The Search Game Has Changed. Most Agencies Haven't. (1 of 2)
Updated: Sep 1
Understanding the AI Search Shift and What It Means for Your Marketing Strategy
Let me tell you about a conversation I had with an SEO expert at a Stockholm agency.
At my last employer, inbound was key for lead generation. We had recently moved our marketing to this agency; it was a local, full-service shop that offered more than the modest but effective agency we'd been working with in Poland. The pitch was compelling: one-stop-shop, integrated services, senior expertise. So we made the switch.
At some point I asked their SEO lead a straightforward question: how are you thinking about AI search? What happens when someone asks ChatGPT or Perplexity a question that our product should be able to answer: will we show up?
The answer was not convincing.
It wasn't evasive, exactly. It was more that the question didn't quite land. The conversation drifted back to keyword rankings, domain authority, and backlink profiles. All legitimate things. But not the thing I was asking about.
That moment stuck with me. And it's part of what led me to build something myself.
But before I get to that, let's talk about why the question matters so much right now.
The Search Landscape Is Shifting Under Our Feet
For roughly two decades, the rules of online visibility were relatively stable. You built a website, optimized your pages for keywords, earned backlinks from credible sources, and hoped Google rewarded you with a spot on page one. Search Engine Optimization (SEO) was the discipline that governed all of it.
That world still exists. But it is shrinking fast.
Today, a growing share of search behavior is being mediated not by a list of blue links, but by AI-generated answers. When someone asks ChatGPT for a recommendation, queries Perplexity for research, or reads a Google AI Overview at the top of their search results, they are not clicking through ten websites and forming their own opinion. They are receiving a synthesized response; assembled by an AI system drawing on sources it deemed credible, authoritative, and well-structured.
AI-powered search is no longer a future trend — it is the present reality. In 2026, over 30% of all search queries globally are processed through AI-powered platforms. Users are not just typing keywords into Google anymore. They are asking ChatGPT for recommendations, querying Perplexity for research, and getting AI-generated summaries in Google Search results. --AdWithChatGPT/ State of AI Search in 2026: ChatGPT vs Perplexity vs Google.
What Agencies Are Working With. And Where It Stops
To be fair to agencies, the tools they use are genuinely good at what they were built for. Most serious SEO practitioners work with some combination of the following:
SEMrush: keyword tracking, site audits, competitor analysis, backlink monitoring
Ahrefs: strong on backlink intelligence and content gap analysis
Moz: domain authority scoring, on-page optimization recommendations
Google Search Console: the ground truth on how Google actually sees your site
Screaming Frog: technical SEO crawling, broken links, redirect chains
Surfer SEO: content optimization against top-ranking pages
These tools do their jobs well. They track traditional ranking signals: keywords, backlinks, domain authority. They surface technical health issues. They benchmark you against competitors on known search terms.
But they were all built for a world where Google returns ten blue links and you optimize to rank in positions one through three.
That world still exists. But it is no longer the whole picture.
The AI Search Black Hole
Here is the honest conversation most agencies are not yet having with their clients. None of the standard tools measure:
AI Overview inclusion. Google's AI Overviews synthesize an answer at the top of the page. Your content might rank second organically but never get cited in the AI Overview. SEMrush doesn't tell you why. Neither does Ahrefs.
Perplexity and ChatGPT citations. These systems pull from the web but use their own relevance signals: answer clarity, source authority, content structure. There is no "Perplexity ranking report" you can pull from any standard tool.
Answer-readiness scoring. Is your content structured in a way that a reasoning engine can extract a clean, citable answer? Current tools do not measure this. It is not about keywords — it is about how clearly you have stated what you know.
Entity and authority signals for large language models. LLMs weight content differently than PageRank does. Clarity of expertise, consistency of voice, and structured factual claims matter in ways that traditional SEO metrics do not capture.
Zero-click erosion. If an AI Overview answers the question completely, the user never clicks. Your traffic drops even if your ranking holds. Current tools show you the ranking. They do not show you the zero-click erosion.
This is not a criticism of agencies. It is just where the industry is. The signals are not published. The tools do not measure it yet. The best practices are still being written in real time. Most agency playbooks were built for 2019 search and updated cosmetically for 2023.
The honest answer from any agency right now should be: "We are watching this closely and testing." Most are not even saying that.
Enter GEO: The Discipline That's Trying to Fill the Gap
A new field has emerged to describe the practice of optimizing content for AI systems: Generative Engine Optimization, or GEO. It isn't a replacement for SEO, but rather, an extension of it. The foundations (technical health, quality content, structured data, authoritative sources) still matter enormously. What changes is the optimization target.
Online search is no longer about links—it is about answers. AI-generated results now surface synthesized responses directly within search interfaces, dramatically reducing click-through to traditional websites. --KPMG/ GEO mandate: optimizing content for AI-generated results
In traditional SEO, you are optimizing for a ranking algorithm that returns a list of links. In GEO, you are optimizing for a reasoning system that synthesizes an answer. The question shifts from "does Google rank this page?" to "would an AI cite this page as a trustworthy source?"
That is a meaningfully different question. And it requires a meaningfully different approach to content.
What Integrated Strategy Actually Looks Like
What does integration actually look like in practice? I read through Semrush's 2026 AI Visibility Index and pulled out some tips and core principles based on what the research and emerging GEO frameworks suggest:
Build traditional SEO first. Ranking in the top 10 for your target keywords is still the foundation for Google AI Overviews, Bing Copilot, and ChatGPT's web search citations. You cannot skip this step.
Implement comprehensive structured data. FAQ Page, How-To Article, and articles about the product, organization, and target personas help every AI platform understand your content.
Format for extraction. Use answer-first paragraphs, numbered lists, comparison tables, and self-contained quotable passages. AI systems are looking for content they can cleanly extract and cite.
Publish fresh content regularly. Perplexity and ChatGPT both favor recently published content. Aim for at least two to four new pieces per month.
Build entity authority. Consistent brand mentions across authoritative sources, named authors with credentials, and clear topical expertise help every platform correctly identify and cite your brand.
None of these are exotic or technically complex. They are extensions of good web practice. The challenge is that most businesses have never audited their websites through this lens because until recently, no accessible tool existed to do it.
The Road Ahead (And Why I Stopped Waiting for Someone Else to Solve It)
The field of GEO is still young. There is no universal standard for what makes a website "AI-visible." The major platforms update their approaches regularly. The research community is still working to establish best practices.
But the direction of travel is clear. AI-mediated search is not a future trend; it is the present reality. And the businesses that start building for it now, rather than waiting for the playbook to be finalized, will have a meaningful head start.
After that conversation with the Stockholm agency, I decided I wasn't going to wait for an agency to figure this out on my behalf. I wasn't going to wait for SEMrush to add an AI visibility tab. I was going to build something.
In Part 2 of this series, "I Built an AI Search Visibility Agent — Here's What Happened" , I'll walk you through exactly what I built: a Search Visibility Agent that crawls a website, scores its technical health, evaluates its AI readiness, and generates actionable reports, including content briefs and page rewrites optimized for AI citation. I'll also be honest about what went wrong along the way. Because getting an AI system to run is very different from getting it to run correctly.
This is Part 1 of a two-part series.
--
Samara H. Johansson is a senior B2B marketing consultant specializing in developing and then activating global brand positioning, messaging frameworks, and AI-augmented marketing strategy to generate leads. She works with companies navigating growth, repositioning, and international market expansion. Learn more at SamaraGlobal.com and subscribe to The AI-Augmented Marketer on LinkedIn.
Comments