The Intersection

Customer world model

Customer world model is an older term for the customer intelligence layer. It refers to the living, structured, queryable representation of everything a B2B company knows about its buyers.

A customer world model is what customer intelligence becomes at maturity. It is not a list, not a dashboard, and not a pipe. It is a living representation of a company's entire buyer universe, updated continuously as new calls, tickets, and messages arrive. The word model is intentional. A world model carries the implication of a structured understanding that can be queried, reasoned over, and acted on.

A customer world model holds who the buyers are, what they said, how they described their problems, which objections recur, which language patterns predict won deals, and how the answers to each of these questions have changed over time. You do not search a world model. You ask it. The answers come back cited to source evidence and structured so that an AI agent can consume them the same way a human does.

The Amdahl view

We have moved away from the term customer world model in favor of customer intelligence layer. The concept is the same: a persistent, structured layer that every downstream system can read from. The simpler name communicates the same idea without requiring explanation.

What people get wrong

The confusions that come up on almost every first call about Customer world model.

  1. 01

    A customer world model is not a knowledge base. Knowledge bases store documents. A world model stores structured meaning derived from live signal.

  2. 02

    A customer world model is not a CDP. CDPs store events. A world model stores what buyers said and what it meant.

  3. 03

    A customer world model is not a chatbot wrapper on top of Gong. It is a persistent structured layer that every downstream system can read from.

In practice

What Customer world model actually looks like in real product work.

  1. 01

    A GTM leader asks the world model how the top objection has shifted over the last two quarters and gets a diff, cited to the underlying calls.

  2. 02

    An AI SDR queries the world model before writing an outbound email and pulls the exact pain language a matching persona used last week.

  3. 03

    A product marketer asks the world model which of three positioning angles best matches how closed-won buyers describe the product.

Frequently asked

Related terms

See customer intelligence running on your own customer conversations.