Speedfactory Messenger Bot
Commerce through a chat.

The AM4LA was part of adidas's Speedfactory city series — shoes engineered around local terrain, climate, and running patterns. Los Angeles got a shoe made with Parley Ocean Plastic yarn, ocean-colored upper, full-length Boost midsole, and an NFC chip embedded in the tongue.

The strategic logic: same-day urban delivery as a competitive advantage for an exclusive drop, with a conversational channel handling the full transaction.

The chatbot handled the full purchase flow inside Messenger. Invitation, product introduction, size selection, payment, confirmation, delivery coordination — all within the conversation thread. No redirects. No external checkout. The consumer never left the app.

Messenger wasn't a marketing wrapper around a traditional checkout experience. It was the checkout experience.

On the back end: coordinated same-day fulfillment and last-mile delivery across the LA area.

Consumer education was a real cost. In 2018, buying something through a chatbot was not intuitive. Operations don't care about your elegant front-end — coordinating same-day delivery across a metro area while managing size inventory in real time is a logistics problem. And measurement was complicated in a closed platform.

• 1 daylast-mile delivery speed
• 0→1proved conversational commerce could drive real transactions
• 50%+more than half of invited customers went on to purchase in the pilot

The platform architecture that routed purchase intents could also route support requests and order inquiries. What started as a limited LA pilot became the validated foundation for adidas's enterprise customer service chat.

In 2018, same-day urban delivery was an experiment. By 2021 it was an arms race. Conversational commerce followed a parallel arc. What we called a chatbot is now being rebuilt with large language models and called an agent — and the underlying premise hasn't changed at all.