Platform & products

Confidence threshold

Definition

The internal setting that decides whether a Brand Agent answers directly or escalates to a human. Set too low, it guesses too often; set too high, it escalates so much the feature stops saving anyone time.

Defined in What Breaks First: 10 Failure Modes of a Bad Agent Install.

Related terms

Brand Agent: Shopify's merchant-controlled AI shopping assistant, shipped in the Spring '26 edition on June 17, 2026, currently available to Shopify Plus merchants only.
Shopify Knowledge Base: The Shopify-built admin app where you review generated FAQs, read buyer inquiries, and publish custom FAQs that AI shopping agents use when they answer policy, service, and brand questions.
Agentic commerce: Commerce transactions initiated, researched, or completed with an AI agent acting on the shopper's behalf, rather than a human clicking through a storefront unassisted.
Conversation flow: A named pattern of shopper intent — e.g. 'compare two products on a specific attribute' — paired with the reasoning logic and catalog data the Brand Agent needs to answer any phrasing of that intent well, not a single scripted exchange.
Launch drift: The gradual mismatch between an agent's configuration and reality that sets in when nobody actively tunes it after go-live — catalog changes go unreflected, new question patterns go unanswered, and the agent that was accurate on day one becomes gradually less accurate with each week of neglect.
Prose-trapped fact: A true, useful product fact that exists only inside a marketing description rather than as a structured field — readable by a shopper, invisible to an agent's filtering logic.

All 112 terms

Want to know where you actually stand?

Submit your store and get a /100 agentic-commerce readiness score across catalog, schema, product pages and checkout — plus three fixes you can ship yourself.

Get your Readiness Score Agentic Commerce Setup