Build Interactive Tableau Dashboards

This skill designs and builds Tableau dashboards following visual hierarchy principles (Z/F-pattern layout, top-left KPI placement, 3-5 key metrics per dashboard, progressive disclosure) and a disciplined color strategy (6-8 colors max, semantic red/green coding, colorblind-friendly ColorBrewer palettes). It optimizes performance through pre-aggregated custom SQL data sources, extract creation for datasets over 1M rows with incremental refresh conditions, and efficient LOD/table calculations (running sums, percent-of-total, date-windowed filtering). It implements KPI scorecards with status-indicator calculated fields and percentage variance formatting, interactive filter controls (parameter actions, cascading filters, show/hide toggles), mobile-responsive layouts sized per device breakpoint, a parameter-driven navigation structure, consistent formatting standards (number/date formats, font hierarchy, standardized chart sizes), highlight/filter actions for interactivity, null-handling and data-validation calculations, dynamic context-aware titles, and automated alert-threshold calculations.

Use this skill when building or optimizing a Tableau dashboard - writing a pre-aggregated custom SQL data source, deciding when to create an extract with incremental refresh, writing an efficient LOD or running-total calculation, building a KPI scorecard with status indicators, setting up cascading parameter-driven filters, designing a mobile-responsive layout, building a parameter-based navigation structure, or writing alert-threshold calculations.

It does not cover other BI tools (Power BI, Looker) - it is specific to Tableau's calculation language, extract architecture, and dashboard design conventions.

Inputs are typically a data source and the business metrics/KPIs to visualize. Outputs include optimized data sources and calculations, for example a pre-aggregated custom SQL source:

SELECT
    DATE_TRUNC('month', order_date) as month,
    region,
    SUM(sales) as total_sales,
    COUNT(DISTINCT customer_id) as unique_customers,
    AVG(profit_ratio) as avg_profit_ratio
FROM sales_data
WHERE order_date >= DATEADD('year', -2, CURRENT_DATE)
GROUP BY 1, 2

Other outputs include an incremental extract refresh condition, an LOD running-total calculation, a KPI scorecard calculation returning a status indicator string based on actual-vs-target thresholds, a percentage-variance formatted calculation, a cascading parameter filter (Region to Country to City), a mobile-responsive layout breakpoint plan, a parameter-driven navigation show/hide calculation, and an alert-threshold calculation flagging critical/warning/on-track status.

It also covers error handling and data-quality validation, for example:

IFNULL([Sales], 0)

IF ISNULL([Critical Field])
THEN "Data Missing"
ELSE "Data Available" END

For publishing and maintenance, it recommends consistent naming conventions for workbooks and data sources, version control through Tableau Server projects, automated refresh schedules during low-usage hours, documentation of data sources/calculations/business logic, user guides with screenshots for complex interactions, and regular review cycles for data accuracy and relevance.

BI developers building Tableau dashboards who need performance-optimized data sources, efficient calculations, and disciplined visual design rather than an unoptimized, cluttered workbook.