Power Bi Design Vault Dashboards

Power BI Design Vault Dashboards

Skill by ara.so — Design Skills collection

Expert guidance for leveraging the Power-Narrative Design Studio repository: a curated collection of Power BI dashboard templates, design systems, DAX libraries, and visual intelligence frameworks for modern business analytics.

What This Project Provides

The Power BI Design Vault is a comprehensive library of:

• Pre-built .pbix dashboard templates for retail, HR, finance, and supply chain scenarios
• Reusable JSON theme files with accessibility-compliant color palettes
• Sample datasets in Parquet and CSV formats with synthetic business data
• Annotated DAX measure libraries with inline documentation
• SVG icon sets and visual assets optimized for Power BI theming
• Responsive layout architectures following 12-column grid systems
• Design pattern documentation for visual hierarchy and cognitive accessibility

Primary use case: Accelerate Power BI dashboard development by adapting pre-built blueprints rather than building from scratch.

Installation and Setup

Prerequisites

• Power BI Desktop (October 2025 release or later)
• Basic familiarity with Power BI data modeling and DAX
• Git or GitHub Desktop for cloning the repository

Clone the Repository

# Clone the repository to your local machine
git clone https://github.com/Lithiumgreentek/power-bi-design-vault.git
cd power-bi-design-vault

Repository Structure

power-bi-design-vault/
├── dashboards/
│   ├── retail-sales-opportunity/
│   │   ├── README.md
│   │   ├── retail-sales.pbix
│   │   └── assets/
│   ├── hr-attrition-forecast/
│   └── supply-chain-bottleneck/
├── datasets/
│   ├── sample-retail-clean.parquet
│   └── finance-trials.csv
├── themes/
│   ├── dark-enterprise.json
│   └── pastel-minimal.json
├── visual-assets/
│   ├── icons/
│   └── background-vectors/
└── documentation/
    ├── design-principles.md
    └── data-modeling-guide.pdf

Working with Dashboard Templates

Opening a Dashboard Template

1. Navigate to the specific dashboard folder (e.g., dashboards/retail-sales-opportunity/)
2. Open the .pbix file in Power BI Desktop
3. Review the model view to understand pre-configured relationships
4. Check the "DAX Dictionary" page for measure explanations

Connecting Your Own Data

Replace sample dataset with your data source:

// Power Query M code to replace data source
let
    Source = Csv.Document(
        File.Contents("C:\YourPath\your-data.csv"),
        [Delimiter=",", Columns=10, Encoding=1252, QuoteStyle=QuoteStyle.None]
    ),
    PromotedHeaders = Table.PromoteHeaders(Source, [PromoteAllScalars=true]),
    ChangedTypes = Table.TransformColumnTypes(PromotedHeaders, {
        {"Date", type date},
        {"Sales", type number},
        {"Product", type text}
    })
in
    ChangedTypes

Using environment variables for data source paths:

In Power Query, create a parameter:

// Create a parameter for flexible data source paths
DataSourcePath = Text.From(Excel.CurrentWorkbook(){[Name="DataPath"]}[Content]{0}[Column1])

Then reference DataSourcePath in your File.Contents() calls.

Applying Theme Files

Import a JSON theme:

1. In Power BI Desktop, go to View → Themes → Browse for themes
2. Navigate to themes/ folder
3. Select dark-enterprise.json or pastel-minimal.json
4. Click Open

Customize theme colors programmatically:

{
  "name": "Custom Corporate Theme",
  "dataColors": [
    "#1F77B4",
    "#FF7F0E",
    "#2CA02C",
    "#D62728",
    "#9467BD"
  ],
  "background": "#FFFFFF",
  "foreground": "#333333",
  "tableAccent": "#1F77B4",
  "good": "#2CA02C",
  "neutral": "#FFC107",
  "bad": "#D62728",
  "textClasses": {
    "callout": {
      "fontSize": 45,
      "fontFace": "Segoe UI",
      "color": "#333333"
    },
    "title": {
      "fontSize": 14,
      "fontFace": "Segoe UI Semibold",
      "color": "#666666"
    }
  }
}

Save this as custom-theme.json and import via the same process.

DAX Patterns from the Vault

Dynamic Measure Titles

// Dynamic title that changes based on slicer selection
Dynamic Title = 
VAR SelectedPeriod = SELECTEDVALUE('Calendar'[Period], "All Periods")
VAR SelectedRegion = SELECTEDVALUE('Geography'[Region], "All Regions")
RETURN
    "Sales Performance - " & SelectedPeriod & " | " & SelectedRegion

Parameter-Driven Measures

// Toggle between different metrics using a parameter table
Selected Metric = 
SWITCH(
    SELECTEDVALUE('MetricSelector'[Metric]),
    "Revenue", [Total Revenue],
    "Profit", [Total Profit],
    "Units", [Total Units Sold],
    [Total Revenue] // Default
)

Time Intelligence with Fiscal Calendar

// Year-to-date calculation respecting fiscal year starting in April
Sales YTD (Fiscal) = 
CALCULATE(
    [Total Sales],
    DATESYTD(
        'Calendar'[Date],
        "3/31" // Fiscal year ends March 31
    )
)

Previous Period Comparison

// Sales vs. previous period with error handling
Sales vs Previous Period = 
VAR CurrentSales = [Total Sales]
VAR PreviousSales = 
    CALCULATE(
        [Total Sales],
        DATEADD('Calendar'[Date], -1, MONTH)
    )
VAR PercentChange = 
    IF(
        ISBLANK(PreviousSales) || PreviousSales = 0,
        BLANK(),
        DIVIDE(CurrentSales - PreviousSales, PreviousSales)
    )
RETURN
    PercentChange

Custom Tooltip Measure

// Rich tooltip text with conditional formatting
Tooltip Text = 
VAR ProductName = SELECTEDVALUE('Product'[Name])
VAR Sales = [Total Sales]
VAR Target = [Sales Target]
VAR Achievement = DIVIDE(Sales, Target, 0)
VAR PerformanceLabel = 
    SWITCH(
        TRUE(),
        Achievement >= 1.1, "Exceeding",
        Achievement >= 1.0, "On Target",
        Achievement >= 0.9, "Near Target",
        "Below Target"
    )
RETURN
    ProductName & UNICHAR(10) &
    "Sales: " & FORMAT(Sales, "$#,##0") & UNICHAR(10) &
    "Target: " & FORMAT(Target, "$#,##0") & UNICHAR(10) &
    "Status: " & PerformanceLabel

Rank with Tie Handling

// Product rank by sales with ties resolved alphabetically
Product Rank = 
RANKX(
    ALL('Product'[Name]),
    [Total Sales] + (DIVIDE(1, UNICODE('Product'[Name]) + 1000000)),
    ,
    DESC,
    DENSE
)

Responsive Layout Best Practices

Grid System Implementation

When adapting templates, maintain the 12-column grid structure:

• Header section: 12 columns × 2 rows (fixed)
• KPI cards: 3 columns each × 2 rows (4 cards per row)
• Main chart area: 8 columns × 6 rows
• Filter panel: 4 columns × 6 rows (right sidebar)
• Footer/details: 12 columns × 1 row

Mobile Layout Conversion

For mobile views, the vault templates use this pattern:

1. Stack KPI cards vertically (12 columns × 2 rows each)
2. Collapse filter panel into dropdown slicers
3. Simplify charts (e.g., clustered column → simple bar)
4. Remove decorative visual elements

Bookmark navigation for mobile:

Create bookmarks for:

• Overview (KPIs only)
• Trend Analysis (time series chart)
• Detail View (table with drill-through)

Dataset Integration Patterns

Loading Parquet Files

// Load Parquet file from datasets folder
let
    Source = Parquet.Document(
        File.Contents("datasets/sample-retail-clean.parquet")
    ),
    Navigation = Source{[Name="data"]}[Data]
in
    Navigation

CSV Import with Type Inference

// Import CSV with automatic type detection
let
    Source = Csv.Document(
        File.Contents("datasets/finance-trials.csv"),
        [Delimiter=",", Encoding=65001, QuoteStyle=QuoteStyle.Csv]
    ),
    PromotedHeaders = Table.PromoteHeaders(Source, [PromoteAllScalars=true]),
    DetectedTypes = Table.TransformColumnTypes(
        PromotedHeaders,
        List.Zip({
            Table.ColumnNames(PromotedHeaders),
            List.Transform(
                Table.ColumnNames(PromotedHeaders),
                each try Table.Schema(PromotedHeaders){[Name=_]}[TypeName] otherwise "text"
            )
        })
    )
in
    DetectedTypes

Multilingual Field Handling

// Display measure value with locale-specific formatting
Localized Sales = 
VAR Sales = [Total Sales]
VAR CurrencySymbol = SELECTEDVALUE('Locale'[currency_symbol], "$")
VAR DecimalSeparator = SELECTEDVALUE('Locale'[decimal_separator], ".")
RETURN
    CurrencySymbol & FORMAT(Sales, "#,##0" & DecimalSeparator & "00")

Visual Asset Integration

Using SVG Icons

Import SVG as image:

1. Insert → Image
2. Browse to visual-assets/icons/
3. Select appropriate icon (e.g., trending-up.svg)
4. Set image fit to "Fit" (not "Fill")
5. Use conditional formatting to swap icons based on measure values

Dynamic icon selection with conditional formatting:

Create a measure that returns image file paths:

Performance Icon URL = 
VAR Performance = [Sales vs Target %]
VAR IconPath = "https://raw.githubusercontent.com/Lithiumgreentek/power-bi-design-vault/main/visual-assets/icons/"
RETURN
    SWITCH(
        TRUE(),
        Performance >= 1.1, IconPath & "trending-up.svg",
        Performance >= 0.9, IconPath & "stable.svg",
        IconPath & "trending-down.svg"
    )

Apply this measure to an image visual's Image URL field.

Background Vector Integration

For dashboard backgrounds:

1. Canvas Settings → Canvas background → Image
2. Browse to visual-assets/background-vectors/
3. Set transparency to 95-98% for subtle effect
4. Use "Fit" sizing to avoid distortion

Common Customization Workflows

Adapting Retail Template for Different Industry

Scenario: Convert retail dashboard to healthcare metrics

1. Replace dataset:

   • Swap sample-retail-clean.parquet with patient admission data
   • Maintain similar schema (Date, Category, Metrics)

2. Rename measures:

   • Total Sales → Total Admissions
   • Average Transaction → Average Length of Stay

3. Update visuals:

   • Product hierarchy → Department/Condition hierarchy
   • Regional map → Hospital campus map

4. Adjust color palette:

   • Apply pastel-minimal.json theme
   • Change accent colors to healthcare-friendly blues/greens

Creating a Drill-Through Page

// Measure to enable drill-through only when single item selected
Enable Drillthrough = 
IF(
    HASONEVALUE('Product'[Name]),
    1,
    BLANK()
)

Configure drill-through in Power BI:

1. Create new report page (e.g., "Product Detail")
2. Add Product[Name] to drill-through filters
3. Add back button: Insert → Button → Back
4. Use Enable Drillthrough measure in visual-level filter

Implementing Bookmarks for Narrative Flow

Create a guided narrative:

1. Bookmark 1: Executive Summary

   • Show only KPI cards and trend chart
   • Hide detail tables and slicers

2. Bookmark 2: Regional Breakdown

   • Show map and regional bar chart
   • Display region slicer

3. Bookmark 3: Product Deep Dive

   • Show product matrix and decomposition tree
   • Display product category slicer

Add navigation buttons:

• Insert → Button → Blank
• Action → Type: Bookmark → Select corresponding bookmark
• Label buttons "Overview" → "Regions" → "Products"

Troubleshooting

Theme Not Applying to All Visuals

Issue: Custom theme colors don't affect certain chart types

Solution:

• Ensure theme JSON includes all visual types in visualStyles section
• Manually set colors for custom visuals (they may not support theming)
• Check Power BI Desktop version supports theme features used

DAX Measures Return BLANK() Unexpectedly

Issue: Measures work in some visuals but return blank in others

Checklist:

1. Verify filter context isn't removing all rows: Use COUNTROWS(ALL('Table'))
2. Check for division by zero: Wrap in DIVIDE(numerator, denominator, 0)
3. Ensure relationships are bidirectional if needed (use sparingly)
4. Test measure in table visual with all dimensions to debug

Dataset Path Broken After Moving Files

Issue: Data source paths hardcoded to specific machine

Solution:

// Use relative path from .pbix location
let
    CurrentFileFolder = Text.BeforeDelimiter(
        Text.From(Excel.CurrentWorkbook(){0}[Content]),
        Text.From(Excel.CurrentWorkbook(){0}[Name])
    ),
    DatasetPath = CurrentFileFolder & "datasets\sample-retail-clean.parquet",
    Source = Parquet.Document(File.Contents(DatasetPath))
in
    Source

Or use parameter with environment variable:

# Set environment variable (Windows)
setx POWERBI_DATA_PATH "C:\Projects\power-bi-design-vault\datasets"

Then reference in Power Query via external parameter configuration.

Performance Issues with Large Datasets

Optimization strategies:

1. Use Import mode for small datasets (<1GB)
2. Switch to DirectQuery for large SQL sources
3. Aggregate data in Power Query:

// Pre-aggregate data before loading
let
    Source = Csv.Document(File.Contents("large-dataset.csv")),
    GroupedData = Table.Group(
        Source,
        {"Date", "Product"},
        {
            {"Total Sales", each List.Sum([Sales]), type number},
            {"Avg Price", each List.Average([Price]), type number}
        }
    )
in
    GroupedData

4. Optimize DAX with variables:

// Cache intermediate calculations
Optimized Margin = 
VAR TotalRevenue = [Total Revenue]
VAR TotalCost = [Total Cost]
VAR Margin = TotalRevenue - TotalCost
VAR MarginPercent = DIVIDE(Margin, TotalRevenue)
RETURN
    MarginPercent

Responsive Layout Not Working on Mobile

Common causes:

1. Fixed size visuals: Convert to relative sizing

   • Select visual → Format → General → Size → Use percentage values

2. Overlapping visuals: Use layout grid

   • View → Gridlines → Snap to grid
   • Ensure no visuals share exact same coordinates

3. Mobile layout not configured:

   • View → Mobile layout
   • Manually arrange visuals for phone view
   • Remove non-essential elements

Advanced Pattern: Custom Tooltip Page

Create reusable tooltip pages for consistent hover experiences:

Tooltip page setup:

1. Create new page, rename to "Product Tooltip"
2. Page Settings → Page Information → Allow use as tooltip: On
3. Set canvas size to 320×240 px
4. Add compact visuals (card, mini bar chart, sparkline)

DAX for tooltip-specific measures:

// Tooltip: Show last 6 months trend
Tooltip Sales Trend = 
CALCULATE(
    [Total Sales],
    DATESINPERIOD(
        'Calendar'[Date],
        MAX('Calendar'[Date]),
        -6,
        MONTH
    )
)

Apply tooltip:

1. Select any visual on main report page
2. Format → Tooltip → Type: Report page
3. Page: Select "Product Tooltip"
4. Hover over data points to see custom tooltip

Working with Documentation

The vault includes extensive documentation:

design-principles.md: Guidelines on visual hierarchy, gestalt principles, color psychology

data-modeling-guide.pdf: Star schema vs snowflake, when to use calculated tables, relationship best practices

Key takeaways to implement:

• Rule of thirds: Place critical KPIs at intersection points
• F-pattern reading: Position navigation and filters top-left
• Color temperature: Warm colors (red/orange) for alerting, cool colors (blue/green) for positive trends
• Whitespace ratio: Maintain 30-40% empty space for visual breathing room

Example: Complete Dashboard Customization

Scenario: Adapt HR Attrition dashboard for your company data

// Step 1: Load your HR data
let
    Source = Sql.Database(
        Environment.GetEnvironmentVariable("SQL_SERVER"),
        Environment.GetEnvironmentVariable("SQL_DATABASE")
    ),
    EmployeeTable = Source{[Schema="dbo",Item="Employees"]}[Data],
    FilteredRows = Table.SelectRows(
        EmployeeTable,
        each [TerminationDate] = null or [TerminationDate] >= #date(2024,1,1)
    )
in
    FilteredRows

// Step 2: Create attrition rate measure
Attrition Rate = 
VAR TotalEmployees = 
    CALCULATE(
        DISTINCTCOUNT('Employee'[EmployeeID]),
        ALL('Calendar')
    )
VAR Terminations = 
    CALCULATE(
        DISTINCTCOUNT('Employee'[EmployeeID]),
        'Employee'[Status] = "Terminated"
    )
RETURN
    DIVIDE(Terminations, TotalEmployees, 0)

// Step 3: Forecast next quarter attrition
Forecasted Attrition = 
VAR HistoricalTrend = 
    LINESTX(
        ADDCOLUMNS(
            SUMMARIZE('Calendar', 'Calendar'[Quarter]),
            "@Attrition", [Attrition Rate]
        ),
        [Quarter],
        [@Attrition]
    )
VAR Slope = [Slope]
VAR Intercept = [Intercept]
VAR NextQuarter = MAX('Calendar'[Quarter]) + 1
RETURN
    Slope * NextQuarter + Intercept

Step 4: Apply dark-enterprise.json theme for executive presentation

Step 5: Configure drill-through from department summary to individual employee analysis

This workflow demonstrates the full power of the design vault: starting with a template, integrating real data sources, applying custom DAX logic, and delivering a production-ready dashboard.

Repository: github.com/Lithiumgreentek/power-bi-design-vault

License: MIT — Use, modify, and redistribute freely for personal or commercial projects.