AI Is Rewriting The Rules Of Brand Loyalty

AI is becoming the starting point for shopping journeys, and that could have significant consequences for brand loyalty. As consumers increasingly turn to AI assistants and conversational search to discover products, they are being introduced to a much wider range of brands. Many are subsequently buying from companies they had never previously considered, challenging the traditional role of brand preference in purchasing decisions.

It is also changing the purpose of brand websites. By the time shoppers arrive, AI recommendations may have already shaped their decisions, leaving marketers to reconsider where and how they invest to influence consumers.

According to DEPT's How AI Is Changing What People Buy report, 61 per cent of US shoppers had used AI assistants within the previous three months. Meanwhile, 56 per cent of shoppers using AI search reportedly converted to a brand that was entirely new to them.

The implications go beyond discoverability. AI could also make it easier for brands to find highly specific audiences, creating opportunities for more niche products.

“If people are coming to a search interface with very long prompts because they've got a very specific niche, then suddenly, as a brand, you can make and define products which fit niche audience needs and actually find your market,” outlines Isabel Perry, chief strategy officer and global executive vice-president of strategy at DEPT.

A rethink for brands

If consumers arrive on a website having already completed much of their consideration journey through an AI assistant, the role of paid media changes. AI may increasingly capture existing demand, but brands still need to create it. That means building awareness and preference before consumers begin asking for recommendations.

Perry reveals that DEPT has seen a sharp increase in briefs focused on AI discoverability. Yet the objective should not simply be to secure a recommendation from an AI assistant. It should be to ensure consumers already have a brand in mind. After all, there is a considerable difference between a shopper asking AI which product to buy and asking where to buy a specific brand.

That distinction could gradually shift marketing investment back towards brand-building rather than an overwhelming focus on performance marketing.

At the same time, websites and product information are becoming increasingly important. AI systems can evaluate products against detailed specifications, customer needs, and available data, potentially giving an advantage to brands with stronger functional credentials. This is particularly relevant to considered purchases such as cars, where practical requirements may outweigh emotional appeal.

“I often use this example of I just bought a car, and I just bought it entirely from a conversation with Claude. It was just the top recommendation for my very specific needs about boot size and customer ratings, and price, meaning I might have bought Skoda. It's like a very LLM car. It's like not sexy, but it has got great specs,” outlines Perry.

Brands that have become synonymous with their categories, such as Hoover, Sellotape, Uber, and Google, may already have an advantage. Their names are so embedded in everyday language that consumers often use them as shorthand for the products or services they represent. But Perry believes brands must also prepare for three distinct AI-driven discovery environments.

Traditional search is becoming more conversational while continuing to present results. Chat assistants allow consumers to explore products through longer, more detailed exchanges. Emerging personal assistants go further still, acting on the behalf of users by navigating websites and completing tasks. Each presents a different challenge for brands seeking to influence the purchasing journey.

Confidence in AI commerce is growing, but trust limited

Despite growing enthusiasm for AI-powered shopping, consumers remain reluctant to hand over control of their money. DEPT's research found that just 7 per cent of respondents would allow an LLM to complete a purchase on their behalf.

“People are extremely comfortable delegating the choice, but not the transaction,” Perry explains.

That distinction matters as brands develop their agentic commerce strategies. Although consumers are becoming comfortable relying on AI for recommendations, allowing an assistant to complete a transaction remains a bigger leap.

Perry points to Amazon's introduction of one-click checkout as an example of how purchasing behaviours can evolve. What initially feels unfamiliar can become routine when the convenience is compelling enough. She expects a similar progression with AI-assisted transactions, although widespread adoption may still be some way off.

“People do trust the LLMs, and that's why they are delegating the shopper journey to it because they trust its recommendation. They just don't trust the final transaction,” Perry admits.

One brand already taking AI discoverability seriously is hardware retailer Logitech. Recognising that consumers have specific requirements for products such as mice, keyboards, and gaming equipment, but may struggle to distinguish between them visually, the company has been preparing its content for AI-driven shopping.

According to Perry, Logitech established a centre of excellence, identified gaps between customer queries and existing content, and introduced FAQs to answer previously overlooked questions. It also developed a generative engine optimisation (GEO) strategy spanning social and creator content, while exploring conversational commerce through Amazon's Rufus shopping assistant. The approach illustrates how brands can improve their chances of appearing in AI recommendations by making their products easier to understand and compare.

“You need to be clear about your product reality and the values that people have. What is the differentiation that people should care about, and those are the prompts that you want to win if you can't win the broad category because you're not Sellotape,” she advises.

AI commerce's unexpected power users

While younger consumers are often assumed to be leading the adoption of emerging technologies, DEPT's findings suggest a more surprising picture. According to its report, consumers aged 45 to 60 emerged as power users of AI commerce, challenging the assumption that younger generations are always the earliest adopters.

Perry says DEPT has also been working with clients to understand which parts of their customers' journeys remain human-led and which are increasingly delegated to AI.

In one project involving an airline, business-class travellers emerged as particularly quick adopters, suggesting that AI usage may be influenced as much by audience needs and purchasing categories as by age.

Another striking finding was the extent to which AI recommendations could disrupt established brand preferences. Perry describes seeing “extreme loyalty changes” among consumers who followed LLM recommendations rather than choosing their usual brands. It raises a difficult question for marketers: how much does brand loyalty count when an AI assistant presents a seemingly better alternative?

Preparing for the AI shopping rush

With Black Friday approaching, Perry warns that brands cannot afford to overlook the practical work required to improve their visibility in AI-driven shopping. Although some content can be picked up by AI systems within days, a comprehensive GEO strategy takes longer to deliver results.

In the short term, businesses should ensure their websites are accessible to relevant AI agents and that outdated firewall settings are not inadvertently blocking them. Product content should also be treated as an essential part of a brand's digital infrastructure, rather than simply material designed for consumers browsing a website. That means clearly communicating promotions, discounts, stock availability, and Black Friday offers in language that reflects how shoppers are likely to search.

Accurate product feeds and inventory data are equally important, alongside the technical foundations that make information easier for AI systems to interpret. Clean semantic HTML, properly structured headings, lists, and tables can all help. Perry also warns against burying valuable product information in PDFs or JavaScript-heavy experiences that may be difficult for AI systems to access. Making that information available on accessible web pages can offer a more immediate opportunity to improve discoverability.

For marketers, the challenge is becoming two-fold: making products easier for AI to find and recommend, while giving consumers a reason to ask for their brand in the first place. As AI takes on more of the shopping journey, being discoverable will matter. But being wanted may matter even more.