“Design Systems for Data Products”
summary
Defines a data product using DJ Patil's definition and offers six concrete design-system moves: three-mode switching (visual/table/SQL) used in Wave and Trifacta, a 22-color accessible palette for multicolor charts, ML-driven predictive interaction showing suggestion cards with yellow previews, and a proprietary eclair engine that lets users edit queries by manipulating charts. Emphasizes dual-process design (fast visual cues vs slower dense views), calls out Excel’s scale limits, and argues for guidelines, composable D3 patterns, and data and graphical integrity.
AI-written summary · about
chapters (12)
- 0:00Introduction and personal story
- 2:32What are data products?
- 7:23Why design systems miss data needs
- 11:42Six ideas for data design systems
- 12:49Identify user cognitive preferences
- 16:47Design for dual process theory
- 19:18Optimize visual perception
- 22:29Machine learning content guidelines
- 25:01Composable visualization guidelines
- 27:36Addressing design exceptions
- 31:43Guiding principles for integrity
- 34:59Q&A session
related talks
27mVisual Design - Do not let Data be your cul-de-sac!
38mDesign systems panel discussion
23mDesign Systems and Creativity
19mDesign Systems for Scalable Cloud-Native Analytics Platforms
24mDesign and Development between Art and Science
50mDVUX: Data Visualization and User Experience
23mDesign Systems 101
42mInteractive Data Analysis: Visualization and Beyond