Data Studio
Intro
Edit Demo Data Like a Spreadsheet
Reprise → 2024-2025, ~8 Months
Principal Product Designer
0-1
New product area buildout
Live
Custom data on your real production app
3
Rapid iterations
AI
LLM-powered data manipulation
The Work
Some upfront complexity couldn't be abstracted
v1 version was too technical
The first version was engineering-driven. Even semi-technical users struggled to understand the product architecture. Setup was painful, adoption was limited.
Hierarchy and mental model changes from v1 to v2
Shifting the mental model
I led a rearchitecture around how users actually think: Product Areas (the pages they demo) containing Widgets (the data elements like tables and charts). Sales people know their story and product, so we let them build around that.
Updated landing page with reworked taxonomy
Spreadsheet-style editing
We invested in making it behave like they'd expect, heavy keyboard navigation, row & column section, point & click to select multiple cells. Familiar patterns, less learning curve.
AI-powered data generation
Natural language prompts to edit data at scale. Worked closely with the senior dev to figure out what was actually possible for v1 and smoothed the UX. This was a high visibility feature, our own demos depended on it working well.
Initial rough AI dialog flow
Outcomes
Non-technical users could self-serve. Faster demo setup, richer customization. The AI feature became a selling point in our own sales process.
- 3x faster demo creation
- Launched Live Overlay product tier
- Sales teams can self-serve custom demos
- Higher close rates from personalized demos
Editing the data driving the chart in the app
Switch Datasets on the fly
Built out the landing page as well
