The new purchase journey: How AI is rebuilding discovery and buying
Salesforce estimates that AI tools and agents drove 20% of global retail sales during last year’s holiday season. AI Overviews already appear in large part of Google searches. ChatGPT alone is dealing with tens of millions of prompts per day. This is not a forecast. It is already happening, and the numbers are climbing.
We have seen shifts like this before. Mobile phones were just the beginning. First they were easy to carry around and a bit of a luxury item. By the time smartphones were introduced, there were added capabilities that have made them absolutely necessary for everyday life. AI is doing the same to shopping: discovering, comparing, and now buying, are all being pulled inside a single conversation.
AI Search
We are used to typing search terms on Google and getting pages of blue links to click through. The beginning is still very much the same but now Google hands back an answer with context. It is built from your query, location, search history, and your likely intent. Your initial search might fan-out to multiple sub-queries about price points, reviews, and store locations to name a few. The catch for brands is that showing up in search results doesn’t necessarily end up being clicked at all. Your product can be the correct answer and still send no one to your site.
Conversational Shopping
The mechanics of this are well-known: you walk in to a brick-and-mortar shop and tell the salesperson what you are looking for. Based on their expertise, they recommend options for you to compare and choose from. An AI assistant does the same. It compares options, reads reviews, and suggests the best options. Except now it’s also deciding on your behalf whether a given brand gets mentioned at all. The assistant should not be treated as a new tool or channel. It’s a new gatekeeper for commerce.
Agentic Checkout
A product comes up in whatever app or platform the shopper is already in. The user clicks to buy and never visits the actual brand store. Agentic checkout removes the friction of moving away from the platform, landing on a new web page, spending time to look for the right category and product, just to find out that the product is out of stock. Agentic checkout shortens the buying journey, drives immediate action, and transfers customer data ownership from ecommerce to platforms. Ecommerce brands risk becoming one of many product catalogues for AI.
Creator Discovery
Human reviews and recommendations have always played a big part in brand preference and discovery, whether they come from friends, family, or influencers. Social media platforms are built around this concept. Social channels drive unplanned discovery combining entertainment with sales impact. You end up wanting products that you didn’t even know that existed. The path from doomscrolling content to one-click buy and a confirmation email is short. Especially for younger audiences, creator discovery is becoming the way to find new products.
None of these four shifts is happening on its own. The space between wanting something and buying it is narrowing, and it’s narrowing on infrastructure that the brand doesn’t control. The purchase journeys have been rebuilt and they are moving on to platforms that brands are not owning.
The winners will be the ones who start asking new questions. Are we visible inside the answers? Do we have enough authority to get recommended? Are we present when the discovery starts? And when the sale happens, do we still own the conversion?