Dashboard Layout and Data Visualization UX

You are a dashboard UX designer who has designed data products for Bloomberg Terminal, Datadog, and Amplitude-level complexity.

Dashboard purpose: [WHAT DECISIONS DOES THIS DASHBOARD SUPPORT]
Primary users: [ROLE AND TECHNICAL SOPHISTICATION]
Data types: [TIME SERIES/CATEGORICAL/GEOGRAPHIC/HIERARCHICAL/REAL-TIME]
Update frequency: [REAL-TIME/HOURLY/DAILY/MANUAL REFRESH]
Key metrics count: [5-10/10-20/20-50/50+]

Design a complete dashboard UX:

**1. Information Hierarchy**
- What metric do users look at first? (hero metric placement)
- What's the natural reading flow? (Z-pattern, F-pattern, or scanning)
- How to group related metrics (by team, funnel stage, time)
- Progressive disclosure: overview to detail drill-down path
- Default view vs customized view

**2. Layout Grid System**
- Widget grid (12-column for desktop, stacking rules for mobile)
- Widget sizes: small (1/4), medium (1/2), large (full), tall (2 rows)
- Recommended layout for your specific metric count
- Drag-and-drop customization UX
- Saved layout / view management

**3. Chart Type Selection Guide**
For each metric type, recommend:
- Single value: big number + sparkline + delta indicator
- Time series: line chart (multi-line, stacked area, or bar)
- Comparison: grouped bar, bullet chart, or slope chart
- Proportion: donut chart (not pie), treemap, or stacked bar
- Distribution: histogram, box plot, or violin
- Correlation: scatter plot or bubble chart
- Geographic: choropleth or point map
- Funnel: horizontal funnel with drop-off percentages
- Table: when and how to use data tables effectively

**4. Interaction Design**
- Global date range picker (presets + custom range)
- Filter panel (position, persistence, clear all)
- Hover tooltips (content, positioning, delay)
- Click-to-drill-down behavior
- Cross-filtering (clicking one chart filters others)
- Zoom and pan for time series
- Export and sharing (PDF, PNG, CSV, link)

**5. Loading and Empty States**
- Skeleton screens for initial load
- Shimmer animation for refreshing data
- Error states for individual widgets vs whole page
- No data available states
- Stale data indicators

**6. Responsive Behavior**
- Desktop (1440px+): full grid, all widgets visible
- Tablet (768-1439px): condensed grid, prioritized widgets
- Mobile (< 768px): single column, swipeable cards, key metrics only
- Chart adaptation rules (sparklines replace full charts on mobile)

**7. Accessibility for Data Viz**
- Color-blind-safe palettes for all charts
- Alt text strategy for charts
- Keyboard navigation within dashboards
- Data table alternatives for every chart
- High contrast mode

Provide exact pixel specs, spacing values, and interaction behaviors. This should be handoff-ready for developers.