Brands trying to improve their visibility in AI Search face a difficult question:
Should you optimize differently for ChatGPT than you do for Google AI Mode?
According to a new Semrush study, the answer is both yes and no.
Semrush’s August 2026 AI Visibility Index Mini Study: ChatGPT and Google AI Mode: A Tale of Two Models analyzed AI visibility across 22 industry categories using data from January through June 2026. The underlying AI Visibility Index includes a database of 126 million U.S. AI search prompts, making this one of the more substantial datasets available for understanding how brands and sources surface inside AI-generated answers.
The findings reveal something marketers need to understand quickly:
ChatGPT and Google AI Mode often agree on which brands matter, but they can differ significantly in where they go for evidence about those brands.
That distinction has major implications for SEO, GEO, content marketing, digital PR, reputation management, and brand authority.
It also reinforces an increasingly important reality of AI Search:
Your website is only one part of the information ecosystem AI uses to understand your brand.
Here are the biggest findings from Semrush’s research—and what we believe brands should do about them.
1. ChatGPT and Google AI Mode Agree on Many of the Brands That Matter
One of the most important findings in the report is how much agreement exists between the two platforms.
Across the 22 categories Semrush studied, the most-mentioned brands on ChatGPT and Google AI Mode overlapped by approximately 69%.
Semrush also found that six brands maintained top-10 positions for mentions on both platforms every month included in the study.
That matters because it suggests AI visibility is not completely platform-specific.
Strong brands appear to develop authority signals across the broader web that can influence multiple AI ecosystems.
In other words, marketers should resist the temptation to think of ChatGPT optimization, Google AI optimization, Gemini optimization, and every future AI platform as completely independent disciplines.
There appears to be a shared foundation.
Brands that are widely discussed, consistently represented, well understood, and validated across credible sources have an opportunity to surface across multiple AI experiences.
This is one of the most important strategic takeaways from the entire report:
Build authority across the web, not just visibility inside one AI platform.
That does not mean every platform behaves identically.
Far from it.
The biggest differences emerge when you examine which sources the platforms cite.
2. ChatGPT and Google AI Mode Often Trust Different Sources
This may be the most actionable finding in Semrush’s study.
ChatGPT and Google AI Mode may recommend similar brands while relying on very different sources to support their answers.
Semrush found that ChatGPT heavily favors community and reference content.
Reddit accounted for 28.9% of ChatGPT citations, while Wikipedia accounted for another 23.1%. Combined with other community and reference sources, those categories represented approximately 59% of ChatGPT’s citations in the analysis.
Google AI Mode looked considerably different.
Its leading citation source was YouTube at 21.1%, followed by Facebook at 17%. Overall, nearly half of Google AI Mode’s citations came from social and video sources.
Think about what that means strategically.
A brand trying to improve AI visibility cannot simply publish another 50 articles on its own website and assume the job is finished.
For ChatGPT, the conversation happening on Reddit or the factual context appearing within authoritative reference sources may matter.
For Google AI Mode, video, social platforms, reviews, and other highly visible third-party ecosystems may play a larger role.
This is why AI Search optimization must extend beyond owned content.
Your website remains foundational.
But the question marketers increasingly need to ask is:
Where else does the internet provide evidence that our brand deserves to be included in the answer?
That is also where strategic third-party brand mentions become increasingly important.
Mention Engine was built specifically around this challenge. Rather than asking marketers to manually search thousands of potential publishers and placements, Mention Engine uses AI to evaluate more than 70,000 brand mention opportunities and recommend a strategic portfolio around the brand, market, goals, and budget.
Explore Mention Engine and see how strategic third-party mentions can strengthen your AI authority.
3. Being Mentioned and Being Cited Are Not the Same Thing
Semrush calls this the Mention-Source Divide, and it deserves much more attention from marketers.
A brand can appear frequently in AI answers without its own website being frequently cited.
Disney provides one of the report’s clearest examples.
Semrush recorded roughly 820,000 ChatGPT mentions of Disney, while disney.com was cited only about 50,000 times.
On Google AI Mode, the gap was even larger: roughly 981,000 mentions compared with approximately 23,000 citations to Disney’s own domain.
Nike, Toyota, Marvel, and Netflix demonstrated similar patterns.
Why?
Because AI does not necessarily need a company’s website to talk about that company.
The broader web talks about Disney.
Publishers review Disney.
Communities discuss Disney.
Travel sites describe Disney.
Reference sites document Disney.
Customers evaluate Disney.
AI can therefore assemble an understanding of the brand from information distributed across the internet.
The opposite pattern also exists.
Semrush found that Wikipedia was cited dramatically more often than it was mentioned as a brand. Tripadvisor, IMDb, LinkedIn, and other information-rich platforms showed similar tendencies. These properties often function less as subjects of AI conversations and more as sources powering those conversations.
This creates an important distinction for AI Search strategy:
Brand visibility and citation visibility are related, but they are not interchangeable.
A sophisticated AI Search program should therefore measure both.
Are AI platforms mentioning your company?
Are they recommending it?
What sources are influencing those recommendations?
Which third-party websites are being cited when your company appears?
And which sources are supporting competitors instead?
Those questions reveal much more than simply tracking whether your own website earned a citation.
4. AI Creates “Consideration Sets”—and Your Competitors Matter More Than You Think
One of the most interesting concepts in the Semrush study is the Consideration Set.
AI systems frequently do not mention brands individually.
They group them.
For example, when ChatGPT mentioned PayPal, Semrush found Apple Pay appearing alongside it 86 times per 100 PayPal mentions, Venmo 77 times, and Stripe 72 times.
In effect, AI has learned that these brands belong in the same competitive conversation.
The same phenomenon appears across industries.
A consumer asking AI for recommendations is therefore not simply triggering a decision about whether your company deserves to appear.
AI may effectively be asking:
Which group of companies belongs in this answer?
That fundamentally changes competitive research.
Traditional SEO often focuses on:
- Which competitors rank for my keywords?
- Who has more organic traffic?
- Who has stronger backlinks?
- Who outranks me?
AI Search adds another question:
Who does AI consider my competitors?
The answer may not match your traditional SEO competitor list.
Semrush also identified asymmetric relationships where smaller brands frequently appeared alongside category leaders, while the category leaders appeared without those smaller brands.
That creates what might be called recommendation dependency.
The strategic objective is not merely to enter the consideration set.
It is to strengthen your authority enough that your brand becomes a consistent, independent member of that set.
For marketers, that means competitive AI visibility monitoring should track co-mentions, not simply raw mention volume.
5. Every Industry Has Its Own “Citation Core”
This may be the Semrush finding most directly relevant to brand authority strategy.
Reddit, Wikipedia, YouTube, and Facebook may dominate citations at the platform level, but Semrush found that when you drill into individual industries, specialized publishers become extremely important.
Semrush calls this the Category Citation Core.
For example:
- Health and fitness relies heavily on sources such as Healthline, WebMD, Healthgrades, and Cleveland Clinic.
- Automotive relies on Cars.com, Carfax, Edmunds, and Kelley Blue Book.
- Travel is heavily influenced by Tripadvisor.
- Sports relies significantly on ESPN and other specialist sports sources.
- Government queries strongly favor authoritative government domains, with NIH accounting for more than half of citations in Semrush’s analysis of that category.
The lesson is simple:
There is no universal list of websites every brand should target for AI visibility.
A healthcare brand and a software company should not have the same third-party authority strategy.
Neither should a hotel, law firm, ecommerce company, financial services provider, or home services company.
Each industry has its own trusted ecosystem.
And importantly, Semrush found that many of these specialist Citation Cores overlap between ChatGPT and Google AI Mode.
That means earning authority within the right industry-specific sources may provide value across multiple AI platforms simultaneously.
This aligns closely with how we think about brand mentions at LSEO.
The objective should not be:
Get more mentions.
It should be:
Get the right mentions, in the right contexts, from the sources that matter within your category.
That is precisely why Mention Engine recommends a portfolio of opportunities instead of treating every mention as equally valuable.
A comparison article can send a different signal than an industry ranking.
An educational resource serves a different purpose from a directory.
A recommendation article creates a different context from a general brand profile.
Context matters.
6. Some Industries Can Use One AI Strategy. Others Cannot.
Semrush divided the industries it studied into three broad groups: settled, contested, and middle-band categories.
Settled categories showed relatively high agreement between ChatGPT and Google AI Mode.
Government led the group, while health, automotive, finance, and education also demonstrated significant overlap.
Semrush argues that these categories tend to depend on factual, authoritative sources where the cost of inaccurate information can be higher. As a result, AI platforms frequently converge around similar authorities.
At the other extreme are contested categories.
Beauty showed particularly low agreement in brand visibility, while Business and Professional Services had only about 42% source overlap, the lowest source overlap Semrush reported among its categories.
Fashion, food, and digital technology also showed meaningful divergence.
The implication is important:
Your AI Search strategy should reflect your industry.
A healthcare organization may be able to build authority around a relatively stable group of trusted sources that influences multiple AI platforms.
A technology or professional services company may require a more diversified strategy because ChatGPT and Google AI Mode are drawing from different information environments.
This is another reason generic “GEO checklists” are insufficient.
The correct strategy depends on:
- Your industry
- Your competitors
- The AI platforms your audience uses
- Your existing authority
- The third-party sources influencing your category
- The contexts in which buyers evaluate your brand
AI visibility is becoming a portfolio problem, not a single-channel optimization problem.
7. The Source Landscape Is Stable—But Not Static
Semrush’s six months of data also reveal an important tension.
At a high level, many of the dominant sources remained remarkably consistent.
Google AI Mode drew its monthly top-five citation sources from the same six domains throughout the study: YouTube, Facebook, Reddit, Wikipedia, Google, and Instagram.
ChatGPT’s leading sources came from a similarly small recurring group that included Wikipedia, Reddit, Facebook, YouTube, Yahoo, Fandom, and later Apple.
But the weight assigned to those sources changed.
Between January and June, YouTube’s share of Google AI Mode citations rose from approximately 18% to 24%.
Meanwhile, ChatGPT’s reliance on Reddit increased sharply, reaching 43% in June, while Wikipedia’s share declined from approximately 26% in January to 17% in June.
Some verticals changed even more dramatically.
In finance, for example, Investopedia’s share of ChatGPT citations reportedly fell from approximately 20% in January to under 3% in June while market-data sources gained ground.
That gives marketers an important warning:
Do not build an AI visibility strategy once and assume it is finished.
The ecosystem is changing too quickly.
Brands should continually monitor:
- AI mentions
- AI citations
- Competitor co-mentions
- Citation sources
- Changes in source weighting
- Category-specific authority sites
- Emerging opportunities
The winning strategy will be adaptive.
What Semrush Recommends Marketers Do Next
The report concludes with a practical seven-part playbook.
Semrush recommends maintaining accurate branded reference information, investing seriously in YouTube, participating authentically in Reddit communities, improving the customer experience to support reviews and discussion, identifying your industry’s Citation Core, creating platform-specific strategies where needed, and tracking both mentions and citations across your Consideration Set.
We agree with the broader principle behind that playbook.
AI Search optimization cannot live entirely on your website.
Companies need to think about their entire authority footprint.
That includes owned content, social content, video, reviews, community conversations, authoritative publishers, directories, rankings, comparisons, and credible third-party mentions.
No single tactic creates AI authority.
The objective is to build enough consistent evidence across the web that AI systems can confidently understand:
- Who your company is
- What you do
- Who you serve
- Where you operate
- What you are known for
- How you compare with alternatives
- Why you deserve consideration
Where Mention Engine Fits
The Semrush study does not test Mention Engine, and it would be inappropriate to interpret its findings as proof that any specific placement will cause an AI platform to recommend a brand.
What the research does demonstrate is why third-party authority deserves a formal place inside AI Search strategy.
Brands are being discussed through sources they do not own.
AI platforms rely on different types of external evidence.
Industries have specialized Citation Cores.
Brands are evaluated alongside competitive Consideration Sets.
And being mentioned is not the same thing as being cited.
Mention Engine was created to help brands operationalize this part of the AI visibility challenge.
Rather than manually researching thousands of potential third-party opportunities, marketers can use Mention Engine to evaluate a large universe of brand mention opportunities and build a strategic portfolio aligned with their business, category, goals, audience, geography, and budget.
For agencies, the same platform can be used across multiple clients, with multi-user access and white-label reporting available to support scalable campaigns.
The philosophy is straightforward:
Don’t simply ask where you can get mentioned. Determine where your brand should be mentioned—and why.
That distinction becomes increasingly important as AI Search evolves.
The Biggest Takeaway: Optimize for Recommendation, Not Just Ranking
Semrush’s study reinforces something we have been saying at LSEO for some time:
The internet is shifting from search toward recommendation.
SEO remains foundational.
Your website still matters.
Technical performance still matters.
Content still matters.
Rankings still matter.
But AI-powered discovery introduces another layer.
A brand can rank well and still be underrepresented in AI recommendations.
A competitor can have a weaker website and still benefit from stronger third-party authority.
And a prospective customer can receive an AI-generated shortlist before ever visiting a search result or company website.
The brands that adapt will therefore think beyond individual webpages.
They will build the broader authority required to become part of the answer.
They will understand which sources influence their category.
They will monitor the competitors AI associates with them.
They will create content for multiple information ecosystems.
And they will actively strengthen their presence across credible third-party sources.
That is what Recommendation Visibility™ is ultimately about:
Not simply whether AI can find your company—but whether it has enough trusted evidence to recommend it.
Semrush’s research gives marketers another compelling dataset showing just how much of that evidence exists beyond the brand’s own website.
The question now is what you do with it.
Ready to start building the third-party authority behind stronger AI visibility? Explore Mention Engine and create your first strategic brand mention plan. New users currently receive a limited-time $100 Launch Credit.
Source: Semrush, AI Visibility Index Mini Study, “ChatGPT and Google AI Mode: A Tale of Two Models,” August 2026. The study analyzed January–June 2026 U.S. data across 22 industry categories using Semrush’s AI Visibility Index and its database of 126 million U.S. AI search prompts.