GEO Starter Guide: How To Get Your Brand Mentioned In Answers From ChatGPT, Gemini & Other AI Models

I originally published this guide in Hebrew last week on my blog, AskPavel. After it gained a lot of traction, I decided to share an English version as well.

For decades, when we talked about “search,” we really meant one thing: Google. A world where asking a question led to a list of clickable links.

But that world is changing. In the age of AI, search looks different. Instead of showing you a set of results and letting you choose a website, generative AI models like ChatGPT, Gemini, Claude, and Perplexity deliver a single, direct answer.

From a user’s perspective, this is a game-changer. After all, when we search for something, we’re not looking for links—we’re looking for answers. Cutting one step out of the funnel makes the process faster and more convenient.

But does this shift mean traditional SEO is dying? Not even close (keep reading).

The goal of this guide is to help CMOs, marketers, website owners, and anyone trying to grow their brand’s digital presence understand the new rules of the game and fold them into their digital marketing strategy. If your audience is already using AI models to search, then your brand needs to be part of that conversation.

With all the hype and disinformation (often fueled by AI-driven founders and investors), it’s more important than ever to cut through the noise and look at what’s really happening—based on real data and trusted sources.

And just to be clear: Yes, we offer traditional SEO services. But we’ve also been helping clients win visibility in GEO. We have no reason to defend SEO out of loyalty—we are writing this for two simple reasons:

1. To lay out the real numbers, because the hype around “the death of search” is wildly misleading.
2. To help you navigate the new GEO landscape and get your brand mentioned in AI-generated answers.

Introduction: AI Search Is Evolving — But Don’t Count Google Out Just Yet

Let’s put things in perspective:

Despite all the buzz around “the death of search” or “SEO is dead,” the numbers tell a very different story. As of March 2025, search volume on Google is 373 times larger than on ChatGPT. Google handles over 5 trillion searches per year, growing at a rate of 21.64%, while ChatGPT accounts for just 0.25% of global search volume. For comparison, that’s even less than Bing (4.10%), Yahoo (1.35%), or DuckDuckGo (0.73%)!

Saying “Google search is dead” today would be like declaring it obsolete when Bing first launched 🙂

I know I’m ruining the hype, but I’ve always believed in deep research and real data over trendy narratives and investor-fueled speculation.

Yes, generative AI is advancing quickly. But if we zoom out and look at actual usage numbers — AI search is still a small player in the search ecosystem (at least for now).

So no, SEO isn’t dying. It’s just time to expand the toolbox. The real opportunity is to combine the best of both worlds: traditional SEO and what we now call GEO — Generative Engine Optimization.

This guide focuses primarily on ChatGPT, which currently dominates the AI search space with 81.39% market share, far ahead of other models like Gemini, Claude, or Perplexity. According to Statcounter, here’s how the AI search market is currently divided (as of March 2025):

Understanding the New Rules of the Game

In traditional search engine optimization (SEO), the goal has always been clear: drive more relevant organic traffic from search engines to your website — and keep that traffic growing over time.

To make it happen, you’d typically aim to:

• Rank for as many relevant keywords in your niche as possible (e.g., by creating keyword-focused pages).
• Climb higher in the rankings for those keywords (via classic on-page and off-page SEO tactics).
• Get users to click your link over the others (with compelling meta titles, schema, featured snippets, etc.).

But in the world of GEO — the rules have changed.

There are no search results to choose from. There’s just a single AI-generated answer, sometimes with references.

There’s no first page, no second page… no pages at all.

And in this game, the winner takes it all — if your brand is mentioned in the answer, you get 100% of the user’s attention.

So what does success look like in this new AI-powered search world?

You now have two clear goals:

1. Get your brand mentioned in as many relevant AI answers as possible (AI Mentions).
2. Get your website cited as a trustworthy source in those answers (that’s how you drive *actual traffic*).

Thinking Like a Prompt Engineer

In SEO, we used to obsess over what keywords people were typing into Google.

In AI search, nobody is searching for vague terms like “car insurance.” Instead, people write detailed prompts, like:

“What’s the best way to insure a used car after an accident?”

This shift means our content needs to evolve. Your website should focus on delivering:

• Complete and direct answers (no fluff).
• Short, skimmable paragraphs.
• Clear definitions.
• Professional tone, without being condescending.
• Tangible examples.

Traditional SEO vs. GEO

What Does Optimization Look Like for Generative AI?

A few months ago, Seer Interactive published a study showing a potential correlation between AI mentions and traditional SEO ranking factors.

Since AI search is still a relatively new field, there isn’t a ton of large-scale data to draw firm conclusions — but early patterns are starting to emerge. In our own testing, we’ve seen some alignment between the two worlds, suggesting that your SEO efforts can directly support your visibility in AI-generated answers.

As we mentioned earlier, there are two primary types of value we’re aiming for when it comes to AI search:

Earning Brand Mentions in AI Answers

In this case, the AI returns an answer that mentions your brand, product, or service — without a link. The user may then decide to search your brand name in Google or visit your site directly.

Example query:

“What’s the leading commercial real estate company in new york?”

If your client’s name appears in that list, that’s a brand exposure win. Even without a direct link, the mention itself creates brand awareness and can lead to increased branded searches or direct visits.

Driving Traffic from AI Answers

This refers to situations where an AI model provides an answer and includes clickable links to external sources. If your page is cited as one of those sources, there’s a good chance the user will click through to your site.

Example query:

“What are the benefits of a CRM system for small businesses?”

If your article ranks well in this answer, it could be included as a reference, earning you direct traffic from the AI interface.

How to Increase Brand Visibility in AI Models Answers (AI Mentions)

Now that we understand the shift, your goal in this new search landscape is clear: get your brand name mentioned inside as many relevant AI-generated answers as possible.

But how do you actually make that happen?

What kinds of sources feed into AI models like ChatGPT? How can you show up in the right places? And what tactics increase the chances of being mentioned in those answers?

That’s where your AI Mentions strategy comes in.

Show Up in Sources That Feed the Training Data

One of the biggest misconceptions is that LLMs like GPT simply crawl the entire internet. They don’t. In reality, their training data is made up of specific and often curated sources.

So where should you focus?

• Major publishers that have official partnerships with OpenAI, including Associated Press, News Corp (WSJ, NY Post), The Atlantic, Vox Media, and Condé Nast (Vogue, The New Yorker).
• Community-driven platforms that consistently show up in model training sets, such as Reddit, Quora, and Wikipedia.

The more your brand is present in these ecosystems, the greater the odds it will be “learned” and recognized as a relevant entity by AI models.

Smart PR Activiy (Even Without a Link)

In classic SEO, links were (almost) everything. In the AI world, a brand mention — even without a link — can matter just as much.

Your goal is to ensure your brand shows up positively and contextually in product roundups, reviews, expert articles, and comparisons, even if those don’t link to your site directly.

Here’s how to make that happen:

• Pitch journalists who are working on upcoming pieces in your niche.
• Offer guest content to relevant publications.
• Proactively manage PR in AI-friendly spaces.

No PR budget? No problem.

You can still appear in AI answers without a full-blown PR campaign. With surgical content optimization, even small sites can break through — as Rand Fishkin’s example clearly shows.

Build a Strong “About Us” Page

When ChatGPT or Gemini gets asked about your brand, one of the first places it looks is your About page. If that page is empty or vague, the model might pull info from boring legal pages—or worse, irrelevant sources.

Here’s what a solid About page should include:

• What your company actually does.
• Areas of expertise.
• A bit of history or backstory.
• Key stats and facts.
• Notable achievements.

Don’t Block the Bots

Make sure your robots.txt file allows crawling by AI tools like OpenAI, Anthropic, and others. Here’s a list of OpenAI’s crawlers.

When ChatGPT launched, many site owners rushed to block AI scrapers out of fear their content would be used without permission.

But here’s the thing: if the bots can’t see your content, don’t expect to show up in their answers.

Target Long-Tail Queries

Instead of optimizing for broad terms like car insurance, think like your audience and go more specific—something like “what’s the cheapest way to insure a new car after an accident?”.

That kind of detailed phrasing is how real users talk to AI. Writing content that matches long-tail queries helps build strong semantic signals around your niche. Tools like SparkToro and BuzzSumo can help you understand where your audience hangs out and what they’re curious about—so you can create content that hits the mark.

How to Drive Traffic from AI Answers (Citations)

So you’ve made it into the dataset — now what? How do you make sure AI models actually send traffic your way?

Beyond just being mentioned by name, one of the key opportunities in AI-powered search is to generate actual traffic through citations, links, and active references.

In this section, we’ll explore how your content structure influences whether your site is selected as a source, how to boost semantic alignment to increase selection odds, and how to monitor and improve your brand’s visibility inside AI-generated answers.

Add Credible Sources and Citations Inside Your Content

In 2024, researchers from the Indian Institute of Technology Delhi and Princeton University published a major study at KDD’24 in Barcelona. They created a large-scale testing framework called GEO-bench, containing 10,000 queries and hundreds of real-world content sources, to evaluate what kinds of content are most likely to appear in AI-generated answers.

The takeaway? Content that includes reputable citations (“According to [source]”), up-to-date stats, and clearly sourced facts was ~40% more likely to be cited as a reference by AI. The quality of the source matters more than the quantity — one credible citation outperforms five vague ones.

Create Content with E-E-A-T Principles in Mind

This SEO-originated framework also applies to AI visibility. E-E-A-T stands for:

• Expertise – Who wrote it? Are they a subject-matter expert?
• Experience – Is there real-world proof they’ve done what they write about?
• Authoritativeness – Is the author or site known and trusted in the space?
• Trustworthiness – Can the user (or model) trust this content?

As with SEO, make sure each article includes an author byline and a short bio with credentials, relevant experience, and domain-specific expertise. Google’s own documentation highlights this as a ranking factor — and generative engines are following suit.

Track and Update Your Content Based on AI Visibility

One of the most important (and overlooked) steps is actively tracking which of your pages are getting referenced by AI engines — and then updating them accordingly.

Here’s how to do it:

1. Identify referenced pages: Try running prompts related to your niche on ChatGPT, Gemini, Claude, or Perplexity. Note which of your URLs appear in citations or source listings.
2. Measure actual traffic: Once you’ve spotted these pages, check if they’re actually bringing in users. In Google Analytics, go to Traffic Acquisition and filter by sources like:
   1. perplexity.ai / referral
   2. you.com / referral
   3. chat.openai.com / referral (only if browsing is enabled)
3. Actively update your content: when we identify relevant pages showing up in AI search results, we’ll want to:
   1. Update stats, dates, and references.
   2. Add clearer semantic context for your core topics.
   3. Rework key paragraphs to answer specific user questions more directly.
4. Go deep, not just wide: AI engines prefer depth and clarity. Expand pages with guides, examples, or studies — this boosts the chance of repeat selection.
5. Re-check after model updates: Since models like ChatGPT are updated frequently, repeat these tests every few months. A page cited in May might disappear by August if it’s outdated or replaced by fresher content.

Content AI Can Easily Extract Answers From

LLMs look for structured, extractable information. The easier you make it to scan and summarize, the more likely it is to show up in AI answers.

• Use lists instead of long paragraphs.
• Keep paragraphs short.
• Provide direct answers — no fluff.
• Use H2 and H3 headers to organize your page.
• Stick to simple language. (If you wouldn’t say it in a real conversation, don’t write it that way.)
• Add FAQ schema wherever possible.

Take a look at ChatGPT’s answer to the question “What basic fishing gear do I need as a total beginner?” and notice how it breaks the answer into subtopics, sometimes linking to separate sources for each one due to the lack of a single comprehensive reference.

When multiple links are used, each receives less traffic. That’s why the goal should be to create individual content pieces that thoroughly address each subtopic, in order to maximize the overall traffic potential.

Think of it like dominating multiple positions on page one of Google — but inside an AI answer.

Use Structured Data to Boost Visibility

When LLMs build answers, they’re not just parsing raw text. They’re looking for clear structure, and that includes Structured Data (like JSON-LD).

Use relevant schemas such as:

• Article
• FAQ
• Product
• Organization
• Review

Make sure your markup is clean, updated, and consistent. Use Google’s Rich Results Test to verify your setup.

Build a Strong Semantic Connection

LLMs don’t just associate your brand once — they “learn” it over time through semantic association. This happens when your brand consistently shows up in the right context.

Here’s how to build that association:

1. Stay consistent around specific topics. Don’t just mention “insurance” — tie your brand directly to “car insurance claims,” “third-party liability coverage,” etc.
2. Create long-tail content around real user queries. Example: instead of “car insurance,” write about “how to reduce your car insurance premium after your first accident.”
3. Expand related context. A page about car insurance should also mention “claims adjusters,” “replacement vehicles,” and “damage appraisals.”
4. Give straight answers. AI prefers pages with clear, focused answers to specific questions.
5. Keep refreshing the connections. Update, refine, and expand your content to stay relevant to how people — and models — phrase things today.

Here’s a great example: in a recent demo, Rand Fishkin showed how Sparktoro appears first in AI answers for “audience research software” — not because of keyword stuffing, but because that exact phrase shows up right next to his company’s name across dozens of bios and articles online. That’s the power of semantic proximity.

Bonus Tip: Optimize for Bing

Did you know Microsoft owns 49% of OpenAI?

This partnership has made Bing a key player in the new search landscape. Why? Because ChatGPT Search uses Bing’s index to retrieve information and generate answers. If your site isn’t properly indexed by Bing—or worse, missing altogether—it might not appear in ChatGPT’s answers either.

So while Google still dominates traditional search traffic, 2025 is the year to start treating Bing as an essential part of your SEO strategy—especially if your goal is to show up in AI-driven results.

Here are a few recommendations from SEO expert Jessica Bowman on how to align with Bing in the ChatGPT era:

• Check your Bing indexing status: Log into Bing Webmaster Tools to see which of your pages are indexed—and which ones aren’t.
• Compare Google and Bing traffic: Look at the URLs that drive traffic from Google and compare them with Bing’s. This helps you identify high-value pages that are missing from Bing and need attention.
• Make sure critical pages are indexed: If you have access to Bing’s API, export your full list of URLs and run a bulk index check. It’s an efficient way to catch gaps fast.
• Fix minor indexing issues: Most problems are simple—like a rogue noindex tag or a blocked robots.txt path—and easy to fix once spotted.
• Bring it up in SEO reports to leadership: Show that you’re forward-thinking by highlighting Bing visibility as part of a broader global trend. It helps position you as someone preparing the organization for the AI future.

Put Your Site to the Test: Ask Gemini How AI-Ready It Is

A simple way to find out how your site appears to AI models? Just ask them. 🙂

At the moment, the most helpful insights come from Gemini. You can ask it to evaluate how likely your website is to show up in AI-generated answers. Just use the prompt below (and don’t forget to enable Deep Research before starting):

“Rate the website [your domain] from 1 (lowest) to 100 (highest) in terms of likelihood to appear in LLM-generated answers. Justify the rating and compare it to competitors.”

In just a few minutes, you’ll receive a downloadable report with a detailed competitive breakdown — including what you’re doing well and where you can improve.

Bottom Line: SEO isn’t dead, but it’s worth building a presence in AI too

This new world of search doesn’t mean you need to throw out everything you know about SEO. It just means adding a new layer — one that helps you work with the tools that are changing how people find information.

Just like we’ve learned to adapt to emerging trends as they gain momentum — it’s worth starting to explore GEO and making it part of your broader search strategy.

Those who get on board early will benefit twice: Once from classic search results, and again from AI-generated answers that are quickly gaining influence across platforms.