Section 2: Data Sources and Data Types

Learning Objectives

By the end of this section, students will be able to:

Identify primary and secondary data sources for real estate analysis
Understand different data types and their analytical applications
Evaluate data quality and reliability for real estate projects
Access and integrate multiple data sources effectively
Handle geospatial and temporal data in real estate contexts

Introduction

Real estate analytics relies on diverse data sources, each offering unique insights into property markets. This section explores the data landscape available to real estate analysts and how to work with different data types.

The real estate industry generates massive amounts of data daily, yet most analysts access only a small fraction of available information when making investment decisions. This disconnect stems from a fundamental challenge: understanding which sources provide reliable information and how different data types require distinct analytical approaches.

Real estate data exists across hundreds of fragmented sources, each with its own access protocols, update frequencies, and quality standards. The same property might appear in multiple databases with conflicting square footage, varying sale prices, and mismatched addresses. Which source should you trust when CoStar shows 12,000 square feet but the county assessor records 11,450? This question directly impacts valuation models, investment decisions, and client recommendations.

Main Content

Primary Data Sources

Property Records:

County assessor databases
MLS (Multiple Listing Service) data
Property tax records
Deed transfers and ownership history

Market Data:

Recent sales transactions
Current listing prices
Rental rates and vacancy data
Market inventory levels

Economic Indicators:

Interest rates and mortgage data
Employment statistics
Population demographics
Income levels and growth

Secondary Data Sources

Geographic Data:

Census tract information
School district boundaries
Transportation networks
Environmental factors

Market Intelligence:

Real estate market reports
Industry publications
Economic forecasts
Regulatory changes

Alternative Data:

Satellite imagery
Social media sentiment
Foot traffic patterns
Crime statistics

Data Types in Real Estate

Structured Data:

Property characteristics (square footage, bedrooms, bathrooms)
Transaction records (price, date, location)
Financial metrics (rent, expenses, cap rates)

Unstructured Data:

Property descriptions and photos
Market commentary and reports
Social media posts about neighborhoods

Geospatial Data:

Property coordinates and boundaries
Neighborhood characteristics
Distance to amenities and services

Time Series Data:

Price trends over time
Market cycle indicators
Seasonal patterns

Data Quality Assessment

Real estate data quality varies significantly across sources. Analysts must evaluate data reliability before building models or making investment decisions.

Data Completeness Score

Formula = (Complete Records / Total Records) × 100

Where:

Complete Records = records with all required fields populated
Total Records = total number of records in dataset
Target threshold = 85% for reliable analysis

Data Accuracy Validation

Formula = (Verified Records / Sample Size) × 100

Where:

Verified Records = records confirmed against authoritative sources
Sample Size = randomly selected records for verification
Target threshold = 95% for high-confidence analysis

Common Data Quality Issues

Missing Values:

Incomplete property characteristics
Missing transaction dates
Unreported financial metrics

Data Inconsistencies:

Conflicting property measurements
Mismatched addresses across sources
Inconsistent date formats

Outlier Detection:

Unrealistic property values
Impossible square footage measurements
Anomalous market conditions

Example: Data Integration for Property Valuation

Building a comprehensive dataset for residential property valuation requires integrating multiple data sources to create a complete picture of property value drivers.

Data Integration Workflow

PROPERTY VALUATION DATASET
┌─────────────────────────────────────────────────────────────┐
│                    TARGET VARIABLE                         │
│                 Property Sale Price                         │
└─────────────────────────────────────────────────────────────┘
                              ▲
                              │
                    ┌─────────┴─────────┐
                    │                  │
                    ▼                  ▼
        ┌─────────────────┐  ┌─────────────────┐
        │   PROPERTY      │  │    LOCATION     │
        │ CHARACTERISTICS │  │      DATA       │
        └─────────────────┘  └─────────────────┘
                    │                  │
                    ▼                  ▼
        ┌─────────────────┐  ┌─────────────────┐
        │    MARKET        │  │    ECONOMIC     │
        │      DATA        │  │    CONTEXT      │
        └─────────────────┘  └─────────────────┘

Data Source Details

Property Characteristics (Internal Factors):

┌─────────────────────────────────────────┐
│ PROPERTY CHARACTERISTICS                │
├─────────────────────────────────────────┤
│ Physical Attributes:                    │
│ • Square footage: 2,400 sq ft           │
│ • Lot size: 0.25 acres                 │
│ • Year built: 2015                      │
│ • Bedrooms: 4, Bathrooms: 3            │
│ • Garage spaces: 2                      │
├─────────────────────────────────────────┤
│ Condition & Upgrades:                   │
│ • Property condition: Excellent         │
│ • Kitchen upgrades: Granite counters    │
│ • HVAC system: New (2020)               │
│ • Roof age: 8 years                     │
└─────────────────────────────────────────┘

Location Data (External Factors):

┌─────────────────────────────────────────┐
│ LOCATION DATA                           │
├─────────────────────────────────────────┤
│ Demographics:                           │
│ • Census tract: 1234.56                 │
│ • Median income: $85,000                │
│ • Population density: 2,100/sq mi       │
│ • Education level: 65% college+         │
├─────────────────────────────────────────┤
│ Amenities & Services:                   │
│ • School district rating: 8.5/10        │
│ • Distance to downtown: 3.2 miles       │
│ • Distance to airport: 12.5 miles       │
│ • Distance to park: 0.8 miles            │
└─────────────────────────────────────────┘

Market Data (Comparative Context):

┌─────────────────────────────────────────┐
│ MARKET DATA                             │
├─────────────────────────────────────────┤
│ Comparable Sales:                       │
│ • Recent sales (6 months): 12 properties│
│ • Average price: $485,000               │
│ • Price range: $420K - $550K            │
│ • Days on market: 28                    │
├─────────────────────────────────────────┤
│ Market Trends:                           │
│ • Price per sq ft: $202                 │
│ • Inventory level: 2.1 months           │
│ • Market trend: +3.2% (YoY)             │
│ • Absorption rate: 0.8 homes/month      │
└─────────────────────────────────────────┘

Economic Context (Macro Factors):

┌─────────────────────────────────────────┐
│ ECONOMIC CONTEXT                        │
├─────────────────────────────────────────┤
│ Local Economy:                          │
│ • Employment rate: 4.2%                 │
│ • Job growth: +2.1% (YoY)              │
│ • Major employers: Tech, Healthcare     │
│ • Population growth: +1.8% (YoY)        │
├─────────────────────────────────────────┤
│ Financial Environment:                  │
│ • Interest rates: 6.8% (30-year fixed)  │
│ • Mortgage availability: High            │
│ • Credit conditions: Favorable          │
│ • Inflation rate: 3.1%                  │
└─────────────────────────────────────────┘

Integration Process

Step 1: Data Collection

Property data from MLS listings
Location data from Census Bureau
Market data from local real estate boards
Economic data from Bureau of Labor Statistics

Step 2: Data Matching

Geocode addresses to coordinates
Match properties to census tracts
Align market data by geographic boundaries
Synchronize economic data by time periods

Step 3: Feature Engineering

Calculate distance-based features
Create market comparison ratios
Derive economic indicators
Standardize measurement units

Result: A comprehensive dataset where each property has 50+ features across four data categories, enabling sophisticated valuation models that capture both property-specific and market-wide factors.

Data Source Reliability Matrix

Different data sources offer varying levels of reliability for different analytical purposes:

Data Source	Completeness	Accuracy	Timeliness	Cost	Best Use Case
County Assessor	High	High	Low	Free	Property characteristics
MLS Data	Medium	High	High	Medium	Market transactions
Census Data	High	High	Low	Free	Demographics
Alternative Data	Low	Medium	High	High	Market sentiment