The AI Visibility Guide: How We Earn Hundreds of Brand Mentions and Citations

AI visibility is now a critical part of SEO.

If you’re serious about making your brand show up in ChatGPT, AI Overviews, Perplexity and more, you need to implement a concerted strategy across multiple fronts.

And you need to understand that there are two very distinct aspects to AI optimization. There’s a significant difference between optimizing for citations (that link to your site) and brand mentions (that are unlinked).

The truth is that fewer than 30% of the brands most mentioned by AI are also among the most cited.

At Exploding Topics, we created a well-thought-out AI visibility strategy for both citations and mentions based on hard data and monitoring from Semrush’s Enterprise AIO and found a lot of success.

Want to copy our strategy? Here’s how we did it.

First, Get Data on AI Visibility

First and foremost, you need accurate data to drive the creation (and effective monitoring) of your AI visibility strategy.

To begin, check out Semrush’s new AI Visibility Index study to understand which brands and sources are winning in your specific industry and identify the gaps between brand mentions and source citations.

The AI Visibility Index analyzed search prompts across five major industries. It also looked at both ChatGPT and Google AI Mode, surfacing some fascinating insights into the differences between the two LLMs.

For instance, whereas ChatGPT leans heavily on Reddit and Wikipedia, Google AI Mode is far less reliant on any single source of information. Its favored sources vary more significantly by industry.

Moreover, Semrush found a significant gap between brands that get cited in AI vs brands that get mentioned. Later, we’ll go over our strategies for both.

While the AI Visibility Index gives you the big picture of your industry, Semrush's Enterprise AIO takes it further by tracking your specific brand's performance in real-time. The platform lets you track, analyze, and benchmark brand mentions across virtually all top LLMs and AI search engines, like ChatGPT, Perplexity, Claude, Gemini, and more, giving you insights into exactly how your brand appears in AI-generated responses.

With Enterprise AIO, you can track your competitors alongside your own performance. Use the Competitor Rankings dashboard to see your competitors' share of voice, position changes, and mention frequency, with filters to see how your competitive landscape differs between ChatGPT, Google AI Mode, and other platforms.

The platform also features hallucination detection, citation tracking, and a lot more.

Together, the AI Visibility Index and Enterprise AIO give you both the data and monitoring you need to build an effective AI visibility strategy. The Index shows you which brands dominate your industry and which sources AI tools favor, while Enterprise AIO tracks your specific performance and lets you respond quickly when your AI presence changes.

This data-driven approach is what allowed us at Exploding Topics to win more AI visibility, and it's the foundation for the specific tactics we'll share next.

The Case Study: Our Original AI Research

We’re going to demonstrate the real-world effectiveness of an AI strategy that implements the findings of the Semrush visibility study, and show you how you can copy it. To do that, we will use the example of an original piece of research that Exploding Topics published in July 2025.

Our study was on the AI trust gap, examining consumer attitudes toward AI content. It has since been shared in major publications like eMarketer, ZDNET, and Inc.

And crucially for our purposes, it has also helped to increase Exploding Topics’ AI visibility.

Let’s dive into exactly what that means.

The Goal: AI Citations and Brand Mentions

It’s all very well saying that you want to increase your AI visibility. But in order to do that, you first need to understand it.

It’s helpful to break AI visibility up into citations and brand mentions. They are two separate concepts, and you will need to adopt different strategies to achieve them both.

AI Citations

An AI citation is when you are listed as a source in an AI response. Here’s an example of being cited in a Google AI Overview:

In order to be cited by AI, you need to create quality content. Ask yourself if your content has these characteristics:

• Directly answers the user query
• Demonstrates expertise with original, non-commodity content
• Transparently discloses other EEAT signals (expert bio with credentials, about page, editorial policy etc.)

You don’t have to be the brand with the most citations. You just need to win citations to queries that matter to your business.

For instance, the Semrush AI Visibility Index found that Zapier is the number one most AI-cited brand in the digital technology and software space.

As a workflow automation tool, Zapier publishes content that touches on A LOT of niches and use cases.

So it’s natural that Zapier should be cited very frequently as long as it’s producing high-quality content.

Your benchmark could be different depending on your particular industry and direct competitors.

Besides, it’s also clear that achieving lots of AI citations does not necessarily set you up for brand mentions.

This is why Zapier is way down in 44th for AI brand mentions despite being the top-cited brand in AI answers.

Brand Mentions

A brand mention is when an AI tool mentions your brand by name within its response to a user prompt.

For example, here’s Exploding Topics getting the top mention from ChatGPT, in response to a question about the best tools to find emerging trends:

One thing you’ll immediately notice is that unlike a citation, a mention does not necessarily lead to a link.

But that doesn’t mean AI brand mentions are a waste of time. Far from it.

Your potential customers are already using AI to help them make decisions. 63% are using it to conduct research.

A (positive) brand mention can be absolutely invaluable. ChatGPT users are closer to making a purchase by the time they reach your site.

According to Semrush, the average AI search visitor (tracked to a non-Google search source like ChatGPT) is 4.4 times as valuable as the average visit from traditional organic search, based on conversion rate.

Findings admittedly vary, with another recent study finding negligible differences in conversion rates between organic search and LLM traffic.

But regardless, common sense makes one thing clear: if a user is asking AI about trend discovery tools, that’s exactly the kind of person we’d want to tell about Exploding Topics.

Strategy 1: Establishing Brand Trust

Our first 4-step strategy looks at building trust in order to earn AI citations.

And it works. Exploding Topics is being routinely cited by AI, and the data proves it.

Our domain features in the Google AI Overview for 3,800 keywords.

The traffic from those citations alone is worth more than $28,000.

Here’s how we do it:

Original Research

It’s always worth remembering that AI is designed to try and provide its users with helpful, insightful information. Becoming a provider of that information is therefore an excellent first step to picking up citations.

We have made a conscious decision to publish more original research. And we have picked up AI citations as a direct result of our proprietary studies.

For instance, here’s what happens when you ask ChatGPT whether users trust AI Overviews. Our study is prominently cited three times within the first three headings:

In terms of direct referrals, our study only received about 4% of its total traffic from AI chatbots. But that still amounted to over 325 visits, picking up referrals from ChatGPT, Perplexity, Gemini, Grok, and Copilot.

And funnily enough, the study itself revealed that being cited as a source in AI is by no means a guarantee that users will click through to the original material. Indeed, only 8% of people who have encountered Google AI Overviews always visit the base text.

So compared to the number of people who actually clicked through to our research, the number of AI citations we received could plausibly be around 10x higher.

Optimize for AI

Original research provides LLMs with the insight they are trying to pass onto their users. But in order to reap the benefit, you have to make sure that AI can easily parse your content.

Simply put, if AI cannot read it, then it will not cite it.

Your content needs to be clear, structured, and data-rich. It needs to strip back the marketing copy and provide genuine facts and figures.

The AI Visibility Index found that Bankrate is the number 1 top-cited finance source in Google AI Mode, and 6th in ChatGPT.

Even its homepage is full of useful information, updated daily.

We take the same approach in our content. Again, our AI study is a great example.

Our “fast facts” section gives an immediate breakdown of some of the standout data. We adopt this approach across most of our blog posts.

Once into the main body, we break down our content into clearly defined passages. Each passage should make sense on its own, so that it can be directly extracted by AI without difficulty.

In this instance, we referenced AI Overviews 4 times in 3 short paragraphs. An LLM reading this content would be able to easily establish the context for the data:

It’s a win-win: what works for AI also works for your readers. Easy-to-read, clearly-presented data without aggressive marketing spin makes for a better user experience.

We also follow technical optimization best practices so our content is easily visible, scannable, and readable to LLMs.

Technical Optimization for AI Visibility

Many technical SEO principles also apply to AI search optimization.

But here are some specific items to pay attention to for AI visibility:

Robots.txt and llmx.txt Settings

Check your robots.txt settings and ensure you’re not blocking any AI crawlers that matter to you.

It’s even better if you explicitly add specific AI crawlers to your allowlist. Here are a few AI crawlers to consider allowing:

• OAI-SearchBot
• ChatGPT-User
• PerplexityBot
• Claude-SearchBot
• Claude-User
• Perplexity‑User

In addition, I also recommend creating an llms.txt file for your site.

The llms.txt file uses markdown format that you can use to highlight important parts of your website.

You can include information about pricing, key features, use cases, documentation, and so on, along with appropriate links.

Even though there’s no consensus yet about llms.txt being helpful for AI search optimization, it’s easy to create one. So you might as well do it now.

That way, if llms.txt becomes more widely accepted in the near future, you'll already have a head start.

Minimal JavaScript Usage

LLMs have limited tokens and computational resources.

That’s why our main blog content uses simple HTML with minimal JavaScript.

Wrapping your content inside JavaScript only adds an obstacle for AI crawlers, so they’ll simply ignore your content if it’s not immediately accessible.

Make sure to frontload your content using only HTML for maximum AI readability.

Additional Technical Optimization Best Practices

• Reduce load speed: Review your website speed using tools like Page Speed Insights. Address any issues reducing your page speed. Target low values for Cumulative Layout Shift (CLS) for AI bots that snapshot your page early and provide minimum friction for agentic browsing.
• Minimize click distance to content: If your content is hidden behind too many clicks, such as dropdowns or interactive widgets, AI is likely to ignore it. Use a simple page structure and make your entire main content visible without additional interaction.
• Intentional schema implementation: Use appropriate schema types to match your content format, but don’t overuse them. Excessive or misleading schema can be seen as spammy by traditional search engines and increase HTML complexity for LLMs.
• Strong meta descriptions matter: Meta descriptions might not matter as much for Google search, but some AI chatbots like ChatGPT still read meta descriptions to determine content relevance. Try to be descriptive and even answer a question directly in your meta description, if possible.
• Provide descriptive image alt text: While LLMs are getting better at reading images, they still prefer extracting meaning from text because it’s less resource-intensive. You can minimize the processing load on AI by providing descriptive alt text for your images.

Other Steps: Price Transparency, Documentation

Not every step of your AI visibility plan will be relevant to every piece of content you publish. Here are some approaches not specifically incorporated into our AI study that have nonetheless helped Exploding Topics grow its authority.

Price transparency: An obvious occasion you might be cited is when an AI user is inquiring about the price of your product. But you’ll only pick up the citation if you’re actually up-front about costs.

“Contact sales” is not an uncommon sight on pricing pages. But LLMs prefer to give clear-cut answers; if you have not supplied your price directly, AI will scour forums to find figures.

That can lead to inaccuracies. It also means you lose control of the conversation, because there’s a risk that third-party discussions about your prices may come with negative sentiment attached.

Exploding Topics has a transparent payment structure. Each subscription tier also comes with a very clear description of the benefits:

As a result, LLMs like Google AI Mode are able to provide accurate information. And Exploding Topics picks up some easy citations.

Documentation: As with the pricing point, documentation is about making sure that you are the authoritative source for AI when it comes to discussion about your own brand.

If users have questions about how exactly your product works, or whether you can use it to perform a particular function, it is far better that AI responses point back to your own content.

That means keeping up-to-date and thorough documentation. This could be crawlable API pages, llms.txt files, or help centers.

It can also take the form of clear walkthrough articles. Exploding Topics has published lots of these:

Sure enough, when asking ChatGPT how to use the Exploding Topics API, this article is effectively treated as the single source of truth.

Strategy 2: Getting Into the Conversation

As well as picking up hundreds of citations, Exploding Topics is also getting brand mentions in AI responses.

Specifically, we’re getting mentions in response to non-branded queries. In other words, even when users aren’t asking about Exploding Topics, AI is mentioning our brand.

Semrush Enterprise AIO allows you to compare your “share of voice” with competitors across multiple AI platforms.

Among an automated selection of main competitors, Exploding Topics receives more Google AI Mode mentions than any other source, with the sole exception of Google Trends, in response to pertinent non-branded queries.

In Perplexity, Exploding Topics actually outperforms all of these competitors, boasting the highest share of voice.

How are we doing it? The key is brand sentiment.

Essentially, users like our product. And they are talking about how much they like it in all the right places for LLMs to find, resulting in overwhelmingly positive AI mentions.

Much of that is organic. But there are also actionable steps you should be taking.

Community discussions

If your brand is being talked about on Reddit, then AI will find those conversations.

ChatGPT in particular is remarkably reliant on Reddit as a source. In fact, Reddit is cited in 141.2% of ChatGPT responses related to business and professional services (i.e. it receives around 1.4 citations per response).

As a result, lots of positive discussion about your brand on Reddit can have an outsized effect on AI brand mentions.

For instance, Zoho has managed to achieve a 17.28% share of voice in the digital technology/software sector, behind only Microsoft, Amazon, and Google. That’s a vast overperformance compared to its market cap, and it can be partly attributed to an active Reddit community.

To a certain extent, you need to just let these conversations happen. Trying to actively engineer them by posting yourself can lead to negative sentiment.

But aside from simply making an excellent product, there are ways you can try and stimulate discussion on Reddit.

When you post, make sure you come with something that adds value and resources rather than strident self-promotion. Original research can come in handy here.

A Reddit user posted the results of another of our original surveys into the ChatGPT subreddit, which sparked significant and positive conversation:

And it’s not just Reddit. Google AI Mode demonstrates a preference for LinkedIn as a source.

As LinkedIn is a professional platform, it can be easier to talk more openly about your brand and its benefits.

But ideally, you still want the posts to be coming from the community. Once again, it all comes back to adding value.

Our AI content survey was shared widely on LinkedIn, resulting in positive mentions for Exploding Topics.

Depending on your niche, community platforms like Quora and Stack Overflow could also be important. Per the AI Visibility Index, Quora is a top-10 source for Google AI Mode for both fashion and business queries.

By getting people talking in the right places, you will improve the chances of AI mentions.

Review sites

As well as forum discussions, you ideally want users to be actively reviewing your product. A succession of high scores and high praise is easy for AI to understand, and boosts the likelihood that it will recommend your brand.

For a consumer market, that means encouraging reviews on sites like Amazon, Best Buy, and Yelp. For SaaS, you’ll want to focus on G2, Capterra, and TrustPilot.

Obviously, you need reviews to be authentic. But when you know you’re giving your users a good product, it’s smart to set up a systematic review request process.

The AI Visibility Index shows that Slack, Asana, and Monday all made it into the top 20 most mentioned brands within the digital technology space. They all fare well on G2.

Meanwhile, Exploding Topics has gained 40,000 likes within the Semrush App Center, as well as 5-star reviews from the founders of ClickFlow and Vivino.

Mentions in publications

AI loves user-generated content, but it also relies on other sources. Every industry has a set of trusted publications, where getting mentioned can boost your AI visibility.

In finance, it’s publications like NerdWallet or Investopedia — NerdWallet is cited in 75% of Google AI Mode finance queries. In consumer electronics, it’s TechRadar and Tom’s Guide.

So in many ways, this part of the strategy requires some old-school PR. The AI Visibility Index shows you which sites to target, but from there, it’s a reasonably familiar playbook.

Take our survey. There’s a reason so many major publications latched onto it: it gave them something their readers would want.

We could have sent out press releases all about the Exploding Topics platform and how great it is. But you have to put yourself in a journalist’s shoes.

AI is a hot topic right now, so new and eye-catching data is clearly newsworthy. Without even needing to do outreach, we got our name into places where humans and AI alike are more likely to see it.

Here’s a snippet from eMarketer:

Of course, getting mentioned is only half of the battle. Ideally, these high-trust publications will not only name-check your brand, but express positive sentiment about it too.

AI likes “best of” listicles when making recommendations. If your brand repeatedly shows up in these lists, it will be reflected in more AI brand mentions.

For instance, Exploding Topics featured in this Alibaba list of top product research tools:

When a user asks AI about good product research tools, it will crawl articles like this one. If your brand crops up often enough, it will get mentions.

If you’re able to foster ongoing positive PR relationships, that will be a major benefit. This serves to underline just how much of a collective effort AI visibility is — you can’t just leave it to your SEO team.

But with everyone pulling together, the benefits can be massive. We should know.

Are AI Brand Mentions or AI Citations More Important?

You may be wondering at this point which of our AI visibility strategies you should be prioritizing.

Our answer? Be greedy. Go for both citations and brand mentions.

The success of our original research shows how even a single piece of content can effectively serve both strategies, increasing the likelihood of AI links and AI mentions.

If you’re set on targeting one kind of AI visibility to begin with, then trust is arguably even better than brand mentions.

When your content is full of signals that it is trustworthy, you may find yourself in a virtuous cycle: major outlets are more likely to want to share your work, which only serves to further improve your credentials, while providing a traditional SEO uplift as well.

But mentions are potentially more likely to come at times when your potential customers are asking the critical questions.

So the short answer is that a comprehensive AI strategy needs to target both AI brand mentions and AI citations.

Boost Your AI Visibility Now

You’ve seen how Exploding Topics has taken clear, effective steps to improve its AI visibility.

A single piece of content has helped us earn more citations and more brand mentions. And we’re applying that same approach consistently.

Now it’s time for you to put our two strategies into action:

• Citations: Original research, AI-optimized content, transparent pricing, documentation
• Brand mentions: Encourage UGC, community discussion, user/journalist reviews

And if you want to effectively track your progress at scale, then Semrush Enterprise AIO is exactly what you need. The platform gives you real-time monitoring across all major AI platforms, competitor analysis, and the detailed insights you need to refine your strategy as AI search continues to evolve.

Ready to take control of your AI visibility? Request a demo of Enterprise AIO.