Zillow Market Snapshot: EDA Dashboard Case Study

A one-day look at new Zillow listings in 10 of the largest U.S. cities, visualized through Power BI.

  • Visual Studio Code
  • Python
  • Pandas
  • Microsoft Power BI
  • DAX
  • Data Pipeline
  • ETL
Project GitHub Repository

At a Glance

  • Built a one-page Power BI dashboard analyzing 380 new Zillow listings (Sept 16, 2025), surfacing insights on pricing, property types, and listing dynamics.
  • Median listing price landed at $439.5K ($292.9/sqft), with single-family homes dominating supply but condos leading in value per sqft.
  • Next up: Scale through incremental data loading, add trend analysis, and integrate valuation analysis (list price vs. tax assessments vs. Zestimate).

Dashboard

Dashboard built in Power BI to compare product-level costs and simulate savings.

Overview

On September 16, 2025, I extracted all new property listings posted on Zillow across ten major U.S. cities using the Zillow API (via RapidAPI). This dataset of 380 listings represents a single-day snapshot of new supply entering the market. The goal was to perform exploratory data analysis (EDA) and build a Power BI dashboard to surface insights on pricing, property characteristics, and listing composition.

Key Objectives

The project was guided by five core questions:

  • What is the overall pricing landscape?

  • How do prices vary across cities and states?

  • What is the composition of the housing market by property type and listing type?

  • How do features like bedrooms, bathrooms, and living area influence price?

  • Which property types deliver the highest value per square foot?

Approach & Analysis

Data Pipeline Creation & Transformation

To keep the workflow reproducible, I built a lightweight ETL pipeline in Python:

  • Environment setup: Created a project folder and virtual environment for isolated development.
  • API ingestion: Queried Zillow’s listings API for ten target cities (Austin, Dallas, Houston, San Antonio, Chicago, Los Angeles, New York, Miami, Phoenix, Seattle).
  • Data handling: Stored results in Pandas, enriched them with city/state metadata, and exported a timestamped CSV for traceability.
  • Cleaning in Power BI: Removed unused columns, standardized headings, extracted clean values from nested fields (e.g., listing subtype), and formatted categorical fields (e.g., hometype, listing type).

Final dataset columns:
HomeType, Bathrooms, Bedrooms, City, State, IsShowcaseListing, ListingType, LivingArea, Price.

 

Feature Engineering

To transform raw records into business-ready insights, I engineered features directly in Power BI:

  • Categorical simplifications:

    • Bedrooms and bathrooms were bucketed (e.g., 3BR, 4 Baths, 5+ BR) to make comparisons more intuitive.

    • Price buckets (0–200K, 200K–400K, …, 1M+) highlighted affordability tiers and simplified price distribution analysis.

  • Core KPIs:

    • Median Price

    • Median Price per SqFt

    • Median Living Area

    • Supporting averages (price, bedrooms, bathrooms, living area).

These KPIs helped counterbalance the distortion caused by extreme luxury outliers in cities like Los Angeles and New York.

  • Exploratory measures: Additional metrics (e.g., average price by state, average price per bedroom, price per sq ft by hometype) powered interactive visuals, making it possible to slice the story by geography, property type, or listing category.

Key Findings

The analysis produced several key insights:

 

1. Pricing Landscape

The median listing price was $439,500, with a median living area of 1,626 sq ft. Median price per square foot stood at $292.88, while the mean values were skewed upward by extreme luxury listings. Most homes clustered in the mid-market range ($400K–$500K), representing typical U.S. family housing.

 

2. Geographic Variability
City-level averages revealed a polarized market:

  • Luxury hubs: Los Angeles ($8.9M) and New York ($4.3M) drove the dataset’s upper extremes.

  • Mid-tier markets: Austin ($1.07M), Miami ($2.15M), and Seattle ($1.43M) showed strong but less extreme averages.

  • Affordable markets: Chicago ($320K), San Antonio ($400K), and Houston ($715K) reflected more accessible housing.

3. Market Composition

  • By property type, single-family homes dominated (74%), followed by condos (17%) and smaller shares of townhouses, multi-family, and manufactured homes.

  • By listing type, traditional agent-listed sales (FSBA) made up 91%. Distressed or alternative sales (foreclosures, auctions, FSBO) were rare.

4. Property Characteristics & Price

  • Bedrooms & Bathrooms: Prices generally increased with additional rooms, though luxury outliers dominated the 5+ category.

  • Living Area: Homes under 1,000 sq ft clustered in the $70K–$250K range; mid-sized homes (1,500–2,500 sq ft) dominated the $200K–$800K band; and very large homes (5,000+ sq ft) pushed into the multi-million tier.

5. Price Efficiency (Value per SqFt)
Condos delivered the highest value density at $1,660/sq ft, followed by single-family homes ($899/sq ft). Manufactured homes trailed at $120/sq ft, reflecting affordability positioning but lower long-term appreciation potential.

Dashboard Design

I consolidated the findings into a 1-page Power BI dashboard:

  • KPIs: Median Price, Median Price per SqFt, Median Living Area.
  • Visuals:

    • Price distribution histogram (by price buckets).

    • Bar/column charts for average price by state, bedrooms, and bathrooms.

    • Pie chart for hometype distribution.

    • Donut chart for listing type distribution.

    • Scatterplot for living area vs. price.

    • Bar chart for price per sq ft by property type.

  • Filters: State, City, HomeType, Listing Type.
Scroll to Top