Emotional Design: Don Norman's Three Levels
Don Norman proposes three levels of design: visceral, behavioral, and reflective. The evidence behind them, how to apply them, and where they turn manipulative.
What are the three levels?
The visceral level is the first impression. It works through the senses: shape, color, texture, sound. Interaction Design Foundation describes it as the first reaction to a design; its example is that an uncluttered interface suggests ease of use (Interaction Design Foundation, “Emotional design”).
The behavioral level is the experience of use. It is the assessment of how much and how easily the design helps someone reach a goal. The person needs to feel in control with minimal effort.
The reflective level is the conscious judgment formed after use: whether the product is worth what it costs, whether it represents who the person wants to be, whether they would recommend it to a friend. Brand, story, and identity live here. Interaction Design Foundation describes it as the conscious judgment that keeps people using a product and forming emotional bonds (Interaction Design Foundation, “Emotional design”).
Why should a designer care?
Because it gives you a vocabulary for discussing what isn’t measured in clicks. When someone says “the app feels cold”, you can ask whether they mean the visceral level (colors, typography), the behavioral level (how the app responds to actions), or the reflective level (what the brand promises and doesn’t deliver).
It also helps you prioritize. A product can have flawless usability and still create no bond. Another can have usability flaws and be deeply loved. The three levels show you where the problem is.
What evidence is there?
The argument most cited in the book is that attractive things work better. The study most often used to support it is by Noam Tractinsky, published at CHI 1997. In the first experiment, 104 engineering students at an Israeli university rated 26 ATM layouts (created in Japan by Kurosu and Kashimura and translated into Hebrew) on two scales, one for apparent usability and one for aesthetics. The correlations were very high: 0.921 in that experiment, against 0.59 in the original Japanese study (Tractinsky, “Aesthetics and apparent usability”, CHI ‘97).
The author himself notes the limits. Usability was judged before the ATMs were used, so it was an impression, not a measurement. There was a possible method bias. And aesthetics was defined narrowly, as visual beauty. Cite the study with those qualifications, not as proof that the beautiful is always more usable.
How do you apply it?
- Visceral. Use color, typography, and space with intent. A visually quiet interface conveys calm; a saturated one conveys anxiety. Check that the first impression matches what the product actually does.
- Behavioral. Care about the response to every action. A clear loading state, a visible confirmation, and an error that explains what happened create a sense of control. Interaction Design Foundation stresses that the user should feel in control with minimal effort.
- Reflective. Be consistent with the brand’s promise. The tone of an error message or a welcome email also builds a relationship. Honest storytelling, with real people, helps more than a slogan.
A practical detail: the tone of an error message is emotional even when it looks technical. A cold “Something went wrong” creates distance; one that explains what happened and what to do next builds trust.
Example
Illustrative example: a personal finance app has a very cold blue interface and a highly technical tone in its messages. People find it efficient but distant. The team reviews the three levels. At the visceral level, it shifts the palette to warmer tones and adds simple illustrations to empty screens. At the behavioral level, it turns errors into messages that explain the next step. At the reflective level, it writes the brand promise on the welcome screen and keeps it in the first monthly summary. The change doesn’t improve clicks. It improves people saying the app “gets” them.
When NOT to apply it
- When emotion replaces function. A beautiful product that doesn’t solve the problem doesn’t have emotional design; it has decoration.
- When it is used to push decisions. Guilt, urgency, or fear used to sell crosses the line into dark patterns.
- When the audience is very diverse. The same color or image doesn’t carry the same meaning in every culture. Test with the people who will use the product.
- When the team uses it as an excuse not to measure. Emotion can be researched too, with interviews, surveys, and observation.
Common mistakes
- Confusing emotional design with decoration or with illustrations everywhere.
- Assuming beauty compensates for poor usability over the long run.
- Using an emotional tone that doesn’t match the brand or the context.
- Designing the reflective level without keeping what the brand promises.
- Forgetting the cognitive biases that shape what people say about how they feel.
How to start tomorrow
Pick a screen in your product that people describe as “cold”, “confusing”, or “boring”. Write down what each level says: what you see on entering, what happens when you use it, and what the brand promises. Change one thing at each level and ask five people to describe the product in three words before and after. If the words change, the emotional design is working. If they don’t, check whether the change was real or just visual.