Enhanced design evaluation with Variant Comparison in Maze

The Variant Comparison block streamlines the process of identifying optimal designs through efficient testing and comparative analysis.

Enhanced design evaluation with Variant Comparison in Maze
image: maze.co

The Variant Comparison Block is a new feature designed to enhance design iteration and selection. It facilitates the creation and testing of multiple design variations within a single maze environment. This allows for efficient comparison of results, enabling users to identify the optimal design solution.

video: maze.co

Variant Comparison Approaches: Versioning vs. Multivariate Testing

The Variant Comparison Block facilitates two primary approaches to variant comparison:

  • Versioning: This approach is ideal for evaluating entirely distinct design versions. For example, versioning can be used to assess which user flow offers the most intuitive experience for participants.
  • Multivariate testing: In contrast, multivariate testing focuses on comparing variations within a single design element. This approach might be used to determine which color scheme or call to action (CTA) copy elicits the strongest user response.
maze-usability-testing-comparison-Versioning-Multivariate.png
image: maze.co

Variant allocation strategies: Exclusive vs. Alternating exposure

The Variant Comparison Block offers two primary strategies for allocating users, or exposing them to different design variations:

  • Exclusive comparison: This approach assigns each participant to a single design version. In the field of product research, this method may also be referred to as monadic, between-subjects, or between-groups testing.
  • Alternating comparison: In contrast, alternating comparison exposes all participants to both design variations, but in a randomized order. This strategy is also known as sequential, within-subjects, within-groups, or consecutive testing.
maze-usability-testing-comparison-Monadic-Sequential.png
image: maze.co

The Variant Comparison Block comprises multiple groups of blocks named variants. Each variant is configurable with a blend of blocks to enable diverse tests demonstrating various designs and supporting questions. A maximum of five variants can be integrated into a single Variant Comparison Block.

The results of all variants can be viewed side-by-side. As all variants are integrated within a single maze, responses for each can be observed in one location or exported as a comprehensive CSV file for thorough analysis.