Shared commerce language

Give agents a map
of commerce meaning.

A shared data vocabulary that helps humans and AI agents understand ecommerce entities, relationships, and decisions.

Get a free commerce data audit Ask iKawn to map this workflow
Shared language

A shared language for products, orders, customers, campaigns, and returns.

Definition

A commerce ontology is a structured vocabulary and relationship model for ecommerce data, enabling AI systems to understand how products, customers, orders, returns, campaigns, and decisions connect.

Why It Matters

  • Agents perform better when commerce data has shared meaning instead of isolated field names.
  • A clear ontology reduces ambiguity across teams and systems.
  • Structured relationships make predictions, recommendations, and audit trails easier to explain.

How It Works

01

Define core entities such as Product, Variant, Customer, Session, Cart, Order, Return, Campaign, Creative, Agent, Policy, and Decision.

02

Map relationships such as customer placed order, order includes variant, variant caused return, campaign promoted product.

03

Attach metrics and events to entities so agents can reason over outcomes.

04

Use the ontology to power internal links, schema, search, recommendations, and agent workflows.

Examples

  • A return-risk model connects a campaign, SKU variant, customer segment, and return reason.
  • An agent understands that a creative asset belongs to a campaign promoting a product family.
  • A decision log records which agent recommended an action, what evidence was used, and who approved it.
  • A topic page can be generated from entity relationships instead of isolated keywords.

iKawn Framework

Entities

the nouns of commerce.

Relationships

how entities affect each other.

Signals

events, metrics, and observations.

Decisions

recommendations, approvals, actions, and outcomes.

FAQ

What is a commerce ontology?

It is a shared vocabulary and relationship model for ecommerce data.

Why does ontology matter for AI agents?

Agents need consistent meanings for products, customers, orders, returns, campaigns, and decisions.

Is this the same as structured data?

Structured data describes pages for search engines. A commerce ontology describes business entities and relationships for internal intelligence and agents.

Where should teams start?

Start with products, variants, orders, customers, returns, campaigns, creatives, policies, and decisions.