GeneyMAP: Exploring the Potential of GenAI to Facilitate Mapping User Journeys for UX Design

Research Background and Issues

• What problems or challenges did the authors identify?

  1. Although Generative AI (GenAI) has been widely applied in various scenarios within user experience (UX) design, its potential in the process of creating user journey maps (Journey Map, JM) has not been thoroughly explored.
  2. Commonly used JM tools face challenges such as difficulty in managing large amounts of data, reliance on manual interpretation of qualitative data (e.g., interview scripts), and rigid creation processes that limit creativity and insights.
  3. Existing GenAI design tools, such as QoQo.ai, often lack input sources from real user data, resulting in generated content that risks being inauthentic or contextually irrelevant.

• Why is this issue important? User journey maps are a critical design tool for visualizing the interaction process between users and services or products. They help designers better understand user needs and identify pain points to develop improvement strategies. If GenAI can effectively support the JM process, it could not only improve data management efficiency but also make design outcomes more insightful and creative.

• Research Motivation and Related Work The motivations for this research are:

  1. To integrate the technical advantages of GenAI (e.g., data processing, insight generation, and visualization) into the JM creation process to address existing issues.
  2. To explore how to accelerate the process and enhance the accuracy of generated content while maintaining the integrity of the user journey. Related work includes preliminary studies on the application of GenAI in UX design tools, but these studies are limited to specific scenarios (e.g., user personas or prototype design).

Solution

• What methods or solutions did the authors propose? The authors designed and developed a tool called GeneyMAP, which integrates OpenAI's GPT-4 and DALL-E 3 APIs to support the creation of user journey maps. GeneyMAP automates the following tasks:

  1. Template extraction and user data mapping.
  2. Generating personalized journey maps from user interview scripts.
  3. Merging journey maps and identifying pain points and design opportunities.
  4. Providing interactive support for generating design inspiration.

• What are the innovative aspects of this solution?

  1. End-to-end support: This is the first exploration of integrating GenAI throughout the entire JM creation process, including data analysis, mapping, pain point identification, and visualizing design opportunities.
  2. Dual-modal data support: The tool combines user interview data with generative content, enhancing the quality of generated outputs through a blend of human and AI inputs.
  3. Customizable design experience: Through a flexible prompt engineering framework, designers can precisely control the output content.

• What are the implementation steps and key technologies used? GeneyMAP consists of four main steps:

  1. Extracting structured JM templates (D1): Automatically analyzing and extracting journey stages and interaction elements from interview data.
  2. Generating personalized JMs (D2, D5): Creating journey maps by combining real user data with synthetic user data.
  3. Merging JMs and identifying design opportunities (D3, D4): Using GenAI to integrate multi-user data and propose pain points and opportunities for design improvements.
  4. Exploring design inspiration (D5, D6): Generating prompts and providing visualization support to inspire and supplement design solutions. The tool was developed using the Unity engine and incorporates a Chain-of-Thought (CoT) prompting structure to enhance the logical consistency and coherence of outputs.

Research Outcomes

• What specific results were achieved?

  1. GeneyMAP significantly improved the efficiency of JM creation, with outputs outperforming existing tools in terms of novelty, detail, engagement, and surprise.
  2. Designers using GeneyMAP were able to map user data more quickly and comprehensively, discovering new design inspirations.
  3. The tool effectively balanced data organization efficiency with the flexibility for designers to customize outputs according to their needs.

• What advantages does it have compared to existing solutions?

  1. Improved completeness of user journey maps, saving time on data analysis and script processing.
  2. Provided more insightful solutions through synthetic users and data visualization.
  3. Offered greater interactivity and iterative support than traditional tools, helping designers focus more quickly on core design aspects.

• What were the experimental or evaluation results?

  1. Comparative experiments showed that participants' average time spent was significantly reduced (from 36.6 minutes to 30.7 minutes).
  2. Expert evaluations and surveys indicated that GeneyMAP significantly outperformed traditional tools in terms of novelty (p = 0.016), engagement (p = 0.006), and surprise (p = 0.007).
  3. Respondents reported that the tool enhanced their ability to integrate data, identify pain points, and improved the transparency of the generation process.

• Limitations and Future Directions

  1. Limitations:
     • Data sources are limited to text (e.g., interview scripts) and cannot integrate multimodal data (e.g., video or audio).
     • Some generated content lacks precision or contextual relevance, such as irrelevant design opportunities or misinterpreted pain points.
     • The tool heavily relies on effective prompt engineering skills.
  2. Future Directions:
     • Expand GenAI capabilities through multimodal input to support data in formats like images and audio.
     • Develop design quality control mechanisms, such as further validation or optimization of synthetic user-generated content.
     • Conduct technical research to improve tool transparency and explainability, for example, by adding visualizations of generation decisions.