When AI Shops for Us, What Do Brands Lose?

Google is turning Gemini into a storefront. By letting people shop and check out directly inside Gemini, partnering with giants like Walmart, Google has solved the convenience problem. For consumers it means fewer clicks and less friction between discovery and purchase.

But for retailers this creates a massive blind spot. When artificial intelligence becomes the interface for shopping, the customer journey disappears into a black box. You see the purchase, but you lose the why.

The LLM Shopping Revolution 

We're in the middle of a massive, structural shift in how people buy. Large language models (LLMs) aren't just for research anymore. They're becoming the actual storefront, moving from the discovery and research phase all the way through to the transaction.

And the data we’re seeing at Contentsquare prove this isn’t “the future,” it’s already happening. While AI-referred traffic still accounts for just 0.2 percent of total traffic, last year we tracked a +632 percent spike in traffic, showing consumers are turning to generative AI tools to research products and narrow down their choices before checking out.

Meanwhile, traditional organic search traffic has declined 9 percent year-over-year (YoY) as AI Overviews increasingly answer questions without requiring a clickthrough. The data shows that the visitors who do arrive from AI sources are coming with significantly stronger intent. In fact, conversions from AI-referred traffic rose 55 percent YoY.

In a world where 30 percent of consumers say they would allow an AI agent to complete a purchase on their behalf, brand influence doesn’t disappear. It moves upstream toward shaping user intent earlier and ensuring relevance within the systems making decisions. And while many internal model interactions are invisible to brands, where brands can create visibility is inside their own AI experiences — e.g., brand-owned ChatGPT apps within those environments.

When you capture the journey happening inside your ChatGPT app — the initial queries, follow-up questions, and the AI’s personalization — you’re tapping into a powerful feedback loop and the data you need to optimize your product, messaging and customer experience.

New Interfaces, New Opportunities

This is where announcements like Google’s become critical. The same goes for the growing ecosystem of ChatGPT apps and other LLM-native commerce experiences.

These platforms aren't just threats; they're a reflection of how consumers want to shop and they represent a huge opportunity for brands to own a piece of the AI-driven convenience economy.

Measure, Optimize, Win

So what should brands actually do? Treat LLM-driven commerce exactly like they treated social media when it emerged as a new channel — measure it, optimize for it, and turn it into a growth engine

This means three things:

1. Focus on what you can actually measure. Just like search, what happens inside LLMs remains largely invisible to brands unless platforms choose to expose it. However, there are tools that monitor specific prompts and brand visibility within LLMs. These are good proxies for this information.
2. Create, optimize and measure brand-owned AI experiences — e.g., chat interfaces on your site, in your app, or embedded as apps within LLMs — where you can understand how customers ask, refine and act, and where journeys can be understood in the context of entry points and customer intent.
3. Optimize deliberately. Just like brands learned to optimize for Facebook and Instagram algorithms, they now need to optimize for how LLMs surface and recommend products. That means investing in generative engine optimization (GEO) alongside traditional SEO. It means structuring product information so AI can parse and present it accurately. Make sure you know which AI platforms are driving qualified visitors, how these visitors behave differently than traditional search or social traffic, and what their conversion patterns look like. It means continuously testing and learning, because these platforms are evolving fast.

The brands that figure this out early will lead the next era of commerce. The ones that ignore it will wake up one day with a very limited understanding of customers, declining visibility, and no clear path to improvement. AI shopping works in theory. But what's at risk is what gets hidden along the way — and the only way to protect that is to get inside these new experiences and measure them relentlessly.

Rachel Obstler is the chief product officer at Contentsquare, an all-in-one experience intelligence platform.