Designing Data-Intensive Applications: Advice for Interaction Designers
This article offers ten strategic recommendations for interaction designers working on complex, data-intensive products. Drawing from the author's thirty years of design experience and recent work as a solo builder in software engineering and data science, the piece argues that screen elements are merely projections of multi-dimensional data landscapes. Key advice includes learning Python to better understand data logic, deeply understanding users' jobs through direct observation and 'dogfooding,' and allowing data structures to dictate interface layout rather than forcing content into pre-designed containers. The author emphasizes minimizing visual clutter ('stripping back chrome'), thoughtfully designing empty states, and using realistic data during development. Additional suggestions focus on intelligent pre-population, bridging the gap between user mental models and underlying data models, precise terminology, and clear navigation. The text highlights how combining familiar graphical conventions, such as whisker plots and Gantt charts, can make abstract concepts like probabilistic milestones accessible. Ultimately, the article encourages designers to look beyond surface-level UI to engage with the foundational data and domain layers, ensuring that the interface effectively communicates real-world phenomena and domain concepts to users.
Wire timeline
Designing Data-Intensive Applications: Advice for Interaction Designers
This article offers ten strategic recommendations for interaction designers working on complex, data-intensive products. Drawing from the author's thirty years of design experience and recent work as a solo builder in software engineering and data science, the piece argues that screen elements are merely projections of multi-dimensional data landscapes. Key advice includes learning Python to better understand data logic, deeply understanding users' jobs through direct observation and 'dogfooding,' and allowing data structures to dictate interface layout rather than forcing content into pre-designed containers. The author emphasizes minimizing visual clutter ('stripping back chrome'), thoughtfully designing empty states, and using realistic data during development. Additional suggestions focus on intelligent pre-population, bridging the gap between user mental models and underlying data models, precise terminology, and clear navigation. The text highlights how combining familiar graphical conventions, such as whisker plots and Gantt charts, can make abstract concepts like probabilistic milestones accessible. Ultimately, the article encourages designers to look beyond surface-level UI to engage with the foundational data and domain layers, ensuring that the interface effectively communicates real-world phenomena and domain concepts to users.
UX Collective - Medium