Bold Metrics, the leading AI-powered fit-and-sizing body data platform trusted by global apparel brands and retailers, today announced the launch of its Agentic Sizing Protocol (ASP), a proprietary protocol built specifically for agentic commerce. ASP enables shopping agents to deliver quick, personalized size recommendations directly within the checkout flow. Signaling early enterprise adoption, a global apparel retailer has joined as a launch partner. 

The Missing Piece in Agentic Commerce

AI shopping agents can currently recommend products, compare styles, and check inventory, but the moment a shopper asks about sizing, the experience falls short. 

“AI shopping agents hitting the market right now struggle to answer the most critical question in apparel,” said Morgan Linton, CTO of Bold Metrics. “The moment a shopper asks, ‘What size should I get?’ the  AI agent does not have a clear path to accurately determine the right size. Agentic Sizing Protocol addresses this gap and enables a much more personalized agentic decision process that goes beyond size and captures a shopper’s unique fit preference. In agentic commerce, every second of delay drives abandonment, so performance isn’t optional; it’s critical.” 

How Agentic Sizing Protocol Works

ASP is a lightweight, token-optimized protocol built specifically for AI-native commerce environments. When a shopper needs sizing guidance, the AI agent calls ASP, asks the shopper a minimal set of questions through natural conversation, and receives a personalized size recommendation along with plain-language fit context, such as “just right in the chest” or “slightly snug in the waist,” so shoppers can make confident and informed decisions. The sizing interaction is transport-agnostic, supporting any transport mechanism, from APIs to CLIs to MCPs, easily integrating with emerging AI standards like Google’s UCP.

Enterprise-Grade Security for Multi-Agent Environments

ASP is built with enterprise-ready security standards in mind. Retailer credentials never touch an AI agent and are never exposed to large language models. Instead, brands authenticate through secure, encrypted tokens, allowing retailers to scale agent integrations without compromising control or compliance. 

A New Layer of the AI Commerce Stack

Bold Metrics is introducing ASP as a proprietary protocol designed for the broader commerce ecosystem, enabling AI agents across any platform to integrate sizing capabilities directly into their emerging agentic shopping experiences. Developers can integrate ASP through standard APIs or agent tooling, making it easy to add accurate sizing expertise to AI-driven shopping experiences. 

The protocol builds on more than 14 years of Bold Metrics’ work in solving one of apparel ecommerce’s most persistent challenges: helping shoppers find the right size online. The core technology generates a highly accurate digital twin with 50+ body measurements from only a few inputs. This core technology, combined with garment intelligence, allows shoppers to select their size based on their unique fit preferences. In agentic commerce, where an AI agent might know, for example, that a shopper wants a jacket a little looser in the chest so they can fit a sweater under it, ASP can enable this decision process in a powerful new way.

ASP fills a critical gap in AI-powered apparel shopping by enabling AI agents to quickly recommend the right size while accounting for each shopper's unique fit preference.

More details on Gap Inc.'s role as a launch partner can be found here.