🔥 SEOs Diners Club #195: Forget Clicks, Focus on Being Selected

Greetings,

With the rise of artificial intelligence, the digital marketing world, especially SEO, is trying to find its way through a fog of uncertainty. Fundamental questions like "Is SEO dead?" and "What does chatting with Google actually mean?" occupy even the most experienced strategists' minds. Amidst this noise, some stand out by offering a clear vision of the future.

At the forefront is Dan Petrovic, director of Dejan Marketing, researcher, and one of the industry's most respected voices. Having deeply examined the relationship between artificial intelligence and digital visibility for years, Petrovic brings surprising clarity to this complex subject. This week, I analyzed the five most enlightening and paradigm-shifting insights Petrovic shared in this video, inviting you to explore the strategic dynamics of this new era of SEO.

🎯 Forget Clicks, Focus on Being Selected: SEO's New Rules

1. No Panic: Instead of Saying SEO is Dead, Accept That It's Simply Evolving

Dan Petrovic takes a clear stance against the industry panic: "SEO + AI = Still SEO." This seemingly contradictory formula actually illuminates a fundamental truth. The core principles of search engines—crawling, indexing, and ranking—remain vital for AI-powered search systems.

The most critical point is that organic search serves as a "memory layer" for AI models. AI assistants depend on the web's rich information pool—organic search results—to generate current and accurate answers to our questions. Petrovic's Swiss Army knife analogy perfectly summarizes this: AI isn't taking the SEO professional's job; it's just expanding the playing field and enriching the toolbox.

SEO is now doing more, and there's no need to rename our profession.

2. Forget Click-Through Rate, Focus on "Selection Rate"

In traditional SEO, one of the most sacred metrics was Click-Through Rate (CTR), measuring how attractive a search result was. However, Petrovic argues that the game's rules have fundamentally changed, and we now need to focus on a new concept he calls "Selection Rate Optimisation."

What does this mean?

An AI assistant takes your complex question like "best family-friendly vacation routes for the weekend" and breaks it down behind the scenes into dozens of sub-queries ("fan-out queries") like "hotels with activities for children," "vacation spots within 3 hours by car," "well-reviewed family restaurants." Your content needs to be "selected" from the pool of results for these queries.

This means content marketers must now focus not just on storefront elements like titles and meta descriptions, but also on the structural integrity and argumentative strength of content. When the model "reads" your content, it must be convinced.

3. The Idea That You're "Chatting" With AI is Just an Illusion

It might be surprising to learn that the interactions many of us have with tools like ChatGPT or Gemini aren't actually real dialogues. Petrovic reveals a critical detail about how this process works: With each new message you send, the entire conversation history is actually sent all at once to a brand new instance of the model, and the model generates a response based on this holistic information. There's no continuous "memory" or "conversation"; each message is a single-use interaction.

According to Petrovic, the chat interface is "a crutch that helps people construct the right prompt." This insight offers a powerful strategic hack for increasing efficiency: Petrovic notes that when solving a problem and finding a good intermediate solution, he goes back to the beginning of the conversation, updates the relevant prompt with this new solution, and continues from there. This way, he cleans unnecessary and incorrect steps from the context, allowing the model to progress more healthily.

When there's no conversation, the conversation is an illusion.

4. Future AI Will Be Smart But Not "Knowledgeable"

Petrovic emphasizes a strategic shift in the design philosophy of next-generation AI models. Future models will be equipped with less "world knowledge" but will focus more on "tool use" capabilities.

The fundamental reason for this approach is efficiency and cost. Continuously retraining a massive model with current world knowledge means enormous energy and time waste. Instead, developing more agile models that are smart but can consult external sources—search engines—when they need current information is much more practical. This situation, far from diminishing the importance of organic search and quality content, actually reinforces it. Because websites will continue to be the "memory" and information source for this smart but "unknowledgeable" AI.

We should think of AI's future not as "knowledgeable" but rather as "agentic," tool-using, and intelligent.

5. Websites and Apps Will Become "Optional"

Petrovic's vision for the next five years is called the "agentic era." In this vision, our personal AI agents will act on our behalf. For example, when it needs to book an appointment or purchase a product, our agent will communicate directly with the service provider's agent to complete the task invisibly.

In this scenario, users' need to visit a website or open a mobile app will significantly decrease; these interfaces will become an "option" rather than a necessity. This will be the ultimate playing field for the "Selection Rate Optimisation" we mentioned above. You'll no longer optimize for a human's click, but for an AI agent shopping on your behalf to choose your brand. That's where the real challenge and opportunity lies. Visiting a website to make a purchase will become more of an option than a necessity, unlike now.

Here is the presentation I prepared for you to understand the main concepts:

📊 Algorithm & Search Engine Updates

Supporting Dan Petrovic's vision, the search world was quite active this week. Signals that Google is preparing to go full AI mode are getting stronger. This transition promises a seismic shift in how brands capture visibility, generate traffic, and build trust. Simply focusing on traditional SEO metrics will no longer be enough; brand visibility and relevance will become more important than ever.

Another reflection of this change is AI Overviews. Research shows organic click-through rates have dropped by up to 61% in queries containing AI Overview. This reminds us once again how vital it is for our content not just to be ranked, but also to be "selected" and summarized by AI.

A great resource for those wanting to explore these topics more deeply: I watched the webinar conducted by industry leaders Lily Ray and Mark Williams-Cook, and I can say that very valuable insights were shared about the biggest SEO changes in 2025 and preparing your search strategy for 2026. I definitely recommend watching it.

🚀 From the World of AI

There are also exciting developments on the AI front. OpenAI, taking user feedback into account, released GPT-5.1, which has a warmer and more conversational tone. This update offers two different models: "Instant" which is faster for simple tasks, and "Thinking" which is more persistent for complex tasks. This is a sign that our interactions with AI will become more personalized and efficient.

However, the rise of AI requires us to rethink our traffic sources. According to comprehensive research by Conductor, AI referral traffic currently accounts for only 1% of all web traffic, and the majority of this traffic comes from ChatGPT. The report's most striking finding is this: Having good rankings on Google doesn't guarantee visibility on AI platforms like ChatGPT. AI prioritizes authoritative sources like Mayo Clinic and NerdWallet, especially in YMYL (Your Money or Your Life) categories. This demonstrates how critical it is to build brand authority and trustworthiness.

Google is also rapidly adapting to this "agentic era." The new AI shopping features announced for the holiday season offer futuristic capabilities like AI calling stores on our behalf to check inventory (Agentic Calling) or purchasing products when it finds prices within our set budget (Agentic Checkout). This could be the first steps of the "optional websites" era Dan Petrovic mentioned.

Finally, a small but fun piece of news: OpenAI announced it fixed ChatGPT's em-dash problem, which had become frequent in its texts and turned into an "AI signal." We can now tell the chatbot not to use this punctuation with special instructions.

💡 Practical SEO Tips

Actionable takeaways we can draw from Dan Petrovic's analyses and this week's developments:

Optimize Your Content for Query Clusters: Instead of focusing on a single keyword, target a series of sub-queries (fan-out queries) that can answer a user's complex question. Your content's structural integrity and argumentative strength will be your most important weapon for being "selected" by AI. The model must be convinced when it "reads" your content.

Strengthen Your Brand Authority: AI prioritizes trustworthy sources. Be present on platforms recognized as authorities in your industry, manage your brand presence with consistent and accurate information. Don't ignore elements like reviews, social signals, and local listings. Remember, ranking well on Google is no longer enough; you need to be recognized as a trustworthy source on AI platforms too.

Use Structured Data: Provide search engines with clear and structured information about your products, services, and content. Leverage Schema markup to display features like shipping and return policies that Google newly announced in search results. This makes your content easier to understand for both traditional search and AI-powered systems.

Prepare for the "Agentic Era": Prepare for a future where AI agents will make decisions on your behalf. Think about what you need to do to make your brand "selectable" not just by humans, but by AI agents too. This includes many factors from pricing transparency to customer reviews, API integrations to service quality.

Remember, succeeding in this new era requires understanding the language not just of search engines, but also of the AI models behind them. If you need a more in-depth strategy on this, Stradiji would be happy to help.

🎬 Closing

Dan Petrovic's five analyses prove that AI hasn't killed SEO; rather, it has transformed it into a more strategic, more technical, and far more layered discipline. Instead of panicking, we must understand this evolution and adapt to the new rules of the game. While fundamental SEO principles remain valid, new metrics like "selection rate" and new paradigms like the "agentic era" require us to reshape our strategies.

In the midst of this transformation, the most critical question we must ask ourselves is: As our websites and apps become "options," how will we make our brands indispensable at the invisible negotiation table of AI agents?

I hope this week's notes guide you in this new playing field.

Wishing you a great week,

Mert Erkal

Want to Support This Newsletter?

I love sharing the latest developments and strategies from the SEO world with you. If you find my content useful, you can support me by buying a coffee. ☕ Click 'Buy Me a Coffee' to contribute to knowledge sharing. Let's achieve more together!

© 2022-2025 SEOs Diners Club by Mert Erkal