Opinions expressed by Entrepreneur contributors are their own.
Key Takeaways
- Consumers are now using AI tools to research, compare and narrow options before they ever visit a retailer’s website.
- Purchase decisions are often being made during the AI conversation itself, not on the checkout page.
Have you noticed how the moment of choice for customers has evolved over the years? Traditionally, the journey began with a Google search, followed by a visit to a retail website where consumers explored products and made purchasing decisions. Today, much of that process is being delegated to artificial intelligence.
Instead of browsing categories, consumers are asking AI models to search, filter, compare and recommend products based on highly specific preferences and past interactions. However, despite this rapid transformation, the AI landscape still faces several challenges and limitations.
In March 2026, OpenAI shut down its ‘Instant Checkout’ feature just six months after launch. Despite Shopify’s vast merchant network, only a dozen businesses went live with the feature. While users research products a lot within ChatGPT, they happen to exit the platform to complete their purchases elsewhere.
Calling this a product failure is a misinterpretation of the data. What OpenAI actually identified is a structural split in the modern buying journey. In this article, I’ll discuss how businesses must restructure their funnels to bridge the gap between AI discovery and the final transaction.
What actually changed in how people buy
According to Adobe Analytics, traffic to US retail sites from generative AI tools grew by nearly 693% during the 2025 holiday season. More importantly, these visitors converted at a rate 31% higher than any other source. This shift is becoming the new standard. Gartner predicts that traditional search engine volume will drop by 25% by the end of 2026 as chatbots replace queries that previously went to Google. People are using AI to research and narrow the options, but they aren’t buying inside the AI, they just use it as a new starting point.
The world’s largest retailers have already begun adapting to this ‘discovery-first’ reality after finding that forcing transactions within AI often backfires. Walmart found that conversion rates inside ChatGPT were three times lower than when users were redirected to Walmart.com.
Instead of abandoning AI, they changed the approach by creating their own shopping assistant within AI platforms. This way, they can help customers right when they start searching while still keeping them connected to the Walmart brand. Sephora launched its ChatGPT integration in March 2026 with a focus on personalised discovery rather than immediate checkout. By linking their loyalty profiles, customers get suggestions based on their actual skin type and shopping history.
Shopify activated ‘Agentic Storefronts’ by default, ensuring millions of products are discoverable to Microsoft Copilot and other models, while keeping the final checkout securely on the brand’s own website. Your website is no longer where you convince the customer, but the place where they go to place an order they’ve already decided on.
The mistake most ecommerce businesses are making
According to research from Attest, nearly half of all consumers use AI to research their purchases, and 40% trust AI results more than traditional search links. The problem is that lots of businesses invest money in the wrong point of the funnel, testing their checkout pages or changing the pricing strategies, still ignoring the stage where the customer is actually making up their mind. If an AI cannot find your products or services and make a compelling selling summary out of it, you’ve lost the customer before they even see your deal.
Data from OpenAI and Adobe proves that while shoppers prefer the security of a familiar checkout over buying inside a chat, AI-driven referrals are now outperforming other traffic sources. The challenge is in how your website can pick up the conversation the AI started. If the landing experience doesn’t match the context they just left, you’re likely to lose the orders.
How to deal with this new funnel
With traditional search traffic dropping by 22% as AI suggestions win over standard clicks, businesses must run a quick diagnostic on AI visibility. If your brand doesn’t work properly with the potential the AI offers, it’s time to work on your product data.
Generative Engine Optimisation (GEO) is becoming as essential as traditional SEO for those businesses that want to keep growing on the AI wave. The goal is no longer just selling through AI, but ensuring you appear correctly within it.
Leading brands are already seeing a 150% increase in citations by making their products understandable to AI. It’s all about working with clear product descriptions, its features, and usage scenarios. Whether it is through Shopify’s ‘Agentic Storefronts’ or Perplexity’s ‘Buy with Pro’ partnership with PayPal, the aim is to bridge the gap between discovery and purchase. Structure your data so models can parse it, and ensure your landing pages pick up exactly where the AI conversation left off.
What competitive advantage looks like
With nearly a third of the US population expected to use generative AI search by the end of 2026, the winners will be the businesses that can shape a consumer’s preferences early on. If you make the jump from AI research to the final checkout feel easy, you’re making an advantage that is hard for competitors to beat, as you move based on a deep understanding of the new way people like to shop.
In this landscape, price becomes a less decisive factor, as the buyer has often made their choice before the traditional comparison-shopping even begins.
The long-term benefits of this strategy go far beyond simple sales volume. Sephora’s integration of loyalty data into AI recommendations is expected to cut return rates by 30% simply by ensuring a better match between products and skin types.
Key Takeaways
- Consumers are now using AI tools to research, compare and narrow options before they ever visit a retailer’s website.
- Purchase decisions are often being made during the AI conversation itself, not on the checkout page.
Have you noticed how the moment of choice for customers has evolved over the years? Traditionally, the journey began with a Google search, followed by a visit to a retail website where consumers explored products and made purchasing decisions. Today, much of that process is being delegated to artificial intelligence.
Instead of browsing categories, consumers are asking AI models to search, filter, compare and recommend products based on highly specific preferences and past interactions. However, despite this rapid transformation, the AI landscape still faces several challenges and limitations.
In March 2026, OpenAI shut down its ‘Instant Checkout’ feature just six months after launch. Despite Shopify’s vast merchant network, only a dozen businesses went live with the feature. While users research products a lot within ChatGPT, they happen to exit the platform to complete their purchases elsewhere.