The Interactive Behavior Change Model (IBCM) is a behavioral science framework that integrates principles from neuroscience, psychology, and behavioral science into a single comprehensive system for understanding, designing, and evaluating anything built to influence how people think, feel, and behave.
There are many behavioral science and behavior change systems out there, but each was designed for a specific application, whether health behavior change, commercial sales, therapy, or social marketing. The IBCM is the only comprehensive behavior change system specifically designed for interactive technology, making it the most relevant framework for websites, apps, and especially artificial intelligence, whether in the form of chatbots, assistants, voice interfaces, or any adaptive system that responds to user behavior.
Why the IBCM Was Built for Interactive Design and AI
Most behavioral science systems were built around one-way, human-to-human communication. They don’t account for the feedback loops that define interactive technology, and without that, they can’t fully support the personalization, adaptation, and real-time responsiveness that modern products and AI systems depend on.
The IBCM was built around feedback loops from the ground up. This makes it uniquely suited not just for creative design but for fully automated, algorithm-based influence technologies. Every domain maps directly to a component of the communication process, which means the entire system can be implemented in technology, not just interpreted by a designer, but encoded into rules, algorithms, and AI.
The nine domains represent the most mission-critical components of behavior change, presented in an intuitive system that works equally well for absolute beginners finding their footing in behavioral design and seasoned experts building hyper-personalized algorithms or artificial intelligence systems. It also covers the unique psychology of human-brand relationships, human-AI relationships, and human-technology relationships, making it just as relevant for branding and brand strategy as it is for product design and AI development.
How the Model Works
At its core, the IBCM describes the flow of communication between a source and an audience within a social and physical context. The source could be a person, brand, product, or AI agent. The audience is whoever you’re trying to influence.
Communication can be one-way, a static message sent without feedback, or two-way, where the audience responds and the source adapts. Two-way communication unlocks a much wider and more powerful range of influence strategies, because many of the most effective behavior change principles only become available once you can capture and respond to audience feedback.
The Nine Domains of Influence
The model organizes all influence principles into nine domains that represent the key parts of the interactive communication process. What makes the IBCM unique is that each domain covers an entire area of psychology as a coherent set of influence principles, and all nine work together as a system, with each one amplifying the others.
The numbers below correspond to the domains shown in the diagram above.
1. Source
Source credibility is arguably the most important principle in behavior change. Without trust, no other influence strategy will work. This domain covers everything that shapes whether your audience trusts your product, brand, or AI.
Example principles: Expertise, trustworthiness, and charm.
2. Source Expression and Audience Interpretation
How we express ourselves can be more powerful than what we actually say. This domain covers the style and structure of communication, where a well-expressed message can elevate a good idea and a poorly expressed one can undermine it entirely.
Example principles: Cognitive ease, preattentive processing, and visual salience.
3. Media
Different media carry different psychological advantages, and choosing the right channel for the right situation can significantly boost your impact.
Example principles: Text, audio, video, single and multimedia formats, and haptic feedback.
4. One-Way Message
One-way messaging is how we communicate when we don’t have an interactive environment or feedback mechanism. Think of it as the spray and pray approach, where we craft the best possible message for a broad audience without knowing exactly who is receiving it or how they’re responding.
Example principles: Value propositions and calls to action.
5. Audience
This domain covers all the human predispositions that motivate and drive behavior, the psychological targets we aim for when designing for behavior change.
Example principles: Controlling attention, comprehension, motivation, and decision making.
6. Audience Feedback and Source Interpretation
This is the feedback collection mechanism that makes two-way influence possible. Without accurate data capture and interpretation, the most powerful behavior change principles become unavailable, or worse, they backfire.
Example principles: Interaction-based data capture and automated feedback processing.
7. Two-Way Message
This is where the largest and most powerful set of behavior change principles lives. Two-way messaging adapts based on audience feedback, enabling personalized influence strategies that go far beyond what one-way communication can achieve.
Example principles: Conditional reinforcement, personalization, and tailoring.
8. Social Context
Making users aware of the presence and behavior of other people unlocks the power of social influence, one of the most potent forces in human behavior, and one that can be layered onto both one-way and two-way messages.
Example principles: Social norms and social proof.
9. Physical and Virtual Context
This domain brings in yet another layer of influence factors, the constraints of time, space, and resource availability that shape decision making. Because the brain uses the same spatial navigation systems for both physical and virtual environments, the same principles apply in digital products as in the real world.
Example principles: Scarcity and urgency.
Putting It All Together
Example: Exercise App
Imagine an exercise app designed to encourage daily push-ups based on personalized goals. The source is the app itself, positioned as a trusted fitness guide. The audience domain shapes how the app understands each user’s fitness level and motivations. Media includes notifications, videos, and text prompts. One-way messages provide general instructions and broad motivation for all users. Two-way messages kick in when users log workouts, allowing the app to personalize encouragement based on individual progress. Source expression ensures workout plans are communicated clearly and engagingly. Audience feedback captures interaction data to refine future messages. Social context lets users share achievements or compete with friends. And the physical and virtual context makes workouts accessible anywhere, with time-based challenges adding urgency to keep users engaged.
Example: Coaching AI
Now imagine an AI coaching agent designed to support behavior change in a corporate wellness program. The source is the AI agent, which builds credibility over time through consistent, competent, and empathetic responses. The audience domain drives deep personalization, adapting the agent’s approach to each employee’s goals, barriers, and motivational profile. Media spans chat interfaces, email nudges, and voice interactions. One-way messages deliver motivational content and general guidance. Two-way messages enable the agent to respond dynamically, offering encouragement when momentum drops, adjusting goals when users overachieve, and applying sequential request techniques to gradually increase commitment. Audience feedback feeds into real-time adaptation algorithms. Social context incorporates peer benchmarking and team challenges. And the physical context shapes prompts around real-world constraints, timing reminders around commute patterns, workday rhythms, and location-based triggers.
For the latest edition, see:
Cugelman, B., & Stibe, A. (2023). Interactive Behavior Change Model (IBCM 8.0): Theory and Ontology. In International Conference on Mobile Web and Intelligent Information Systems (pp. 145-160). Cham: Springer Nature Switzerland.
Read on ResearchGate