Build This Project · Portfolio Guide
Hiring managers have seen many sales dashboards. The ones that impress answer specific business questions, use a clean data model, and come with a short note explaining what the numbers mean. Follow the steps below in Power BI; the same plan works in Tableau.
1. Start with the questions
Write down four or five questions a sales manager would actually ask, for example:
- How are sales and profit trending month by month, compared with last year?
- Which regions and product categories drive profit, and which lose money?
- Are discounts eroding margin?
- Who are the top customers, and how concentrated is revenue?
Every visual on the dashboard should answer one of these. If it does not, remove it.
2. Get a suitable dataset
You need order-level data with dates, products, customers, regions, sales, discount and profit. The "Sample Superstore" dataset that ships with Tableau, or similar retail datasets on Kaggle, fit well. Check that dates cover at least two years so year-on-year comparisons work.
3. Build a star schema
| Table | Type | Key columns |
|---|---|---|
| FactSales | Fact | OrderID, OrderDate, ProductID, CustomerID, Region, Sales, Discount, Profit |
| DimDate | Dimension | Date, Month, Quarter, Year (marked as the date table) |
| DimProduct | Dimension | ProductID, Product, Sub-Category, Category |
| DimCustomer | Dimension | CustomerID, Customer, Segment |
Split the flat file into these tables in Power Query, then create one-to-many relationships from each dimension to the fact table. A proper date table is what makes time-intelligence functions work.
4. Write the measures
Total Sales = SUM ( FactSales[Sales] )
Total Profit = SUM ( FactSales[Profit] )
Profit Margin % = DIVIDE ( [Total Profit], [Total Sales] )
Sales LY = CALCULATE ( [Total Sales], SAMEPERIODLASTYEAR ( DimDate[Date] ) )
YoY Growth % = DIVIDE ( [Total Sales] - [Sales LY], [Sales LY] )
Check them with numbers you can verify by hand. If a month's sales are 1,20,000 and the same month last
year was 1,00,000, YoY Growth % should show 20%. If that month's profit is 18,000, margin is
15%. Testing measures against a manual calculation catches most DAX mistakes.
5. Lay out the page
- Top row: KPI cards for Total Sales, Total Profit, Profit Margin % and YoY Growth %.
- Middle: a line chart of monthly sales this year vs last year.
- Lower left: profit by region and category as a bar or matrix, with negative values highlighted.
- Lower right: a scatter of discount vs profit margin by sub-category.
- Slicers: year, region and segment, kept to one side.
Use one accent colour for "good" and one for "bad", consistent titles that state the insight ("Furniture loses money in the Central region"), and enough white space. Add a second page for top-customer analysis.
6. Write the insights
Add a short text panel or a README with three findings and one recommendation each, plus limitations: for example, that the data has no cost of returns or marketing spend. This step is what separates an analyst's work from a chart collection.
7. Publish and present
- Upload the .pbix, screenshots and README to GitHub; building a GitHub portfolio shows a structure.
- Record a two-minute walkthrough video explaining one insight.
- Be ready to explain the model, one DAX measure and one decision you made about the data.
For tool depth, see the Power BI toolset; for a Tableau version, advanced Tableau dashboards. Round out your portfolio with other beginner analytics projects or Power BI project ideas, and read building a job-ready analytics portfolio.
Uncodemy's Power BI course includes guided dashboard projects and Dedicated Placement Support.