Applying Cognitive Models to Interface Design

Observations on improving user interaction by considering System 1 thinking

March 28, 2025, by Kareem Aboualnaga, Software Engineer @ AWS

Interface design often focuses on guiding users through intended workflows. An alternative approach considers designing for how users behave under conditions of limited attention.

In 2022, I maintained a simple League of Legends champion randomizer tool hosted at kkmet.com/champions. The tool received steady traffic, approximately 1,000 daily visitors. However, engagement metrics indicated brief interactions: users typically landed, used the randomize function once or twice, and exited within about 30 seconds. While functional, I considered if the user experience could be improved.

My initial approach involved adding features, such as filters for excluding champions, role-based filtering, and options specific to competitive play formats. These additions did not significantly alter user behavior; the pattern of brief, focused interaction persisted.

Considering the User's Cognitive Mode

This led me to reconsider the design assumptions. I had been designing for a user engaging in deliberate consideration, corresponding to what psychologists Daniel Kahneman and Amos Tversky termed "System 2" thinking. It seemed more likely users were operating primarily in "System 1" mode – quick, intuitive, and requiring minimal cognitive effort.

This aligns with common web browsing behavior, where users often scan content quickly and interact with the most prominent element related to their immediate goal.

A key realization was that users likely had minimal investment in the website itself. Their objective was to get a champion suggestion quickly and return to their primary activity (playing the game). They were unlikely to invest effort in learning a complex interface.

Redesign Based on System 1 Principles

Based on this understanding, I redesigned the interface focusing on principles aligned with System 1 processing:

  1. Make the primary action highly salient: The randomize button was made large, visually distinct (orange), and centrally located.
  2. Minimize cognitive load: Filtering options were simplified to single-click toggles instead of multi-step selectors.
  3. Reduce attentional demands: Non-essential elements unrelated to the core randomization task were removed.

Observed Results of the Redesign

The changes produced measurable shifts in user behavior metrics:

The reduction in session duration was interpreted positively, suggesting users were achieving their goal more efficiently. The increased feature usage and return visitor rate indicated the redesigned interface was more effective and better received.

Practical Considerations for System 1 Design

Designing for System 1 is not about simplification for its own sake, but about aligning the interface with how people process information under low engagement conditions. Some applicable observations include:

Ethical Considerations

It is important to note that techniques for streamlining user interaction share similarities with manipulative "dark patterns." The distinction lies in the designer's intent. Optimizing for System 1 should aim to help users achieve their own goals more easily, not exploit cognitive biases for outcomes detrimental to the user.

For the champion randomizer, the goal was clearly aligned with user intent (quick results). In contexts involving sensitive actions like purchases or data sharing, careful consideration is needed before heavily optimizing for low-attention interaction patterns.

Applying Learnings in Other Contexts

I have found these principles applicable even in the context of enterprise software development at AWS. Users performing routine tasks often operate with reduced attention. Identifying these workflows and applying System 1 design considerations where appropriate has shown potential for improving user satisfaction without sacrificing necessary complexity.

Concluding Observations

When building utility tools where users prioritize speed and efficiency:

Users of utility tools typically seek results with minimal friction. Designing for this reality can lead to more effective and well-received products.

← Back to Home