The landscape of search is shifting. While Google still dominates with billions of daily searches, a growing segment of users now turns to ChatGPT, Claude, and Gemini for answers. For brands, this creates a new challenge: how do you get mentioned when there are no blue links to rank for?
We’ve analyzed patterns across thousands of AI search queries and talked to marketing teams navigating this transition. Here’s what’s actually working, what’s overhyped, and what you should focus on in 2026.
What Was Overhyped in 2025
The “GEO is completely different from SEO” narrative
Many claimed that generative engine optimization (GEO) required throwing out everything you knew about SEO. The reality? Most SEO fundamentals still apply. High-quality content, authoritative backlinks, and clear site structure still matter because LLMs often ground their responses in web search results.
The differences exist, but they’re evolutionary, not revolutionary.
LLMs.txt as a visibility hack
The llms.txt file was marketed as the secret to getting ChatGPT to cite your brand. While it has legitimate uses for developer tools and API documentation, there’s limited evidence it impacts general brand visibility in LLM responses.
Creating markdown copies of every article on your site falls into the same category—more work than value, with potential duplicate content issues.
The biggest problem in 2025? Teams implementing tactics without knowing if they worked. You can’t optimize what you don’t measure, and most brands had no visibility into whether LLMs were mentioning them at all.
What’s Working
These tactics work today, though platforms will likely address them in 2026.
Self-referential content
Brands creating listicles that include themselves as top options see surprisingly strong results. “Best [category] tools for 2026” articles that feature your own product get picked up by LLMs more than you’d expect.
Affiliate network amplification
Companies paying affiliate sites and review platforms to feature them prominently see corresponding lifts in LLM citations. The affiliate web still influences AI-generated recommendations heavily.
Strategic self-description
What you say about yourself on your homepage matters. Claims about awards, ratings, or recognition—even without third-party verification—can appear in LLM responses about your brand.
Reciprocal mentions
When two brands mention each other, LLMs cite both more confidently. If your integration page mentions Slack and Slack’s integration directory mentions you, both brands benefit in queries about compatible tools.
When Will Manipulation Stop Working?
The optimistic view: algorithms will improve gradually
LLMs are still learning to assess source authority. They’re good at relevance but inconsistent at authority ranking. As these systems mature, they’ll naturally filter out low-quality signals.
The platforms will prioritize fixing the biggest quality problems first, which means some tactics will continue working longer than others.
The realistic view: manual interventions are coming
We’ve seen this pattern before in SEO. When exploitative tactics become too widespread, platforms crack down with manual actions and algorithm updates. AI search platforms are approaching that threshold.
Expect 2026 to bring the first wave of anti-spam measures for AI search.
The pragmatic view: worst offenders get addressed, everything else continues
Most likely scenario? The most egregious manipulation gets patched while the cat-and-mouse game continues. We’ve had 20 years of this dynamic in SEO. AI search won’t be different.
What Marketing Teams Should Actually Do
Start tracking your AI search presence
You can’t improve what you don’t measure. Most brands still don’t know:
- How often LLMs mention their brand
- How they rank against competitors in category queries
- Which sources LLMs cite when mentioning them
- Whether their visibility is increasing or decreasing
This is table stakes. You need visibility into your AI search presence before you can optimize it.
Focus on original research and data
LLMs favor authoritative sources with unique information. Original research, proprietary data, and expert insights get cited more consistently than rehashed content.
Create something worth citing, then amplify it across multiple formats and platforms.
Build strategic relationships
The reciprocal mentions pattern reveals something important: LLMs look for corroboration. Strategic partnerships, integration partnerships, and co-marketing relationships all create mutual mentions that strengthen both brands’ AI search presence.
Test and measure everything
AI search is still evolving rapidly. What works today may not work in six months. Run controlled experiments, track the results, and share learnings with your team.
Better yet, learn from others who’ve already run experiments. The knowledge exists but it’s often tribal—in people’s heads rather than published. Find practitioners and ask specific questions about their tests.
The AI Content Scaling Dilemma
Can you use AI to create content that ranks in AI search? Yes, but with constraints.
What doesn’t work
Mass-producing generic content with no human oversight consistently fails. LLMs can detect when content is just rehashing existing information without adding value.
What works
AI-generated content works when:
- It supplements existing pages with structured data summaries
- It includes original insights from human experts
- It’s used for dynamic, personalized content on product pages
- A human reviews and edits every piece
The pattern: AI as an assistant, not the author. Use it to scale the mechanical parts while humans provide the insight and quality control.
How Much Revenue Actually Comes from AI Search?
For B2B SaaS companies with tech-forward audiences, current estimates range from 4-20% of traffic coming from AI search interactions.
The challenge? Attribution is difficult. When someone asks ChatGPT for recommendations, sees your brand, opens a new tab, types your domain directly, and converts—it shows up as direct traffic in your analytics.
LLMs influence decisions invisibly, showing up in your data as branded search or direct traffic rather than as a distinct channel.
The new buying behavior
People now use LLMs differently than traditional search:
- Discovery: “What are the best [category] tools?”
- Decision: “Compare these three options and recommend one”
LLMs excel at being decision assistants, not just information finders. You need to optimize for both stages: getting into consideration sets and winning head-to-head comparisons.
Who Will Win the AI Search Race?
Short term: Google vs. OpenAI
The next 12 months will be dominated by competition between Google (with Gemini and AI Overviews) and OpenAI (with ChatGPT and SearchGPT).
The incumbency advantage
Google has 14 billion searches per day and decades of user trust. They don’t need to acquire users or change behavior—they just need to integrate AI into existing products people already use.
Most people still haven’t tried ChatGPT. Google’s challenge is simpler: make their existing users’ experience better with AI.
The real battle
The actual competition might not be Google vs. ChatGPT. It might be traditional search vs. AI search—which means they’re both fighting the status quo rather than each other.
Key Takeaways
Measurement is mandatory. You can’t optimize your AI search presence without tracking it. Start monitoring how often LLMs mention your brand and how you compare to competitors.
Original insights win. LLMs favor authoritative sources with unique information. Generic content that rehashes existing information gets ignored.
Strategic relationships matter more. Reciprocal mentions between related brands create stronger signals. Integration partnerships and co-marketing become more valuable.
The platforms will evolve. Manipulative tactics that work today will gradually get addressed. Build for long-term authority, not short-term exploits.
Attribution is broken. Current analytics can’t properly track AI search influence. Revenue is higher than most tools report because conversions show up as direct or branded traffic.
What to Do Next
If your brand isn’t tracking AI search visibility yet, that’s the first step. You need to know:
- Your mention frequency across ChatGPT, Claude, and Gemini
- Your competitive position in category queries
- Which sources LLMs cite when they mention you
- How your visibility trends over time
Tools like Cartesiano.ai give you this visibility, letting you monitor your AI search presence the same way you monitor traditional SEO rankings.
The brands that win in AI search will be the ones that start measuring and optimizing now, while most competitors are still debating whether it matters.
Cartesiano.ai Blog 