlds, Monday.com, Pipedrive, SE Ranking, and HP successfully appear in AI Overview comparisons for their categories.

This works because it fills a gap where independent comparison content is sparse, the vendor content ranks well due to topical authority, and AI Overviews use it without checking for bias. The AI picks comprehensive, well-structured comparison content over objectivity verification, so if your comparison ranks well and covers everything thoroughly, the AI doesn't filter it out.

Create extensive comparison guides on your site featuring your product or service first, followed by 5-10 legitimate competitors with honest feature comparisons, structured with clear tables, pros/cons sections, and FAQ schema (this works especially well in niche B2B categories where independent review sites haven't covered the space yet, similar to how we identify content gaps in our market clarity reports).

4/ You need quality schema markup to appear in Google AI Overviews

We have learned from controlled experiments that schema markup quality matters way more than just having schema present on your pages.

A controlled experiment published by Search Engine Land tested three identical single-page sites with different schema setups. The page with complete, well-implemented schema ranked for 6 keywords, hit Position 3, and appeared in AI Overviews, while the page with incomplete schema ranked for 10 keywords but peaked at Position 8 and never appeared in AI Overviews.

AI Overviews don't just check for schema presence, they evaluate schema completeness, proper formatting, and semantic richness. Poor setups signal low content quality to Google's algorithms, so complete, accurate schema helps AI parse content architecture, understand entity relationships, and extract relevant information efficiently.

Audit your existing schema using Google's Rich Results Test and Schema Markup Validator, then go beyond basic Article schema to add FAQ schema for Q&A content, HowTo schema for instructions, comprehensive author info, reading time, word count, and educational level targeting.

Each of our market clarity reports includes a study of both positive and negative competitor reviews, helping uncover opportunities and gaps.

5/ You can get cited in AI Overviews even when ranking outside top 10

A lot of entrepreneurs reported discovering that AI Overview citations don't require top rankings the way traditional SEO does.

Writesonic's study revealed that user-generated content and comprehensive resources have approximately 44% chance of appearing in AI Overviews even when ranking outside the top 10 organic positions. Advanced Web Ranking's research backs this up, showing 46.5% of URLs cited in AI Overviews rank outside the top 50 positions, which throws out traditional SEO assumptions that only top 3-5 positions matter.

This happens because AI Overviews work on dual architecture: the LLM generates summary text while Google's search index separately provides citation links. The citation selection picks content that semantically matches the AI-generated answer and satisfies search intent, regardless of traditional ranking position.

Stop obsessing over top 3 rankings for every keyword and instead create exceptionally comprehensive, well-structured content targeting informational queries where you can show clear expertise, focusing on "citability" factors like clear Q&A structure, concise direct answers, strong E-E-A-T signals, proper schema markup, and logical content hierarchy.

6/ You should prioritize Reddit for disproportionate AI Overview citation weight

We have seen people mentioning that Reddit's influence on AI systems far exceeds what its size would suggest.

Industry analysis reveals Reddit made up only 4% of GPT-3 training data yet holds approximately 22% weight in LLM influence and visibility, creating a 5.5x multiplier. In Google AI Overviews specifically, Reddit became the #1 most cited domain, growing from 1.30% of citations in March 2024 to 7.15% in June 2024 (a 450% increase), while Reddit traffic nearly tripled from 132M visits in August 2023 to 420M globally in February 2024.

LLMs weight authentic, conversational, experience-driven content more heavily than formal content. Reddit's discussion format provides the natural language patterns, diverse perspectives, and real-world experiences that AI models prioritize, plus the upvoting system creates quality signals and threaded discussions offer nuanced information that AI systems value (similar to how we prioritize real user signals from forums and review sites when compiling our market clarity reports).

Develop an authentic Reddit engagement strategy by identifying subreddits where your target audience discusses relevant topics, building karma by providing genuine value for 2-3 months without self-promotion, then strategically mentioning relevant projects when contextually appropriate.

Our market clarity reports contain between 100 and 300 insights about your market.

7/ You'll see impressions rise 49% but CTR drop 30% with AI Overviews

We have learned that AI Overviews create a paradoxical pattern that confuses most site owners looking at their Google Search Console data.

BrightEdge's year-long study tracking 300,000+ keywords revealed that when AI Overviews appear, page impressions increase 49% year-over-year while click-through rates decline 30%. Multiple publishers report the same thing: Charleston Crafted saw impressions stay stable but traffic dropped 75%, and one Reddit user reported impressions up 66% while clicks dropped 50%.

Google counts showing your URL in AI Overview sources as an "impression" even though users read the AI-generated summary without clicking through. The AI Overview answers the question directly, so users see your brand but don't need to visit your site.

Completely reframe your success metrics by stopping treating impressions as a proxy for traffic potential (they now represent brand visibility and authority signaling, not click probability) and focus on getting cited in AI Overviews for brand visibility while optimizing different content for transactional, high-intent keywords that still drive clicks.

8/ You must optimize for mobile to capture 81% of AI Overview citations

A lot of entrepreneurs reported being surprised that mobile optimization directly impacts AI Overview citation probability more than expected.

Ahrefs analysis of AI Overview citation patterns discovered 81% of citations originate from mobile search results, despite desktop and mobile showing dramatically different user experiences. On mobile, AI Overviews take up entire screen space (48% vs. 42% on desktop), pushing organic results below the fold, yet the citation selection heavily favors mobile-optimized content.

Google's mobile-first indexing means the mobile version of your site determines rankings and AI Overview eligibility. Combined with the fact that 67% of searches happen on mobile devices, the algorithm heavily prioritizes mobile-optimized content for citations.

Run mobile-specific content audits using Google's Mobile-Friendly Test and PageSpeed Insights focusing exclusively on mobile scores, optimize Core Web Vitals specifically for mobile, make sure critical content appears above-the-fold on mobile without horizontal scrolling, and test all structured data on actual mobile devices (not just responsive design testing).

9/ You'll maintain stable conversions despite 60% traffic drops from AI Overviews

We have seen people mentioning something counterintuitive about how AI Overviews affect their business metrics beyond just traffic numbers.

Multiple case studies reveal that when sites lose 60% of traffic due to AI Overviews, conversion rates often stay stable or even improve. Conductor reported traffic drops of 60% on some pages but conversions only fell slightly, suggesting lost traffic wasn't highly valuable, while HubSpot's CMO stated they "never cared about vanity metrics like traffic" as conversion metrics held firm despite 75% traffic decline.

AI Overviews answer basic informational queries without clicks, filtering out low-intent browsers. Users who click through have already consumed the AI summary and want deeper information, specific products, or services, so the AI Overview acts as a pre-qualification mechanism that reduces traffic volume while increasing traffic quality.

Completely reframe your success metrics from traffic volume to traffic value by segmenting analytics to track which queries show AI Overviews and analyzing conversion rates specifically for AI-referred traffic versus traditional organic (you may discover AI Overview clicks convert at 2-4x higher rates despite lower volume, which is exactly the type of conversion insight we track in our market clarity reports when analyzing different customer acquisition channels).

In our market clarity reports, for each product and market, we detect signals from across the web and forums, identify pain points, and measure their frequency and intensity so you can be sure you're building something your market truly needs.