01. Why does ChatGPT recommend my competitors?

There's a question I hear a lot lately – sometimes from veteran SEO professionals, sometimes from marketing managers who track their numbers and notice something that doesn't add up:

"We rank well on Google, but when I ask ChatGPT about our field, the competitors get all the love. Why?"

It's not a rhetorical question. It has an answer, and part of it you won't like.

When an AI engine generates an answer, it doesn't run a real-time search and return results. It draws on what it absorbed during training, and additionally, when it has browsing ability, it picks sources that look credible and relevant to the question. Both paths arrive at the same place: the engine looks for content where it's clear what's being said, where the field of authority isn't blurry, and that someone else has already relied on.

The competitor who gets cited in ChatGPT usually didn't pull off any technical magic. They simply wrote content that directly answered the questions people ask, without trying to cram in more and more keywords. They built a reputation over time as a source others cite, link to, and mention in professional contexts.

The takeaway: This isn't about technology. It's about trust built over time. And the good news – it's something you can build.

02. What's the difference between GEO and classic SEO?

To understand the difference, you need to understand what changed in the way people get information.

Classic SEO – the world of links

Google shows a list of links. The user chooses where to click. The game is buying real estate on the results page – position 1, 2, 3.

✓ The goal: drive clicks to your site

GEO – the world of answers

Perplexity generates one complete answer. The user doesn't need to visit your site. If you weren't mentioned – you don't exist.

✓ The goal: become part of the answer itself

The key difference is this: SEO asks Google to send people to you. GEO asks the engine to cite you as an authority. That's a fundamental shift in logic. You're not trying to buy real estate on a results page; you're trying to become part of the knowledge the engine relies on when it answers questions.

Check now: does your brand show up in ChatGPT and Gemini?
A unique visibility checker – answer in 90 seconds, no signup

03. How do AI Overviews affect organic traffic?

Google added AI Overviews – detailed answers shown above the search results – to a large share of searches. The first thing many people noticed is that organic traffic is dropping for certain pages. Not for all of them, and not always, but it's happening.

What turns out to be true in the field is that the drop in traffic is not uniform:

  • Pages that provide general information that can be summarized in a few sentences – suffer the most.
  • Pages with unique information, tools, analysis, content that requires interaction – hold up better.
  • Sites mentioned within the AI Overview itself sometimes see an increase in clicks.

Data from BrightEdge: CTR on organic results has dropped in certain categories, mainly informational queries. If the engine answers you directly – why would you click? But whoever is mentioned inside the answer becomes part of the narrative, not just part of the list.

04. How to build a content strategy that works in both Google and AI?

There's a mistaken belief that you have to choose: write for Google or write for people. In practice, what works in AI search also works well in classic SEO, because both engines are looking for the same thing: content that answers real questions, clearly, from a source that looks credible.

What changes is the format and the level of precision. When an AI engine reads, it looks for the piece that directly answers the question. If it's buried inside two paragraphs of introduction, it may miss it – or prefer a source that phrased it more directly.

The practice I recommend is to break the content into questions and answers – not in an artificial FAQ style, but in a way where every part of the article answers a clear question you'd be glad to see asked.

A real-world example

Not good: "Pensions – what you need to know," with ten mixed-up topics. An AI engine struggles to cite it.

Better: "When should you start saving for a pension?" and answer it directly ← "What's the difference between a pension fund and a managers' insurance policy?" and answer that too. Each question as a standalone unit – an engine can pull exactly what fits the question that was asked.

GEO article writing checklist
Every criterion AI engines check before deciding to cite

05. What content makes AI cite you?

There's no single rule that works in every case. But there are three traits that keep coming up:

  • 1

    Specificity – facts, numbers, reasoned arguments

    "Digital marketing matters for every business" – an engine won't cite that. "Businesses that invest more than 10% of their budget in paid search report 40% higher ROI on average compared to offline channels" – now that's something an engine can work with.

  • 2

    Topical consistency – authority in a specific field

    Sites that write about everything get fewer citations. A site where all the content deals with project management, from every possible angle, will be seen as an authority on the topic.

  • 3

    Citations pointing to you – others mention you

    When other sites link to you, mention you, cite you – the engine learns that something here deserves attention. The engine doesn't only read your site. It reads everything written about you.

06. The four actions that increase your chances of appearing

There's no "button" that adds you to the answers. What there is, is four actions that significantly increase the chance you'll be included:

  • 1

    Clear content structure – Schema, headings, question-and-answer

    AI engines rely on schema markup, hierarchical headings, and passages of text that read like answers. A well-organized FAQ page with clear questions and answers helps a lot. Google itself has stated that structured content is preferred in AI Overviews.

  • 2

    External presence – LinkedIn, podcasts, mentions

    You wrote on LinkedIn, you were invited to a podcast, you were mentioned in a news article, your site earned links from relevant sources – all of these build the presence the engine relies on. It's neither quick nor cheap. But there are no shortcuts.

  • 3

    Freshness – recently written content

    Engines like Perplexity that use real-time search tend to prefer content that was written or updated recently. An article you haven't touched in three years – even if it was excellent – loses out to newer content.

  • 4

    Mentioning your brand in the right contexts

    When someone writes "according to the people at [your company name], the right approach to this is…" – that's a mention the engine can learn from. Building a reputation among colleagues, partners, and customers who write reviews – that's the foundation for GEO.

Full GEO audit – 75+ criteria
Check how ready your site is to appear in AI search, with a detailed score for each category

07. How to create content that delivers results over time?

"Evergreen content" is an old phrase, but the idea behind it matters more than ever. Content that answers questions people ask every year, in any season, and doesn't depend on today's news – keeps working.

The question worth asking before writing an article: "Will this question still be relevant three years from now?" If yes – it's worth the investment. If not – keep it short and move on.

A real case

An accountant's site that wrote seasonal articles about "the deadline before the holidays" got spiky traffic. When it added articles like "What's the difference between a licensed dealer and a limited company?" and "When are you required to file an annual report?" – questions people ask anew every year – the traffic started to compound. A year later, those articles brought in more traffic than all the seasonal articles combined.

08. How to measure how often your brand is mentioned in AI?

Automatically and precisely – there's still no perfect solution. There's no API that tells you "how many times you were mentioned in ChatGPT this week." But there are a few practical approaches:

  • The manual way (the most reliable): Pick 10–20 questions your customers ask and run them through ChatGPT, Gemini and Perplexity once a month. Note who's mentioned, who isn't, and the style of the answers.
  • Emerging tools: Brandwatch, Mention and Semrush are adding AI monitoring features. A dedicated tool like Profound is trying to solve exactly this problem.
  • Monthly benchmark: Build a "benchmark questionnaire" with the relevant questions, check it once a month, and see whether the changes you made to your content are starting to make a difference.
AI engine visibility checker – free, in 90 seconds
See how your brand shows up in ChatGPT and Gemini right now

09. A few honest truths about the process

GEO isn't an add-on to SEO. It's not "one more thing to check off." It requires a shift in content logic – from "how many keywords are we covering" to "how much are we regarded as an authority worth relying on." It's a change that takes time, and isn't always easy to measure on a dashboard.

In practice, much of what GEO requires is exactly what we always said good SEO was: credible writing, specific content, building authority, managing reputation. What changes is that the result doesn't always arrive as direct traffic to your site – it arrives as presence in the answers, as a mention, as trust built before the potential customer ever reached you.

Those who adapt early will enjoy an advantage that's hard to replicate later. Because authority can't be faked overnight – not in Google's eyes, and not in the eyes of AI engines.

The book: "Get Cited, Not Just Ranked"
The first practical guide to optimizing for answer engines. 15 chapters + a ready-to-use toolkit. First chapter free.

Frequently asked questions

Why does ChatGPT recommend my competitors instead of me?

When an AI engine generates an answer, it draws on its training and a selection of sources that appear credible. The competitor who gets cited wrote content that directly answers questions and built a reputation as a source others mention. It's a matter of accumulated trust, not technical magic.

What's the difference between GEO and classic SEO?

SEO asks Google to send people to you through ranking in the results. GEO asks the engine to cite you as an authority inside the answer itself. A fundamental shift in logic: not buying real estate on a results page, but becoming part of the knowledge the engine relies on.

Is a drop in clicks from AI Overviews a problem?

It depends. Pages with general information that can be summarized – yes, they suffer. Pages with unique information, tools and analysis – hold up well. And whoever is mentioned within the AI Overview itself sometimes sees a rise in exposure. The solution: become a source that's mentioned, not just linked.

How long does it take to see results from GEO?

Usually 2–6 months, depending on how competitive the field is, the volume of existing content, and the strength of your external presence. GEO isn't magic – AI engines learn from historical data and verified sources over time.

Does GEO include classic SEO too?

Yes. A professional GEO service includes full SEO (technical SEO, content, link building) on top of optimization specifically for AI search. You essentially get two agencies for the price of one.

How do you measure success in GEO?

Key metrics: GEO Authority Score, the number of direct citations in AI engines, the change in branded search (searches for your brand name), and the quality of incoming leads. Commercial tools like Semrush and Profound are starting to add dedicated measurement features.