From Searching to Asking: How AI Is Reshaping Brand and Content Discovery - brightonSEO Spring 2025

Session Title：The new world of AI optimisation and how it differs from SEO

In April 2025, MediaReach, Inc. participated in BrightonSEO Spring 2025, one of the world’s largest search marketing conferences held in Brighton, UK. This report covers a standout session by Marcus Tober of Semrush, titled "The new world of AI optimisation and how it differs from SEO."

As generative AI platforms like ChatGPT and Perplexity become mainstream tools for information discovery, the rules of SEO are being fundamentally redefined. It’s no longer just about keywords and backlinks—AI-native optimization now demands a deeper understanding of brand credibility, machine-readable content, and platform-specific user behaviors.

In this article, our SEO Consultant Ayaka Uchida offers on-the-ground insights from the session, highlighting key shifts that marketers and organizations must navigate to stay discoverable in an AI-driven search landscape.

Written by Ayaka Uchida
SEO Consultant, MediaReach, Inc.

1. Executive Summary

In this session, Marcus Tober from Semrush discussed how the world of AI-driven search is transforming the fundamentals of SEO. Drawing on examples like Google's growing emphasis on brand signals and how AI engines like OpenAI retrieve and use online content, the speaker emphasized that SEO isn't dead—it's evolving.

Rather than optimizing for traditional keyword-based search engines alone, marketers now have to think about content credibility, machine learning interpretation, and user signals in AI-native contexts. He also addressed how customer demographics, platforms like Reddit, and case studies like Mazda's can affect how AI systems present and prioritize content.

2. Session Details

Session Title: The new world of AI optimisation and how it differs from SEO
Speaker: Marcus Tober (SVP, Head of Enterprise Solutions, Semrush)
Date / Time: Friday, April 11, 2025 — 09:30 AM
Venue: Auditorium 1, Brighton Centre, Kings Road, Brighton and Hove, Brighton, BN1 2GR, United Kingdom
Event: BrightonSEO Spring 2025
Session Link: https://brightonseo.com/sessions/is-seo-dead-or-dying

3. Report Details

3-1. Context and Background

As generative AI tools like ChatGPT become default discovery platforms for many users, SEOs must understand how optimization changes when content is processed by AI, not humans. This talk broke down the distinction between traditional SEO and "AI optimization," which is less about rankings and more about training data, citations, and user feedback loops.

The session emphasized the importance of understanding how machine learning models learn and decide what content to surface—and how that differs from Google's algorithmic methods. Tober presented data showing that while AI platforms are growing, traditional search still dominates traffic, particularly among older demographics.

3-2. Key Messages and Takeaways

• AI optimization is context-dependent—the experience will differ based on user profile, device, and task intent.

• Google AI overviews appear in 75% of problem-solving queries and 1/3 of questions with how/why words.

• AI crawlers do not render JavaScript, making JS-heavy websites essentially invisible to AI systems—as demonstrated by the Mazda example.

• Keywords aren't dead—they carry critical user intent that marketers need to understand.

• Reddit is an official OpenAI partner, meaning even unreliable comments can significantly influence AI responses about brands.

• AI search engines favor smaller websites more than Google does, creating opportunities for sites that struggle with traditional rankings.

• Demographic differences matter—younger users prefer platforms like TikTok/Instagram while older audiences rely more on Google.

3-3. Visual Materials and Slides

Figure 1: Session agenda challenging "keywords are dead" claim：

Figure 2: Comparison between SEO and AI-driven discovery models：

Figure 3: Market competition analysis between search platforms：

Figure 4: Google Search Console data contradicting "keywords are dead"：

Figure 5: Search engine overlap analysis：

Figure 6: Mazda case study on JavaScript limitations：

3-4. Practical SEO Implications

Short-term:

• Review brand representation across public platforms like Reddit and forums.

• Adapt content to answer-like formats AI systems can parse easily.

• Monitor what parts of your content are being used by AI through citation tracking.

• Ensure critical content isn't JavaScript-dependent for AI visibility.

Long-term:

・Invest in brand-building that enhances credibility in AI models.

・Audit how your content appears in open training sets and AI overviews.

・Redefine SEO as source eligibility optimization for generative platforms.

・Consider demographic differences in platform usage when creating content strategies.

3-5. On-site Impressions

This session opened day two in the main auditorium with the eye-catching theme "Is SEO dead or dying?", which naturally drew strong attention. The topic was provocative, and the room filled quickly—demonstrating how much relevance SEO still holds in the AI era. The speaker delivered both data and humor, keeping the audience fully engaged throughout.

3-6. Personal Reflection

The point about age groups and search behavior really resonated with me—SEO isn't one-size-fits-all. For older audiences, traditional Google results are still central, and in such cases, classic SEO remains crucial.

Reddit being frequently cited as a source by AI systems was another key takeaway. Living in Europe, that feels natural, but it reminded me that in Japan, Reddit isn't widely used—so we might easily underestimate its influence. This session was a strong reminder of the importance of maintaining a global perspective. The anecdote about an 11-year-old's comment being cited by AI was both funny and telling. It highlighted how brand perception is no longer shaped solely by official content, but by unexpected third-party platforms. This isn't just a marketer's concern—it's something entire organizations need to take seriously.

The Mazda example was also powerful. It clearly showed how JavaScript-heavy websites can go uncrawled by AI systems—a practical and overlooked risk. It made the technical side of AI optimization very real.

4. Supplementary Materials

Written by Ayaka Uchida
SEO Consultant, MediaReach, Inc.