Liberty Global CEO Releases New AI Marketing Guide: Measuring Model Share to Win in Generative Engine Optimization

Jon Williams, founder and CEO of Liberty Global (Nasdaq: LBTYA), a leading global communications and media company, recently published a column in industry media outlet The Drum, proposing systematic strategies to address the growing concern over AI agents' impact on brand visibility. The article points out that in the era of AI-driven search and recommendations, brands need to shift from traditional search rankings to measuring "share of model" and optimize for generative engines to gain priority recommendations from AI systems.

Background: AI agents are not brand-blind, but brand-obsessed

Previously, renowned marketing scholar Mark Ritson warned that large language models (LLMs) are not "blind" to brands but rather "brand-obsessed"—where awareness is no longer a soft asset but data itself. Brands that can "train" into the model get named, while weaker brands are quietly ignored. Williams believes this trend has been clearly articulated, but most marketers have yet to take concrete action.

Core Strategy: Measuring 'Share of Model'

Williams proposes "Share of Model" as the core KPI for the AI era and recommends the following quantifiable metrics:

• Brand Mention Rate: How often the brand appears in AI responses
• Recommendation Rate: How frequently the brand is the top recommendation rather than just one of several mentions
• Prompt Coverage: The number of relevant query types where the brand appears
• Share of Voice: Frequency of mentions compared to direct competitors
• Model-Specific Visibility: Performance differences across various AI engines like ChatGPT, Perplexity, Gemini, and Claude
• Visibility Fluctuation: The phenomenon of a brand suddenly disappearing, known as a "citation cliff"

Williams also lists three tools that can measure these metrics: Semrush AI Visibility Toolkit, Profound, and Peec AI, which track citation shares, visibility scores, and competitive voice across different engines.

From SEO to GEO: Optimizing to Win Recommendations

The article notes that traditional search engine optimization (SEO) focuses on ranking positions, while generative engine optimization (GEO) aims to become the preferred citation in AI responses. A joint study by Princeton University, Georgia Tech, the Allen Institute for AI, and the Indian Institute of Technology Delhi shows that the following content optimizations can increase visibility in generative engines by up to 40%:1. Authoritative statistics with citations: Use hard data from trusted sources. 2. Fluid, confident, and declarative language: Avoid ambiguous expressions. 3. Answer-first structure: Clearly answer the question as early as possible, then expand on the discussion. 4. Precise, accurate terminology: Use industry terms appropriately.

Williams emphasizes that GEO is especially suitable for B2B enterprises, requiring the establishment of authority across multiple channels beyond their own website.

The Paid Recommendation Layer Has Arrived: Seizing the First Move Is Key

Looking back at Google's launch of advertising products in 2002, early adopters acquired keyword real estate at extremely low costs, while latecomers had to pay several times more. Williams believes that the current AI recommendation channel is at a similar stage. The following three platforms have already started offering paid or commercialized AI discovery products:

• Google AI Mode: Testing conversational ad formats, including shopping ads and sponsored placements.
• Perplexity: Introducing sponsored and commercial formats, building shopping and product discovery features.
• ChatGPT: Launching shopping research, product discovery, instant checkout, and advertising products; its Atlas browser integrates ChatGPT into browser-based research tasks.

Third-Party Validation: Let the Brand Be Defined by Others

Williams points out that in the GEO environment, brands should actively encourage third-party sources to define themselves, rather than forcefully controlling the narrative. Key signals include:

• Reddit: Actively participate in community discussions, establishing relevance to keywords.
• Wikipedia: Ensure entries are accurate and complete to increase the probability of being cited by AI.
• Review platforms: Accumulate genuine reviews on channels such as Google Reviews, Trustpilot, G2, and Tripadvisor.
• Editorial coverage: Independent media endorsements can strengthen entity prominence, which is difficult to replace with brand-owned content.

Conclusion

"Every new technology seems obvious in hindsight—the printing press, the internet, AdWords, social media. Each time, a small group of early adopters seize the channel, then spend a decade watching others catch up at ten times the cost," Williams writes. "GEO is such a moment. Write well, earn trust, build a presence where real conversations happen, and preempt before the channel becomes crowded and expensive. The great fundamentals of marketing haven't changed; they just need to appear in unfamiliar places."

About Liberty Global

Liberty Global is a world-leading provider of broadband, video, and communication services, operating in multiple European countries, offering high-speed internet, television, and mobile communication services through its brands. The company is committed to technological innovation and continues to produce industry insights in AI marketing.