Agent-influenced shopping will account for $190 billion to $385 billion of U.S. e-commerce spending by 2030, according to Morgan Stanley estimates — between 10% and 20% of the market. That figure frames a structural shift for chief executives: the most important intermediary between brands and shoppers is no longer a retailer or a search ranking, but artificial intelligence.

The mechanism is already live. A shopper asks an AI assistant for sneakers for a 23-year-old; another searches Google for a boutique hotel in Rome. In both cases, AI helps decide which brands deserve consideration before the customer sees a single product image or room photo. Mental availability — the likelihood a brand comes to mind in a buying situation — is no longer formed directly by communications and shopper experience alone. A machine intermediary now sits in the path.

What does AI gatekeeping mean in practice?

AI shapes choices without shoppers opening a chatbot. It is embedded in familiar services such as Google, where AI-generated answers appear before users reach a brand's own website. Fragmented messaging or unsupported claims leave AI systems with conflicting information, which makes them less likely to recommend the brand. Humans tolerate shifting messages within an existing emotional attachment; machines do not share that attachment.

Scale offers no protection. Authors writing in Harvard Business Review in 2026 found that Brooks appeared more reliably than Nike in their tests of AI recommendations for running shoes.

Why do distinctive assets alone fail?

Nike illustrates the problem. The Swoosh and "Just Do It" make the brand instantly recognizable to people, yet when AI was asked to recommend a running shoe for a 23-year-old, Nike did not appear in the top three. The AI cited competitors that had built strong associations with emerging buying needs: New Balance on authenticity, Lululemon on wellness. Nike's own narrative, according to the analysis, suffers from semantic contradictions — "Just do it," "Believe in something," "Why do it?" The machine registers the contradictions and declines to recommend the brand regardless of human emotional appeal.

Dove shows a different failure mode. Its real-beauty positioning has built a powerful association with inclusion, yet when an AI assistant was asked to recommend a good cream for sensitive skin, Dove did not appear. Asked why, the assistant associated the brand with affordability. The broader message had not translated into relevance for that specific skincare request.

Research into AI and brand magnetism found that brands need at least three ownable distinctive assets — recognizable visual or verbal cues — plus a clear, relevant narrative to build both mental availability and machine awareness. KitKat's "Take a break" narrative is cited as a clear example: it gives people a reason to reach for the brand beyond wanting chocolate, so the brand surfaces when they want a pause.

How much of AI visibility is long-term brand equity?

Research by Charlie Oscar, reported by WARC, analyzed 30 brands across six categories and attributed 63% of LLM visibility (how often large language models surface a brand) to long-term brand equity. AI pulls from everything, so years of consistent brand building compound. Volvo's decades-long association with safety is cited as the strength of a settled meaning.

The corollary is discipline. A social post may feel fresh in isolation, but if it introduces a conflicting message, it adds confusion to the brand's wider digital footprint. Marketers' impatience becomes expensive: a team may tire of a campaign long before customers have properly registered it. Internal boredom is not a reason to discard an association still under construction.

What must change, and what must not?

The framework is a two-part split:

• Fixed: ownable visual assets and the core narrative. Heinz is the case in point. Years ago, the company discussed removing its keystone logo because it seemed old-fashioned; the surrounding communications could be refreshed instead, without discarding the symbol that made the brand recognizable.
• Flexible: executions. Agencies should receive recognizable assets and a clear meaning to reinforce, then find fresh expressions. Change the shirt, change the earrings, but remain recognizably yourself.

The same discipline applies to promises. Claims must reflect what the business actually does, with evidence to support them. Brands racing into SEO, GEO (generative engine optimization — shaping how AI systems retrieve and present brand information) and Reddit strategies are chasing a moving target; the durable response, the analysis argues, is returning to the core idea the brand stands for and executing it consistently wherever the brand appears.

The new standard of brand craft is dual-audience design: emotional appeal for people, alongside structured information machines can use to validate relevance. Brands that master both early will be harder to displace. The strongest brands give people reasons to remember them and machines clear signals to work with — built together, consistently, over time.