Agentic commerce: from hype to hard reality
Agentic commerce is the shift from shoppers manually searching and clicking through online stores to AI shopping agents that interpret intent, search across catalogs and complete purchases on behalf of users, turning product data quality and machine-readable context into the real battleground for retail competition.
This new model is no longer a thought experiment; it is colliding with the day-to-day of retail. At a recent gathering of commerce and fashion leaders in New York, the launch of an “agentic storefront” put AI at the centre of the customer journey, signalling how brands expect agents to mediate discovery and purchase decisions. Yet the promise is outpacing execution: shoppers are still overwhelmed by choice and brands still bleed attention and loyalty, with global cart abandonment stuck at around 70 percent for two decades. In other words, agentic commerce is exposing structural problems rather than magically fixing them. If retailers treat agents as another marketing channel instead of a structural reset, they will lose both visibility and trust to those who rebuild around this new reality.
AI shopping agents change discovery, not desire
AI shopping agents are quietly becoming the new front door to retail. Product discovery is shifting from keyword searches and endless scrolling to agents doing that work on a shopper’s behalf. One analysis found that a leading conversational AI accounted for 16 percent of inbound referral traffic to one major fast-fashion retailer over a three‑month period, and 8 percent to two other global brands, evidence that research is moving from search engines to chatbots.
This does not mean shoppers suddenly want different things; it means their path to the same desires is being filtered through machine judgement. When 53 percent of people who use generative AI for search also use it to help them shop, the retailer’s real customer is increasingly an intermediary agent. AI-driven traffic to one large commerce platform’s stores has grown eight times year over year, while orders from AI-powered searches rose nearly 13‑fold, and AI channels now bring in new buyers at almost twice the rate of other channels. The implication is blunt: if your products are invisible or misrepresented to agents, human demand will not save you.
Product data organization becomes a competitive weapon
Agentic commerce retail does not run on brand campaigns; it runs on structured, consistent product data. Retailers are rushing to make listings discoverable by AI shopping agents, but assistants struggle when different listings describe the same item in inconsistent ways. That chaos lives in every catalogue: one merchant lists all flavours of a product under a single entry, another splits them into separate products. Humans can understand this; current agents cannot, unless we teach them.
One major platform’s answer is a system called Catalog, which uses large language models to organize merchant data into a format agents can use to identify and compare products. Catalog groups related listings under a Universal Product Identifier so agents can recognise when different listings refer to the same product and distinguish variants such as flavours or sizes. This is more than a technical upgrade; it is a statement that product data organization is now part of merchandising, not an IT afterthought. Retailers and technology firms are working on standards that make it easier for agents to read listings, and even rethinking site formats so agents can reliably parse product information. Those who do not invest here will watch their products fall out of AI-driven rankings, no matter how beautiful their photography looks to humans.
Operational shock: clean data or invisible products
The hard truth is that AI agents punish messy operations. Because agents depend on consistent, structured product data, retailers are being forced to rebuild creaky back-end systems that were good enough for human browsing but unusable for machine reasoning. This is an operational shock, particularly for fashion and luxury brands that historically prioritised store experience over information architecture.
Executives at the agentic storefront launch framed loyalty as both an emotional bond and a data discipline: knowing who the customer is, listening to what they want and reacting in a timely, personalised way. That is only possible when every SKU, attribute and variant is defined in a way AI can interpret. Beyond tools like Catalog, retailers are reworking how they present information online, experimenting with formats that are easier for agents to read. According to one search and AI visibility leader, client conversations now revolve around building an “agentic catalog” on or off major platforms. The message is clear: if product data remains fragmented, AI shopping agents will misinterpret assortments, undercutting both conversion and brand equity.
AI-powered personalization will decide who wins loyalty
While catalog work feels unglamorous, AI-powered personalization is where shoppers notice the difference. One styling service’s Vision platform uses AI to create outfit imagery tailored to each client’s preferences and has now added a “See it on me” feature that lets shoppers generate images of themselves wearing recommended looks as they browse. Instead of waiting for weekly outfit drops, clients can tap the feature in an on‑demand shopping experience and immediately see a personalised image of themselves in that outfit.
This is what agentic commerce should aspire to: agents that do not only fetch items, but place them in the shopper’s context. The company reports that clients who used the original Vision experience delivered more than a 100 percent lift in spending over 90 days, and its leadership says it is building toward a future where clients see themselves reflected at every step of the journey. Meanwhile, owned AI-powered storefronts promise conversational browsing, virtual try-on and checkout in one controlled experience. Botify plans to launch its own agentic catalog product with large retail brands. Brands that combine clean data, agent-friendly catalogs and emotionally resonant personalization will not just adapt; they will shape how agents “decide” what shoppers see.
