Jim Yu is the Founder and CEO at BrightEdge, a platform for real-time insights into search and SEO - which gives him a front row seat for the changing state of discovery and the huge role being played by AI.
He spoke to Econsultancy about the elephants in the room for search and discovery; why 'share of conversation' is becoming as important as traditional search rankings; and why businesses should go back to basics in order to successfully show up for AI agents.
Jim Yu: It's less of an elephant and more of a giant that grows: Google. Despite the explosion of new LLMs and AI-powered discovery engines like ChatGPT, Perplexity, and Claude, BrightEdge's latest data shows Google still captures over 90% of total search referrals nationwide. That dominance means even the most forward-thinking 'AI-search strategies' need to account for Google-not replace it.
There's a significant gap between perception and reality here. We're seeing tremendous growth from AI discovery platforms, but in terms of actual market share and referral traffic, we're still talking about roughly 1%. The headlines don't always reflect what the data shows.
The real elephant, if I'm being honest, is the misconception that GEO and AEO are replacing SEO. They're not. SEO has never been more vital-it's the foundation for success everywhere, including in generative AI. Technical SEO-site crawlability, indexation, architecture, Core Web Vitals, structured data-is what makes your content machine-readable for LLM crawlers and AI Overview systems.
The classic SEO pillars-intent-mapped content, E-E-A-T signals, internal linking, and performance-are precisely the signals AI systems and answer engines lean on when deciding which sources to surface and trust.
GEO and AEO build directly on these foundations, restructuring sites and content so generative engines can parse entities, relationships, and FAQs, then confidently quote or cite them in answers. And with the rise of agentic AI, SEO is only becoming more important.
The other elephant? Marketers need to rethink what drives conversions versus discovery, where to focus their efforts, and how to measure success with new standards and metrics.
Source: BrightEdge AI Pulse data
Expert content with clear structure continues to dominate because AI engines reward authority, not volume. If you're producing content at scale without genuine expertise behind it, you're wasting resources.
Reviews and user-generated content have become critical trust signals. BrightEdge data shows AI agents increasingly lean on sentiment and reviews when making product recommendations, especially in retail. This is a multi-format world for multi-model LLMs and multiple platforms-YouTube, UGC, and social all play important roles.
But ultimately, what matters most is content that AI can understand. Formats with strong metadata, schema markup, structured Q&A sections, and clear authority signals perform better because they're easier for LLMs to parse and cite.
Here's something marketers need to understand: AI doesn't cite content that simply repackages existing information-it can generate that itself. What AI systems do cite are unique insights, original data, and trusted sources. Quality E-E-A-T content remains foundational. Earned media and PR have also become increasingly valuable as external validation signals.
I often say that SEO isn't just search-visible anymore-it's AI-visible. Measuring impressions, citations, mentions, and sentiment in AI responses is the new competitive edge. Your 'share of conversation' across AI engines is becoming as important as traditional search rankings.
Category-specific AI behaviour is another consideration. AI engines surface different types of brands depending on the category-finance, retail, and B2B all behave differently. Marketers must understand these nuances rather than assuming one-size-fits-all visibility.
In terms of organisational changes, we're seeing companies lean into SEO as a discipline and invest in the people and expertise behind it. Integration has become essential-SEO, PPC, design, UX, content, and PR all need to work together.
Originally, organisations that achieved this integration gained a competitive advantage. Now, it's simply a requirement for survival. Smart brands are investing in opportunities for revenue now while future-proofing their brand for what's coming.
AI visibility literacy is at the top of the list. Marketers need to understand how AI engines interpret, summarise, and surface their brand. Most teams can optimise for Google rankings, but few know how to optimise for AI engines-and these are different disciplines.
Technical foundations that make content 'agent-ready' are essential. Skills like structured data implementation, clean HTML, schema markup, accessibility, and real-time feed management are now critical because AI engines rely on these signals to confidently use and recommend content.
New measurement skills are becoming increasingly important. As discovery moves from clicks to answers, teams need people who can measure brand presence without relying solely on traffic. Skills in log analysis, AI impressions, citation tracking, and 'share of conversation' metrics are becoming critical differentiators.
Experienced technical SEOs and content and earned media professionals remain in high demand. We're also seeing demand for what I'd call 'marketing engineers'-people who understand LLMs, prompting, and how to translate visibility and presence into perception and performance.
What's most often missing? Brand storytellers who can weave together different types of top-of-funnel and bottom-of-funnel content cohesively. The technical skills are easier to hire for than the strategic narrative capability.
I wish everyone would stop with the 'shiny new object syndrome.' Every week there's news about another AI player, another feature launch, another platform shift. Stop chasing the news. Focus on what matters now for revenue while keeping a close eye on the broader AI and LLM landscape.
'AI will replace SEO' is another one I'd like to see retired. Google is still the top winner. AI hasn't killed search-it's expanding it. Similarly, the idea that a volume strategy still works is outdated. AI-driven ecosystems punish low-quality or duplicative content. The mass-production era is over.
As for key areas of focus, brands need to prepare for AI Agent Optimisation. AI will soon make decisions on consumers' behalf-purchases, anticipating needs, routine tasks. Brands need to ensure their data, content, and product feeds are 'agent-ready.'
Don't get lost in the noise. Go back to basics with your brand entity: Who you are, what you do, why your products or services matter, and what differentiates you from the competition. Don't build content for content's sake-that's just adding to a problem that existed before.
Build on your brand foundation and use AI to achieve AI success at scale. Otherwise, AI will define your brand for you-and you might not like what it says.
Six months ago, CMOs were asking whether they should invest in this space. Now it matters more than ever. Brands earn the right to win-don't lose it.
For a safe bet: nothing is safer or surer than Google continuing to innovate and dominate. Their foundational moats-data, intent signals, behaviour patterns, knowledge graphs, the Shopping Graph-are enormous. The entire search and AI ecosystem feeds into an even wider ecosystem of products, assets, and cloud services.
For something more outrageous: by 2026, AI agents will make more everyday shopping decisions than humans. For routine purchases-household basics like shampoo, detergent, everyday essentials-consumers will hand full control to AI. Brands will be competing for the agent's choice, not the shopper's click.
