Looker Studio for Analysts
🎯 Purpose
This guidebook teaches analysts how to use Looker Studio (formerly Data Studio) for creating interactive dashboards powered by BigQuery, with a focus on best practices for filters, layout, speed, and interpretation.
🔌 1. Connect to BigQuery¶
- Use "BigQuery" as your data source
- Navigate through project → dataset → table or custom query
- Recommended: use a VIEW or saved query to control logic outside Looker
✔️ Helps reduce dashboard load time and manage logic externally
🧱 2. Set Up a Reusable Dataset¶
- Use a table or view with pre-filtered, pre-aggregated data
- Include clear field names and calculated fields (like conversion_rate, retention_day, etc.)
- Avoid raw event-level logs if not needed
-- Ideal for Looker: Pre-aggregated metrics
SELECT cohort_month, days_since_signup, retained_users FROM my_project.my_dataset.cohort_summary;
📊 3. Core Visualization Types¶
| Chart | Use Case |
|---|---|
| Scorecards | KPIs like DAU, revenue, churn rate |
| Time Series | Trend tracking over days/weeks/months |
| Bar + Column | Channel comparison, cohort volume |
| Pie / Donut | Share by segment or category |
| Pivot Table | Category breakdown over time or dimension |
| Tables | Underlying drill-down for raw views |
✔️ Add comparison metrics (vs previous period, % change)
🧭 4. Filters, Controls, and UX¶
| Feature | Notes |
|---|---|
| Date Range Picker | Add global or chart-level selectors |
| Dropdown Filter | Segment by user type, region, channel |
| Custom Calculated Field | Build ratios or conditional flags inline |
| Chart Filter | Use filter controls on single charts for targeted views |
✔️ Label everything — filters, legends, and axis titles must be clear for stakeholders
🧪 5. Design Best Practices¶
- Group related metrics together (use layout grids)
- Color consistently across views (same colors for channels, statuses)
- Show last updated timestamp
- Use bold scorecards at top for high-priority KPIs
- Avoid overloading pages with too many visuals
📤 6. Sharing, Export, and Embedding¶
- Dashboards can be shared via Google permissions or public links
- Charts and tables can be exported (CSV, PDF)
- Embed dashboards in docs, sites, or email reports
✅ Dashboard Readiness Checklist¶
- [ ] Connected to clean BigQuery view or stable table
- [ ] Visual layout groups KPIs and trends logically
- [ ] Filters and dropdowns are intuitive and labeled
- [ ] Date ranges are consistent and default correctly
- [ ] Charts include legends, axis labels, and titles
- [ ] Dashboard optimized for speed (pre-aggregated views)
💡 Tip¶
“Looker dashboards should answer questions at a glance — if it takes more than 10 seconds to explain, simplify it.”