Visualization Intermediate · 6 min read

Build a Sales Dashboard in Power BI: End-to-End Tutorial

Follow a real-world project from raw data to finished dashboard. Learn data import, Power Query cleaning, star schema modeling, DAX measures, and visual design — all in one walkthrough.

Power BI Tutorials Team
Published

Most tutorials teach Power BI features in isolation. This guide is different — we’ll build a complete sales dashboard from scratch, covering every stage: data import, cleaning, modeling, DAX, and visual design. By the end, you’ll have a repeatable workflow you can apply to any business dataset.

The scenario

You’re a data analyst at a retail company. The sales director wants a dashboard to answer:

  1. What’s our revenue trend over the last 12 months?
  2. Which product categories are over/under performing vs. target?
  3. Which regions need attention?

You have three data sources:

  • Sales transactions (CSV, 50,000 rows)
  • Product master (Excel, 500 products)
  • Sales targets (Excel, monthly targets by category)

Step 1: Import and clean data

Import the sales CSV

In Power BI Desktop: Get Data → Text/CSV → select sales.csv → Transform Data (not Load).

Common issues in raw sales data:

  • Dates as text (2026-01-15 instead of a date type)
  • Revenue with currency symbols ($1,200.00)
  • Inconsistent category labels (“Electronics” vs “Electronic”)
  • Duplicate rows from system exports

Clean in Power Query

Apply these transformations:

  1. Set data types:

    • OrderDate → Date
    • Revenue → Currency (remove $ and , first if needed)
    • Quantity → Whole Number
    • ProductKey → Whole Number
    • Region → Text

  2. Remove duplicates:

    • Select all columns → Right-click → Remove duplicates

  3. Filter out test data:

    • Filter Revenue > 0 (remove nulls and test orders)

  4. Remove unused columns:

    • Keep only: OrderDate, ProductKey, Region, Quantity, Revenue, Cost

Import product master

Get Data → Excel → products.xlsx → Transform Data:

  1. Set ProductKey to Whole Number
  2. Remove columns you don’t need (keep: ProductKey, ProductName, Category, Subcategory, Brand)
  3. Standardize category labels (Replace Values: “Electronic” → “Electronics”)

Import sales targets

Get Data → Excel → targets.xlsx → Transform Data:

The targets file likely has months as columns (wide format). Unpivot:

  1. Select Category column
  2. Transform → Unpivot Other Columns
  3. Rename: Attribute → Month, Value → TargetRevenue
  4. Set TargetRevenue to Currency
  5. Convert Month text (“Jan 2026”) to a date that matches your date table

Step 2: Build the star schema

Create a date table (if you don’t have one):

Date = CALENDAR(DATE(2025, 1, 1), DATE(2026, 12, 31))

Add columns:

Year = YEAR('Date'[Date])
Quarter = "Q" & QUARTER('Date'[Date])
Month = FORMAT('Date'[Date], "MMM yyyy")
MonthKey = FORMAT('Date'[Date], "YYYYMM")

Mark as date table: Model View → select Date table → Mark as date table → choose Date column.

Set up relationships

In Model View, create these relationships (all many-to-one, single-direction):

Date[Date]           → Sales[OrderDate]
Product[ProductKey]  → Sales[ProductKey]
Date[Date]           → Targets[Month]  (may need a MonthKey relationship)

Your model should look like a star: Sales at the center, with Date, Product, and Targets as dimensions radiating out.

Step 3: Write DAX measures

Basic measures

Total Revenue = SUM(Sales[Revenue])
Total Quantity = SUM(Sales[Quantity])
Total Cost = SUMX(Sales, Sales[Quantity] * Sales[UnitCost])
Gross Margin = [Total Revenue] - [Total Cost]
Gross Margin % = DIVIDE([Gross Margin], [Total Revenue])

Target comparison

Target Revenue = SUM(Targets[TargetRevenue])

Revenue vs Target = [Total Revenue] - [Target Revenue]
Revenue vs Target % = DIVIDE([Revenue vs Target], [Target Revenue])

Time intelligence

Revenue LY = CALCULATE([Total Revenue], SAMEPERIODLASTYEAR('Date'[Date]))
Revenue YoY = [Total Revenue] - [Revenue LY]
Revenue YoY % = DIVIDE([Revenue YoY], [Revenue LY])

Revenue YTD = TOTALYTD([Total Revenue], 'Date'[Date])

Rolling 12-month

Revenue Rolling 12M =
CALCULATE(
    [Total Revenue],
    DATESINPERIOD(
        'Date'[Date],
        MAX('Date'[Date]),
        -12,
        MONTH
    )
)

Step 4: Build the dashboard

Page 1: Executive summary

Purpose: Answer “are we on track?”

Layout (top to bottom):

  1. KPI cards (row of 4): Revenue YTD | Target | Variance % | YoY Growth %
  2. Revenue trend (line chart): Date[Month] on x-axis, [Total Revenue] and [Target Revenue] on y-axis
  3. Category performance (bar chart): Product[Category] on y-axis, [Revenue vs Target %] on x-axis
  4. Region breakdown (matrix): Region on rows, [Total Revenue] and [Revenue vs Target %] on values

Page 2: Category deep-dive

Purpose: Answer “which categories need attention?”

  1. Slicer: Product[Category]
  2. Revenue by subcategory (bar chart)
  3. Margin % by product (bar chart, sorted descending)
  4. Revenue vs target by month (clustered column chart)

Page 3: Regional analysis

Purpose: Answer “which regions need attention?”

  1. Slicer: Region
  2. Revenue by region (map visual)
  3. Top 10 products by region (table with conditional formatting)
  4. Monthly trend by region (line chart with small multiples)

Apply design principles

  • Color: one accent color (blue) for current period, gray for comparison
  • Conditional formatting: red for negative variance, green for positive
  • Labels: clear KPI titles, axis labels with units
  • Tooltips: customize to show Revenue, Target, Variance, and LY comparison

See our dashboard design principles for the full guide.

Step 5: Add interactivity

Slicers

Add a date range slicer and a category slicer to the top of each page. Sync slicers across pages using the View → Sync Slicers pane.

Drill-through

Enable drill-through from the regional matrix on Page 1 to Page 3:

  1. On Page 3, drag Region to the “Drill-through” field well
  2. On Page 1, right-click a region in the matrix → Drill through → Regional analysis

Bookmarks

Create bookmarks for different views:

  • “All regions” view
  • “Underperforming regions” view (filtered to negative variance)
  • “Growth opportunities” view (filtered to high YoY growth)

See our bookmarks guide for setup.

Step 6: Optimize and test

Performance check

Run Performance Analyzer (View → Performance Analyzer):

  • Each visual should load in under 2 seconds
  • DAX query time should dominate (not visual display)
  • If slow, check for FILTER(ALL(...)) patterns and unused columns

See our performance guide for optimization techniques.

User testing

Before publishing, watch someone use the dashboard:

  • Can they answer the three business questions?
  • Do they know where to click?
  • Are the numbers formatted clearly?

Step 7: Publish and share

  1. Publish to Power BI Service: File → Publish → select workspace
  2. Set up scheduled refresh: Workspace → Datasets → Settings → Scheduled refresh
  3. Share with stakeholders: Workspace → Reports → Share

For row-level security (each regional manager sees only their region), see Microsoft’s RLS documentation.

What you built

A three-page sales dashboard that:

  • Tracks revenue vs. target with variance analysis
  • Shows YoY growth and rolling 12-month trends
  • Drills from executive summary to category and regional detail
  • Updates automatically with scheduled refresh

The same workflow — import, clean, model, DAX, design, test, publish — applies to any business dashboard you’ll build.

FAQ

How long should this take to build? For a beginner: 4-8 hours. For an experienced analyst: 1-2 hours. The first build is always slowest — once you have the pattern, subsequent dashboards go faster.

Can I use this as a portfolio project? Yes. Use the sample data from Microsoft’s Contoso dataset (free download) and build this dashboard. It demonstrates data cleaning, modeling, DAX, and visual design — all the skills employers look for.

What if my data sources are different? The workflow is the same. The specific transformations change, but the structure (clean in Power Query → star schema → DAX measures → designed pages) applies universally.

What’s next

#project#sales#dashboard#end-to-end#tutorial

Related Tutorials