Brands Must Optimize for LLMs, Not Just Search

The rules of digital marketing are being rewritten not by Google, but by the conversational AI tools that consumers are increasingly turning to for product recommendations, research, and purchasing decisions. While brands have spent decades perfecting their search engine optimization strategies, a seismic shift is underway that threatens to render those efforts insufficient, if not obsolete.

The numbers tell a compelling story. In January 2024, just 25% of consumers reported using generative AI tools for product recommendations. By early 2025, that figure had surged to 58%, according to research published by Jellyfish and INSEAD. This isn't a gradual evolution—it's a wholesale migration of consumer behavior happening in real time.

The retail sector provides perhaps the most dramatic evidence of this shift. Adobe Analytics, which tracks over 1 trillion visits to U.S. retail sites, reported a staggering 1,300% year-over-year increase in traffic from generative AI sources during the 2024 holiday shopping season. On Cyber Monday alone, AI-driven traffic jumped 1,950% compared to the previous year. By July 2025, AI referrals had exploded by 4,700% year-over-year.

This trend extends far beyond retail. Travel websites saw a 1,700% increase in AI-sourced traffic, while banking sites experienced a 1,200% surge over the same period. ChatGPT now accounts for nearly 21% of Walmart's incoming referral traffic, over 20% for Etsy, and approximately 15% for Target.

The implications are clear: large language models like ChatGPT, Claude, Gemini, and Perplexity have evolved from experimental chatbots into critical intermediaries between brands and consumers. As Jack Smyth, Chief Solutions Officer at Jellyfish, puts it:

"LLMs are no longer just tools; they are a critical part of the customer journey and an audience in their own right."

When Brand Awareness Doesn't Translate to AI Presence

Research by Jellyfish and INSEAD reveals a troubling disconnect: a brand's visibility to large language models can differ dramatically from traditional consumer awareness of that brand.

In their groundbreaking study using proprietary "Share of Model" methodology, researchers analyzed how various LLMs perceive and recommend brands across different markets. The findings expose a new digital divide. In Italy's laundry detergent market, for instance, the brand Chanteclair enjoys a 19% share of model on Perplexity but disappears completely from Meta's Llama. Other established brands show similarly erratic visibility across different AI platforms.

David Dubois, Associate Professor of Marketing at INSEAD and co-author of the research, explains the stark reality: "While search engines still display less popular brands on later pages, AI models are merciless. If your brand doesn't register with an AI model, it simply won't appear at all."

The researchers identified four distinct categories of brands based on their AI visibility versus traditional awareness:

Cyborgs like Tesla and BMW maintain strong awareness among both humans and AI models. Tesla's success stems partly from Elon Musk's omnipresence but also from content that emphasizes specific, measurable features—precisely what AI models value.

AI Pioneers such as electric vehicle startup Rivian score high with AI models despite limited mainstream consumer awareness. These brands typically focus on solution-oriented content that aligns with AI processing preferences.

Legacy Leaders maintain high traditional awareness but struggle with AI visibility due to outdated digital content strategies.

The Invisible lack both traditional awareness and AI presence, facing an uphill battle on both fronts.

This categorization reveals an uncomfortable truth: decades of brand-building and traditional marketing excellence provide no guarantee of visibility in AI-mediated consumer journeys.

The Economics of AI-Driven Discovery

The financial implications of this shift are becoming increasingly clear. According to Semrush research, the average LLM referral visitor is worth 4.4 times more than the average traditional organic visitor. Why? Because AI tools better equip users with information before they decide to visit a website, resulting in higher-quality, more purchase-ready traffic.

Adobe data supports this conclusion. While AI-driven retail traffic initially had conversion rates 82% lower than non-AI traffic in July 2024, that gap had narrowed to just 9% by February 2025. AI-referred visitors demonstrate 23% lower bounce rates, spend 41% longer on sites, and generate 12% more page views per session compared to traditional traffic sources.

For travel websites, the quality differential is even more pronounced. AI-referred visitors generate 80% higher revenue per visit, with bounce rates 45% lower than traffic from other sources.

Gartner predicts that by 2028, 50% of search engine traffic will shift to AI-powered platforms. Market projections suggest LLMs will capture at least 15% of the search market by that same year, with the global LLM market expected to grow by 36% from 2024 to 2030.

Consumer adoption patterns reinforce these projections. Among Gen Z users (ages 16-24), 66% now turn to AI tools like ChatGPT for brand, product, and service recommendations. Even Baby Boomers, initially skeptical of generative AI, demonstrated 63% adoption growth in just five months during late 2024 and early 2025.

The Share of Model Imperative

Traditional marketing metrics, brand awareness, share of voice, share of search—are being joined by a new critical KPI: Share of Model. This metric, pioneered by Jellyfish, measures how often brands appear in LLM-generated responses relative to competitors.

Early adopters are already seeing results. Catherine Lautier, VP and Global Head of Media at Danone, notes: "LLMs will increasingly influence customer behavior, and the Share of Model platform enables us to track and compare each LLM's perception." Gokcen Karaca, Head of Digital and Design at Chivas Brothers, adds: "As brands, we must ensure our products are represented in those critical AI-driven responses."

Beta trials of Jellyfish's Share of Model platform revealed significant opportunities for brands willing to optimize for AI visibility. Participants expanded their keyword strategies, adjusted website text and image assets to appeal to different models, and gained competitive intelligence by understanding how AI systems perceive rival brands.

The platform's three-pronged analysis framework provides actionable insights:

Mention rate: How often brands appear in LLM responses
Human-AI awareness gap: Disparities between consumer surveys and LLM visibility
Brand sentiment: How AI models perceive brand strengths and weaknesses

Practical Strategies for LLM Optimization

The good news is that many principles of effective LLM optimization build upon—rather than replace—sound SEO practices. Research analyzing over 10,000 real-world search queries reveals that brands ranking on Google's first page appear in ChatGPT answers 62% of the time. Strong SEO performance remains foundational to AI visibility.

However, LLM optimization (LLMO) requires specific tactics beyond traditional SEO:

1. Content Structure Optimization

LLMs favor content that provides clear, self-contained answers to specific questions. Structure matters enormously. Content featuring original statistics and research findings sees 30-40% higher visibility in LLM responses, according to multiple studies. This preference stems from LLMs' built-in verification processes that seek to support claims with concrete data.

Effective approaches include:

Creating FAQ-style content that directly answers common questions
Using clear hierarchical structure with descriptive headers
Providing complete context within individual sections rather than requiring readers to piece together information from multiple pages
Including specific, measurable details rather than vague marketing language

2. Building Topic-Entity Associations

LLMs learn about brands by analyzing patterns across millions of web pages. When your brand consistently appears alongside specific topics on authoritative sites, AI systems begin to associate your brand with expertise in those areas.

This requires a strategic approach to digital PR and content distribution. Effective tactics include:

Creating newsworthy content like original research or industry surveys that earn media mentions
Responding to journalist queries through services like HARO to build authoritative backlinks
Contributing expert commentary to industry publications
Ensuring product descriptions emphasize unique features and use cases rather than generic benefits
Maintaining consistent terminology around your core offerings across all digital properties

3. Authority Signal Reinforcement

Quality backlinks remain crucial, but the context of those links matters more than ever. Links from authoritative sources that mention your brand in the context of topics you want to be associated with are particularly valuable for LLMO.

Wikipedia presence proves especially important, as it's frequently referenced by LLMs. While maintaining accuracy on Wikipedia requires careful adherence to their guidelines, verified presence on the platform provides significant authority signals.

4. Multi-Platform Optimization

Different LLMs prioritize different content attributes. Google's AI Overviews favor established brands with strong traditional SEO foundations. Perplexity emphasizes real-time accuracy and freshness signals. Claude prioritizes well-structured, comprehensive content with clear sourcing.

A robust LLMO strategy must account for these differences while focusing on universal best practices that work across all platforms. Brands should test standardized sets of industry-relevant questions across multiple LLMs monthly to track progress and identify platform-specific optimization opportunities.

5. Leveraging Structured Data and Technical SEO

While LLMs process content differently than search engines, technical foundations remain critical. Strong site architecture, structured data markup, clear metadata, and mobile optimization all contribute to how effectively AI systems can parse and understand your content.

Bain & Company reports that 80% of consumers now rely on AI-written summaries for at least 40% of their searches. This means your content must be easily extractable and quotable—characteristics enabled by solid technical implementation.

The Tools and Measurement Challenge

One of the biggest obstacles to effective LLMO is measurement. Unlike SEO, where tools like Semrush, Ahrefs, and Google Analytics provide comprehensive visibility into performance, the LLMO ecosystem is still developing its measurement infrastructure.

Adobe's LLM Optimizer, launched in June 2025, represents the first enterprise-grade solution for tracking and improving brand visibility across major LLMs. The platform provides:

Real-time insight into "agentic traffic"—activity from AI crawlers and LLM-based assistants accessing site content
Benchmarking capabilities to compare brand visibility against competitors
Automated recommendations for improving content discoverability
Integration with Adobe Experience Manager Sites for rapid implementation of suggested changes

HubSpot's AI Search Grader offers similar capabilities for analyzing brand visibility across OpenAI and Perplexity platforms, while newer tools from Writesonic, Analyzify, and others are emerging to fill specific niches.

However, manual monitoring remains valuable. Brands should regularly query major LLMs about their products and services, tracking:

Mention frequency and context
Sentiment of recommendations (positive, negative, neutral)
Accuracy of information presented
Presence or absence compared to competitors
Which specific content gets cited most frequently

The Content Quality Imperative

While technical optimization matters, the fundamental driver of LLM visibility remains content quality. AI systems are designed to surface helpful, accurate, authoritative information. Gaming these systems through keyword stuffing or manipulation tactics not only proves ineffective but can actively harm your brand's reputation.

Recent research has exposed attempts at "strategic text sequencing" and other black-hat tactics designed to manipulate LLM recommendations. Studies show that in about 40% of cases, carefully crafted prompt injections can influence product rankings. However, as AI platforms become more sophisticated, such manipulation becomes both easier to detect and more likely to result in brand penalties.

The sustainable path to LLM visibility lies in creating genuinely valuable content that:

Provides unique insights or data not available elsewhere
Thoroughly addresses user questions and concerns
Demonstrates clear expertise and authority
Remains current and accurate
Cites sources and provides verifiable information
Uses natural language that serves human readers first

The Timeline for Results

Unlike traditional SEO, where building authority can take months or years, LLM optimization can show faster results because you're optimizing for extractability and clarity—factors more directly under your control.

Most brands implementing systematic LLMO strategies begin seeing measurable improvements within 60-90 days, according to multiple case studies. The typical optimization timeline follows this pattern:

Days 1-30: Baseline measurement and low-hanging fruit fixes (adding FAQ sections, improving page structure, fixing technical issues)

Days 31-60: Strategic content optimization (restructuring high-value pages, enhancing citations, adding missing context)

Days 61-90: Content creation and refinement (developing new LLM-optimized content, measuring performance, iterating based on results)

Results vary based on starting authority and competitive intensity, but the trajectory is clear: brands that begin optimizing now gain first-mover advantages in an increasingly critical channel.

The Cost of Inaction

Perhaps the most compelling argument for prioritizing LLM optimization comes from understanding the cost of neglect. When consumers turn to AI for recommendations and your brand doesn't appear, you don't get demoted to "page two"—you simply cease to exist in that consumer's consideration set.

Flow's analysis of 5,000 HR and workforce management keywords revealed substantial traffic changes as Google integrated AI Overviews directly into search results. Some brands experienced click-through rate declines of nearly 20% for non-branded keywords as AI summaries displaced traditional organic listings.

The estimated 25% drop in organic traffic that some categories are experiencing represents not just lost clicks, but lost opportunities to influence purchasing decisions, build brand affinity, and capture market share.

As Jon Sica, COO of Batteries Plus, explains: "Currently, about 1% of the referral traffic we receive comes from an AI source. But by monitoring the sources in citations from answers where our brand is being mentioned, we can take action to increase our share of mentions by AI models." His team's proactive approach—auditing AI results, identifying citation sources, and optimizing accordingly—exemplifies the strategic mindset required in this new landscape.

A New Marketing Discipline Emerges

We're witnessing the birth of a new marketing discipline. Just as SEO evolved from a technical afterthought to a core marketing function over the past two decades, LLM optimization is rapidly becoming essential to digital strategy.

Harvard Business Review suggests that SEO professionals will soon be known as LLMOs (LLM Optimizers). Whether that prediction proves accurate, the underlying reality is undeniable: the skills, strategies, and metrics that drove success in the search engine era must evolve to address the AI-mediated discovery era.

The good news is that this evolution builds upon—rather than replaces—existing marketing fundamentals. Strong brands with authoritative content, clear value propositions, and genuine expertise are well-positioned to succeed. The challenge lies in translating those assets into formats and structures that AI systems can effectively process and recommend.

The Urgency of Now

More than a billion people in over 100 countries now engage with Google's AI Overviews every month. ChatGPT processes over 1 billion user messages daily. Perplexity, Claude, Gemini, and other AI platforms are seeing exponential user growth and increasingly sophisticated capabilities.

This isn't a future trend to monitor—it's a present reality demanding immediate strategic response. Brands that master this transition position themselves as essential participants in the algorithmic conversations increasingly shaping consumer decisions. Those that don't risk becoming invisible in an AI-driven marketplace where LLMs serve as gatekeepers to consumers.

The playbook is still being written, and the tools are still maturing. But the direction is clear, the data is compelling, and the window for establishing early leadership is open. The question isn't whether to optimize for LLMs, but how quickly you can begin and how thoroughly you can execute.

Your customers have already changed how they search. The only question is whether your brand will be there when they do.

Key Takeaways

Consumer behavior has shifted dramatically: 58% of consumers now use AI tools for product recommendations, up from 25% in 2023, with traffic from AI sources growing by 1,300%+ in retail alone.
Traditional brand awareness doesn't guarantee AI visibility: Brands can have strong consumer awareness yet be completely invisible to LLMs, or vice versa, creating a new strategic challenge.
AI-driven traffic is higher quality: LLM referrals generate visitors worth 4.4x more than traditional organic traffic, with better engagement metrics and improving conversion rates.
Share of Model is the new critical metric: Tracking how often your brand appears in LLM responses relative to competitors provides essential insight into AI visibility and influence.
Optimization requires specific tactics: While building on SEO fundamentals, LLMO demands focus on content structure, topic-entity associations, authoritative citations, and multi-platform optimization.
Measurement tools are emerging: Solutions like Adobe LLM Optimizer, HubSpot AI Search Grader, and Jellyfish's Share of Model platform enable systematic tracking and improvement.
Results can come relatively quickly: Unlike traditional SEO, LLMO can show measurable improvements within 60-90 days due to its focus on extractability and clarity.
The cost of inaction is invisibility: In AI-mediated search, there is no "page two"—brands either appear in responses or cease to exist in consumer consideration sets.
This is happening now, not later: With over 1 billion daily ChatGPT users and AI accounting for 15-20% of major retailer referral traffic, the future has arrived.
Quality content remains paramount: Sustainable LLM visibility comes from genuinely valuable, authoritative content rather than manipulation tactics or gaming the system.