AI Can Own Discovery. Brands Still Need to Own the Customer Relationship.

The most important shift in AI shopping may not be what marketers think.

The story is not simply that consumers are asking ChatGPT, Gemini, and other AI tools what to buy. The bigger change is that product discovery is moving onto platforms brands do not own, while the transaction, customer data, and long-term relationship are still very much up for grabs.

Reuters reported this week that retailers including Walmart, Ulta Beauty, and Wayfair are actively optimizing their sites to show up in AI-generated shopping recommendations. At the same time, they are pushing to keep the actual transaction on their own properties, where they can capture browsing behavior, purchase history, loyalty activity, and the data that powers future marketing.

That distinction matters. AI may become one of the most important acquisition layers in commerce. It does not automatically have to become the owner of the customer relationship.

AI Is Becoming an Acquisition Layer

The early numbers suggest AI-referred traffic is not just curiosity traffic. Adobe Analytics data cited by Reuters found that 41% of U.S. consumers used generative AI for online shopping in June, and visitors referred by AI services generated 41% higher revenue per visit than shoppers arriving through traditional channels.

Ulta Beauty says shoppers finding its products through Gemini and ChatGPT are showing roughly double the conversion and intent. The company is even working with Google to bring shopping carts and its rewards program into AI-powered shopping experiences.

For marketers, this starts to look familiar. We have spent decades dealing with acquisition channels that sit between the brand and the customer. Search engines, social platforms, affiliates, marketplaces, and app stores all became powerful gatekeepers. AI shopping assistants are shaping up to be the next version of that dynamic.

The difference is that AI can compress far more of the consideration process into one interaction. A customer may no longer browse 20 product pages, read a dozen reviews, compare brands, and then decide. They may simply tell an AI agent what they want and receive three recommendations.

That can make AI extraordinarily valuable for acquisition, but it also creates a much bigger question about who owns what happens next.

Discovery and Ownership Are Two Different Things

A lot of the conversation around AI commerce focuses on visibility. How do brands get recommended? How should product pages be structured? What information do AI models need? What replaces traditional SEO when the customer is asking a chatbot for an answer instead of scrolling through search results?

Those are important questions, but lifecycle marketers should be asking another one: What happens after the recommendation?

If an AI platform sends a customer to your site and you complete the transaction, you have the opportunity to turn that moment into a relationship. You can capture consent. You can understand the order. You can onboard the customer. You can connect the purchase to a loyalty account. You can personalize future communication. You can learn whether the customer buys again.

If the entire transaction happens inside someone else's ecosystem, much of that value may disappear or become dependent on the platform's rules.

That is why the fight over checkout is really a fight over customer ownership.

Why Lifecycle Marketing Becomes More Valuable

If AI makes discovery easier and acquisition more efficient, the value of the post-purchase relationship goes up, not down.

The brand that wins is not necessarily the brand that gets the first AI recommendation. It may be the brand that takes an AI-referred first purchase and creates enough value that the customer no longer needs an intermediary the next time.

That puts more pressure on the parts of marketing that have historically been treated as downstream execution: onboarding, CRM, loyalty, retention, customer service, subscriptions, replenishment, and personalization.

These become strategic defenses against platform dependency.

If a customer comes through ChatGPT once and then purchases directly from your brand for the next five years, the AI platform helped you acquire a customer. If that same customer goes back to an AI agent before every purchase and the AI chooses among you and five competitors each time, you never fully earned the relationship.

The New Funnel Starts After the Click

The traditional funnel taught marketers to obsess over everything that happened before conversion. AI could invert some of that thinking.

When discovery becomes increasingly automated, one of the most important competitive advantages may be what a brand does immediately after the first transaction.

• Can you identify the customer and earn permission to communicate?

• Can you make onboarding useful enough that the customer understands the product and gets value quickly?

• Can you connect the purchase to loyalty, membership, subscription, or another reason to return directly?

• Can you use first-party behavior to make the next experience better than a general-purpose AI recommendation?

• Can you create enough trust and habit that the customer thinks of your brand first next time?

Those are lifecycle questions, and they are becoming acquisition questions too.

What Brands Should Do Now

There are five things I would be thinking about now if I owned CRM, lifecycle, or retention for a consumer brand.

1. Treat AI referrals as a distinct acquisition source. Measure conversion, AOV, repeat rate, retention, and LTV separately. The acquisition economics may look very different from paid social or search.

2. Protect the transition from anonymous visitor to known customer. Identity capture and consent will matter even more when the discovery experience happens somewhere else.

3. Build onboarding for AI-referred customers. Someone who arrives after a chatbot recommendation may have a different level of product understanding than someone who spent 20 minutes browsing your site.

4. Give customers reasons to return directly. Loyalty, replenishment, subscriptions, saved preferences, exclusive benefits, great service, and genuinely useful content all reduce the need to restart the discovery process elsewhere.

5. Measure relationship ownership, not just conversion. The most interesting metric may eventually be the percentage of AI-acquired customers whose next purchase happens directly with the brand.

The Bigger Strategic Question

For years, brands have worried about becoming too dependent on Amazon, Google, Meta, and other platforms. AI commerce does not eliminate that problem. It creates a new version of it.

The good news is that the answer is not to avoid AI. The traffic appears valuable, consumers are clearly experimenting with it, and brands should absolutely make themselves easy for AI systems to understand and recommend.

But getting recommended is only the beginning.

The strategic goal should be simple: let AI help introduce the customer, then build a relationship strong enough that you do not need to be introduced again.

That is where lifecycle marketing becomes much more than a retention function. It becomes the system that turns rented discovery into an owned customer relationship.

Source: Reuters, August 7, 2026, “Retailers tap AI shopping traffic but fight to keep customer data.”