The Future of Agentic Commerce

Reducing the friction between AI driven discovery and purchase

Traditional ecommerce asks customers to leave the conversation, navigate to a retailer, and begin the buying journey from scratch. Agentic Commerce shortens that path by allowing customers to purchase at the moment they’ve decided what they want.

Customers expect a seamless path from product discovery to purchase. However, transitioning from ChatGPT recommendations to Best Buy’s website creates a fragmented experience, forcing customers to restart their journey in a new environment and introducing friction at a critical decision point.

Meeting customers in new discovery environments

AI is changing where and how customers begin their shopping journeys. As product discovery moves beyond traditional retail channels, Best Buy has an opportunity to extend its commerce capabilities into new customer touchpoints.

As the Experience Designer representing Best Buy, I was responsible for defining the end-to-end purchasing experience within ChatGPT in the form of an MCP application. Working closely with product managers, engineers, Stripe, and OpenAI, I helped establish the interaction patterns, product scope, and UX strategy for a platform with few existing conventions and shared ownership across multiple organizations.

Designing across organizational boundaries

Although customers experience a single purchase flow, ownership shifts across three independent platforms.

Defining MVP scope & design direction

One of the earliest decisions was whether to recreate Best Buy’s full shopping experience or focus on the moment of purchase. To reduce complexity and validate the core concept, the MVP prioritized a streamlined checkout flow over a comprehensive shopping experience.

The team then worked through a series of scoping sessions to translate the vision into a defined MVP, identifying which capabilities were essential for launch and which would be deferred to future iterations.

With the MVP defined, the remaining work centered on resolving a handful of foundational design questions. These decisions established the interaction model for the experience and influenced how customers would move from product discovery to purchase within ChatGPT.

Because this experience introduced a new commerce model across three organizations, successful delivery required close collaboration beyond the traditional product team. I partnered directly with Stripe and OpenAI throughout implementation to validate assumptions, uncover edge cases, and refine the experience as the technical foundation evolved.

I ran it as a live loop. We worked through design reviews and moved quickly over Slack on the smaller calls. Once Stripe had a test build running on a ChatGPT account, I went through it in context, found where it drifted from the Best Buy system, and gave specific direction on what to change and how.

A single purchase that moves across three companies without requiring the customer to think about the handoffs between them.

Bringing this experience to life required designing not only the interface, but also how commerce should behave inside a conversational environment.

Because this was one of the first implementations of its kind, many interaction patterns were still being defined. I partnered closely with Stripe and OpenAI throughout development, using a dedicated ChatGPT test account to validate real purchase flows and identify gaps between the intended experience and the platform behavior.

This project reshaped how I approach design, largely because I did not own the full system and could not design as if I did.

Designing commerce inside ChatGPT challenged traditional e-commerce patterns and required the team to rethink how customers discover, evaluate, and complete purchases in a conversational environment.

The biggest lesson was learning how to design responsibly within emerging technology. By partnering with Stripe and OpenAI, I learned the importance of understanding constraints early, making thoughtful tradeoffs, and protecting customer trust when the ideal experience is not always technically possible.

This project reinforced that great experiences are created through partnership, bringing together customer needs, business goals, and technical realities to create something intuitive and reliable.