As data professionals, we often face one common challenge — understanding the story behind the data before building predictive models. Whether you’re building a machine learning pipeline or preparing insights for business decisions, Exploratory Data Analysis (EDA) is the first crucial step.
In this tutorial, I’ll show you how I created an interactive Car Price Prediction dashboard in Power BI, where we visualize patterns and relationships between features like brand, fuel type, transmission, mileage, and price.
Get the dataset here: https://www.kaggle.com/datasets/aliiihussain/car-price-prediction
This Power BI dashboard helps answer key questions:
- Which car brands have the highest average price?
- How does fuel type affect car prices?
- What’s the price distribution across different conditions and models?
- Do newer cars really command higher prices?
Let’s begin building this visualization from scratch.
🔹 Step 1: Understanding the Dataset
We’ll use a CSV file that contains details of car listings for price prediction. The columns include:
| Column | Description |
|---|---|
| Car ID | Unique identifier for each car |
| Brand | Car manufacturer (e.g., Toyota, Honda, Ford) |
| Model | Specific car model name |
| Year | Manufacturing year |
| Mileage | Total distance driven (in kilometers) |
| Fuel Type | Type of fuel (Petrol, Diesel, Hybrid, Electric) |
| Transmission | Gear system type (Manual / Automatic) |
| Condition | Overall condition (New, Used, Excellent, etc.) |
| Price | Market price of the car (Target variable) |
Each row represents a unique car, and our goal is to analyze the key features influencing price variations.
🔹 Step 2: Importing Data into Power BI
- Open Power BI Desktop and click on Get Data → Text/CSV.
- Browse to your dataset file and click Load.
- You’ll see a table preview — verify that columns like Mileage, Price, and Year are correctly recognized as numeric fields.
Once loaded, Power BI automatically adds your table to the Fields pane on the right-hand side.
🔹 Step 3: Data Cleaning and Preparation
To perform meaningful analysis, data preparation is key.
1. Rename Columns
In the Power Query Editor, rename columns for clarity (e.g., rename “Mileage” to “Distance Driven (Km)”).
2. Handle Missing Values
Check for missing entries in Price or Brand. If a car’s price or brand is missing, either fill with the average price of that brand or remove the record for simplicity.
3. Change Data Types
Ensure the following data types are correct:
- Year → Whole Number
- Mileage → Decimal Number
- Price → Decimal Number
4. Create Calculated Columns
You can add more analytical insights using calculated columns. For example, convert car Age from Year:
Car Age = YEAR(TODAY()) - 'Car Data'[Year]
This new column helps analyze how car age impacts price.
🔹 Step 4: Creating Key Performance Indicators (KPIs)
KPIs give us an at-a-glance view of the dataset summary. In the report canvas, use Card visuals to display:
- Count of Cars:
Count of Cars = COUNT('Car Data'[Car ID]) - Average Engine Size: (if available)
Avg Engine Size = AVERAGE('Car Data'[Engine Size]) - Average Distance Driven (Km):
Avg Distance Driven = AVERAGE('Car Data'[Mileage]) - Sum of Price:
Total Price = SUM('Car Data'[Price]) - Average of Price:
Avg Price = AVERAGE('Car Data'[Price])
Arrange these KPIs at the top of your dashboard for a clean summary section.
🔹 Step 5: Visualization 1 – Avg Price & Sum of Price by Brand
We start with a Combo Chart (Line and Clustered Column) to compare total sales value with average price per brand.
- Axis → Brand
- Column Value → Sum of Price
- Line Value → Avg Price
This chart instantly shows which brands dominate in both total sales and price — for instance, Tesla may have higher average prices, while Toyota might lead in total sales volume.
🔹 Step 6: Visualization 2 – Avg Price by Condition
Next, add a Horizontal Bar Chart to analyze how car condition affects price.
- Axis → Condition
- Values → Avg Price
Here, you’ll see insights such as “Like New” cars commanding the highest average price, while Used or Old conditions yield lower prices.
🔹 Step 7: Visualization 3 – Avg Price by Fuel Type
Create a Donut Chart to show the distribution of average prices by fuel type.
- Legend → Fuel Type
- Values → Avg Price
Use data labels to display both percentage and value.
You’ll likely observe that Electric and Hybrid vehicles are priced higher, reflecting market trends.
🔹 Step 8: Visualization 4 – Avg Price by Model
For this, use a TreeMap visual — perfect for showing models with relative price differences.
- Group → Model
- Values → Avg Price
Larger blocks represent models with higher average prices. You can quickly identify top-performing models like Model S, Explorer, or E-Class.
🔹 Step 9: Visualization 5 – Avg Price by Transmission
Add another Donut Chart to compare Manual vs Automatic cars.
- Legend → Transmission
- Values → Avg Price
You’ll notice patterns such as automatic cars typically costing slightly more due to advanced features.
🔹 Step 10: Add Filters and Slicers for Interactivity
On the right-hand side, we’ll add slicers to make the dashboard interactive.
Add the following slicers:
- Brand
- Condition
- Fuel Type
- Model (allow multiple selections)
- Transmission
- Year
Use the Dropdown style slicer layout for a professional and compact design.
Users can now dynamically explore the dataset — for example, filter only Electric cars from 2020 onwards or analyze Used Toyota cars by Transmission type.
🔹 Step 11: Add Styling and Formatting
To make your dashboard visually appealing:
- Choose a consistent color palette – e.g., shades of green and blue for car brands and fuel type.
- Apply a light shadow effect and rounded corners for visuals.
- Use a bold title:
Car Price Prediction Data - Exploratory Analysis - Place a car image or logo for branding.
- Align all visuals neatly and maintain spacing for readability.
🔹 Step 12: Dashboard Insights
After completing the visuals, here’s what we can derive:
- Brand Performance: Tesla and BMW tend to have higher average prices, reflecting premium positioning.
- Condition Impact: “Like New” cars have 2–5% higher prices than used ones.
- Fuel Type: Electric and hybrid cars show growing dominance, with over 25% higher average price.
- Transmission Trend: Automatic vehicles slightly edge out manual ones in pricing, possibly due to luxury segments.
- Model Analysis: Among models, Model S, 3 Series, and Camry show consistently high resale value.
This exploratory dashboard provides a solid foundation for a Machine Learning model to predict car prices based on key features.
🔹 Step 13: Save and Publish
Once satisfied:
- Save your Power BI report file (
.pbix). - Publish to Power BI Service to share online.
- Add interactive filters so viewers can adjust parameters directly from their browsers.
If you want to embed the dashboard in a website or Power BI app, enable Public Web Sharing (with caution for sensitive data).
🔹 Step 14: Next Steps – Predictive Modeling
Now that you’ve explored the data, the next step is building a Car Price Prediction Model using tools like:
- Python (Scikit-learn or XGBoost)
- Azure Machine Learning Studio
- Power BI Python Integration
You can connect Power BI to your trained ML model to predict prices dynamically, bringing AI and BI together in one platform.
🧠 Key Takeaways
- Exploratory dashboards help you understand variable relationships before modeling.
- Power BI’s combination of visual interactivity and data modeling makes it ideal for business-friendly analytics.
- Always complement BI analysis with predictive techniques for end-to-end data-driven decision-making.
🎥 Conclusion
This Power BI dashboard project demonstrates how raw car listing data can be transformed into actionable insights. Through KPIs, interactive visuals, and intuitive filters, we can quickly explore market trends, understand what drives car prices, and prepare data for machine learning.
If you found this useful, check out my Power BI EDA Series — where we explore real-world datasets step-by-step and connect them with AI-based predictions.
📍 Learn, build, and grow your data analytics skills with me at Slidescope.com