Brand identity in the LLM era: the touchpoint you're not managing

Luigi Greco, founder of Conflux, recently published an analysis that highlights a phenomenon many brands are already experiencing without fully understanding it yet: brand identity is forming in a new place, outside the touchpoints we've always designed and controlled.

Imagine a team that spent months working on a brand's new packaging, meticulously crafting every detail to achieve something distinctive and authoritative. Now consider that nine out of ten users, in their first interaction with that brand, will never see that packaging. A new reality that cannot be ignored.

The click collapse: when users stop at the AI answer

Pew Research published a study in July 2025 showing that when an AI summary appears in Google search results, only 1% of users click on the cited sources. The overall click-through rate for traditional results drops from 15% to 8%, a 47% reduction. Meanwhile, Semrush, analyzing Google's AI Mode in September 2025, found that 93% of those searches end without any clicks.

In practice, when a user asks, for example, "best skincare for sensitive skin," they read the AI's answer and move on. They will never visit the brand's website. Bain & Company has mapped this behavior on a broader scale. 80% of consumers now rely on AI-generated results for at least 40% of their searches, and 60% of searches end without the user clicking on any external site. 

Furthermore, Seer Interactive (2025) measured a 61% year-over-year drop in organic click-through rate for queries where an AI Overview appears. This decline signals the end of the click economy for entire categories of informational search.

The consideration set is no longer formed during the search process

51% of consumers (Gartner, 2025) state that their search habits have changed due to generative AI. In the UK, Klaviyo's AI Shopping Index (October 2025) reports even higher adoption rates: 80% of UK shoppers already use AI tools for shopping or product research, and 70% expect AI assistants to be the norm by 2026.

This shift transforms how brands are selected and evaluated. Every consumer has their own consideration set, which is the group of three or four brands they consider before making a purchase. Until recently, this set was built by exploring websites, reading reviews, and comparing options. Today, it's pre-selected by AI, based on a representation of your brand that you didn't directly design.

The three layers of brand identity

To understand what's happening, we need to map out how brand identity is structured today.

Layer 1 is the classic territory of brand design: visual system, packaging, retail, website, advertising. Everything you directly design, own, and control. It's the work that has brought brands this far.

Layer 2 is the extended digital identity: SEO, social media, reviews, third-party articles. Here, control is partial. You can influence it through content and strategies, but you don't directly control what is said about you.

Layer 3 is new. It's the brand as reconstructed by ChatGPT, Perplexity, Claude, and Google AI Overviews when someone asks a category-level question. The answer isn't a page you wrote. It's generated in real-time, probabilistically assembled from what the model saw during training. This is the layer where the first impression of your brand is formed for a growing share of consumers. And first impressions always carry significant weight.

AI responses as new packaging: what the research says

Academic research is beginning to document the underlying dynamics, confirming the evidence from market reports. Lee and Park, in a study published in the Journal of Consumer Behaviour in 2022, demonstrated through a structural equation analysis that the quality of communication with an AI chatbot significantly predicts whether consumers will continue to use that channel. The medium of interaction shapes the relationship before any physical contact with the brand. The chatbot functionally becomes the new packaging, but it wasn't designed by the company.

Khan and Mishra (Journal of Service Theory and Practice, 2023) formalized a perceived AI credibility framework based on Source Credibility Theory. They identify four dimensions of the consumer-AI experience through which AI credibility shapes brand-relevant outcomes: data acquisition, classification, delegation, and social interaction. These are the same dimensions that brand teams have always designed for other media (posters, in-store assistants, digital interfaces) but now the medium carrying them is a large language model that processes co-occurrence statistics across billions of documents.

Beyond marketing literature, a growing body of work from Stanford HAI, University of Chicago, and Hugging Face documents systematic representational biases in how LLMs represent commercial entities and demographic groups. Biases that no brand team chose and few are currently measuring.

The irony: the more AI there is, the more classic brand building matters

Research confirms the phenomenon from an academic perspective. But the strategic implications are even deeper. For two decades, digital marketing has repeated the same mantra: optimize the funnel, personalize everything, obsessively measure CTR, chase last quarter's performance. Today, the wind is blowing in the opposite direction, and it's doing so with a force few had anticipated.

LLMs don't "click," they remember. To appear in their responses, one needs to be densely present in the reference corpus: editorial coverage in authoritative publications, credible and consistent reviews, presence in curated lists, citable papers, industry directories. It's exactly the classic work of brand building that Byron Sharp and Mark Ritson have always advocated, translated into the new medium.

McKinsey, in its report State of Marketing Europe 2026 (November 2025), reported that branding has returned to being the number one priority for European marketing leaders in 2026, with 72% planning to increase budgets, a return to fundamentals after a decade dominated by performance marketing.

The more the world becomes AI-mediated, the more the fundamentals of classic branding regain decisive importance. Brands that have invested in authority sustained, communicative distinctiveness and quality editorial presence are selected. Brands that have pursued last-quarter tactics and short-term performance optimizations lose visibility precisely where the new consideration set is forming.

This phenomenon is directly linked to what we described when talking about how UX and CRO are evolving: algorithms mediate the entire purchasing process. Decisions are still made by people, but the touchpoints and factors influencing those decisions have radically changed. Being excluded from an AI-generated shortlist is the equivalent of not being indexed by Google ten years ago.

Three implications that change brand strategy

The first implication is structural. Brand identity today must be conceived and measured through three layers, not two. Layer 1 remains fundamental. Layer 2 is not superseded. But ignoring Layer 3 is like having designed the most beautiful packaging in the supermarket in an era when the majority of consumers no longer enter that supermarket. They buy elsewhere, guided by descriptions that companies do not control.

The second implication concerns metrics. The relevant question is no longer "is the brand mentioned?" but "how is it represented?". A brand frequently mentioned but poorly framed is in a worse position than one mentioned less but represented accurately. The way an LLM describes the brand when it mentions it shapes the perception users will have of that brand. And that narrative emerges from what the model has absorbed: reviews, articles, mentions, context.

The third implication is the most subtle. The touchpoint is no longer the origin of perception, but its confirmation. When a user arrives on the website or in a store, the first impression has already been formed in Layer 3. The touchpoint's role is to confirm that narrative, not to create it. If it confirms it, trust is strengthened. If it contradicts it, trust is broken before there's even an opportunity to build it.

How brand identity changes with AI

Understanding what brand identity means in a world where an increasing share of discovery is mediated by machines is becoming ever more relevant. It's an open question, worth reflecting on, and one whose effects are essential not to ignore.

Tools and methodologies for measuring the consequences of Layer 3 are emerging. But what seems most important right now is for brand strategy teams to start considering Layer 3 with the same seriousness they apply to designing packaging, retail, and advertising campaigns.

Layer 3 is the place where the brand's first impression is formed, built from the synthesis of everything done in Layers 1 and 2 (editorial authority, reviews, presence in credible sources) and reconstructed by probabilistic systems operating outside the brand's direct control. Ignoring this level means surrendering the most important narrative, the one that precedes every touchpoint, to an algorithmic synthesis that the brand did not design.

Is your brand ready for Layer 3?

Most companies are still only measuring what they can directly control, while their brand's first impression is formed elsewhere, in AI-generated responses. The question is no longer whether this is happening, but how your brand is represented when a model answers a category question, and what you can do to influence that narrative.

At Conflux, we help brands understand how they are reconstructed by LLMs, measure the quality of that representation, and build the editorial presence and authority that determine selection in the new consideration set. If you want to know how your brand appears in Layer 3 and set up a strategy to manage it, contact us: we'll analyze your situation together and identify concrete levers for action.

‍