Skip to article frontmatterSkip to article content
Site not loading correctly?

This may be due to an incorrect BASE_URL configuration. See the MyST Documentation for reference.

What Is a Marketing World Model?

Subtitle: Entities, beliefs, clocks, and the art of building worlds you can run — starting with the marketer that is you.


“We think we design messages. Often we only design moods — until we specify what changes, for whom, and when.”
— adapted from Herbert A. Simon


Opening provocation

Marketing fails in two opposite ways: vague story (no mechanism) and narrow metric (mechanism without meaning). A marketing world model is the middle path: a computable story about who attends, what they believe, how messages move, and what you measure.

In AINS-M6001, the smallest honest world is your own influence system: attention, credibility, clarity, and effort — before you pretend to model “the market.”


1. A marketing world model in four parts

  1. Entities — audiences, messages, channels, assets (including you as marketer).

  2. Properties — attention stocks, trust, message vectors, costs.

  3. Rules — how signals propagate, how trust updates, how effort maps to reach.

  4. A clock — tick size (day, week, campaign flight).

If the skeleton is vague, “insights” are theater.

Key concepts


2. Ontology: what exists in this world?

Ontology answers: what kinds of things are real here?

2.1 Design choice, not discovery

You choose fidelity vs parsimony vs computability. A personal marketing ontology might include Marketer, AudienceSlice, Message, ChannelEffort. It might omit Mood in v1 until rules justify it.

2.2 Boundaries

Outside the boundary: macro trends, platform algorithm details, luck. You may treat them as shocks or constants — but say which.

Key concepts


3. Properties: where meaning becomes measurement

Examples for Marketer in v1:


4. Rules: what updates when the clock ticks?

Illustrative rule families:


5. What this course will not do


6. Bridge to the notebook

01_first_marketing_world.ipynb implements a minimal Marketer and a toy reach rule so you can run something on day one. Your job is to own the ontology choices and list what is missing.


Further reading (optional)


Lecture checklist