Hook Model

What is the Hook Model?

The Hook Model, also called the Hooked Loop, is Nir Eyal’s framework for repeated product engagement from Hooked: How to Build Habit-Forming Products. It proposes a cycle of trigger, action, variable reward, and investment. Eyal’s explanation describes the framework; our Hook Model overview examines it in more detail.

What are the four stages of the Hook Model?

Trigger

A cue prompts interaction. Eyal distinguishes external prompts, such as notifications, from internal cues, such as an emotion associated with using the product.

Action

The user does something, such as opening an app or checking a feed. The model emphasizes making the action easy to perform.

Variable reward

The action produces a rewarding result with some uncertainty: new content, a response from someone else, or another outcome the user values. Variable rewards are part of Eyal's proposed loop; they are not a universal requirement for habits.

Investment

The user contributes something intended to support later engagement, such as saved information, preferences, or effort. A customized product may offer a reason to return.

What does “hook” mean in psychology?

In this context, “hook” is a product-design metaphor for drawing someone back. It is not a separate scientific diagnosis or a general law of attention.

Is the Hook Model scientifically valid?

The model’s appeal is its simplicity. That is also the problem: a description can start to sound like an explanation. Finding a trigger, action, reward, and investment in a successful app does not show that assembling those ingredients will produce another successful app.

Habit research concerns learned responses in recurring contexts. Returning to a useful product, enjoying it, and using it automatically are different things. Evidence about rewards does not validate the entire sequence.

Two articles develop this objection: Hooked: How to Build Habit Forming Products Is Wrong and An Incomplete Loop. Start with the problem your product solves and the reasons someone would use it again. The diagram cannot supply those reasons.

For related approaches, see the Behavioral Design overview, Behavioral State Model, and Fogg Behavior Model. Each still needs to be judged by how well it helps solve the actual problem.