Project 4: Business Intelligence & Data Visualization

Building an Airbnb Analytics Dashboard & Data Model for Business Decision-Making

Temp Text

The Problem/Opportunity:

A company considering the acquisition of Airbnb required a complete analytical solution that would allow stakeholders to better understand the platform’s operational performance, pricing behavior, booking activity, and customer engagement.

To support strategic decision-making, the organization needed:

  • Clean and reliable datasets
  • A structured analytical data model/li>
  • Interactive dashboards and reports
  • Clear business insights accessible to non-technical stakeholders

The project focused not only on dashboard development, but also on the full preparation of the analytical environment, including:

  • Data loading and validation
  • Data cleaning and transformation
  • Data modeling using Fact & Dimension tables
  • Relationship management
  • Cross-validation and data reliability checks
  • Interactive dashboard creation and storytelling

The dashboard was designed to answer key business questions such as:

  • Which hosts and listings perform the best?
  • Which locations show the highest demand?
  • What are the average rental prices and booking trends over time?
  • Which opportunities or operational risks can be identified from the data?
  • How do reviews, bookings, pricing, and host activity evolve over time?

A major focus was placed on usability and storytelling so that stakeholders could quickly understand the business performance and make informed decisions.

The key stakeholders include:

  • Executive stakeholders and decision-makers
  • Strategy and operations teams
  • Pricing and performance teams
  • Data analysts and BI teams

Key Insights:

Data Loading & Initial Exploration

  • Loaded Airbnb datasets from CSV files and Data Warehouse tables into Power BI
  • Imported data related to::
    • Listings
    • Hosts
    • Reviews
    • Bookings
  • Opened and managed transformations through Power Query
  • Explored datasets to identify:
    • Missing values
    • Incorrect data types
    • Formatting inconsistencies
    • Outliers and data quality issues
  • Validated all data types using the provided data dictionary
  • Examples included:
    • Numbers → Whole/Decimal
    • Dates → Date
    • Text fields → Text
  • Additional preparation tasks included:
    • Converting neighborhood names to proper Greek text format
    • Renaming columns into business-friendly naming conventions
    • Applying standardized prefixes such as:
      • listing_*
      • host_*
      • review_*
      • booking_*
    • This improved readability, maintainability, and model clarity.

Data Cleaning & Transformation

Performed extensive cleaning and transformation processes in Power Query to ensure data reliability and analytical usability.

Missing Value Handling
  • Applied different handling strategies depending on the data type:
    • Numerical Fields
      • Avoided replacing missing prices or quantities with zero where it would distort business meaning
      • Used null handling or average-based imputation where appropriate
    • Text Fields
      • Standardized missing text values using:
        • “Unknown”
        • “N/A”
    • Date Fields
      • Applied placeholder dates where necessary for consistency
      Formatting & Type Transformations
    • Cleaned columns containing symbols or formatting inconsistencies, such as:
      • Currency symbols ($)
      • Percentages (%)
      • Text-based numeric values
    • Converted them into proper numeric formats for analytics and KPI calculations.

      • Advanced Power Query Transformations

      Host Data Transformation (DimHost)
    • Split host_location into:
      • City
      • Country
    • Standardized inconsistent location naming conventions:
      • Athens
      • Αθήνα
      • Athina
    • Used:
      • Replace Values
      • Column Quality
      • Column Distribution
      • Column Profile

      to identify and correct data inconsistencies and transformation errors.

      Listing Information Transformation (DimListingInfo)

      Performed multiple business-focused transformations, including:

      • Converted bathrooms_text into numeric values
        • Example:
          • “1 bath” → 1
          • “1.5 baths” → 1.5
    • Renamed the field to bathrooms
    • Removed redundant columns
    • Applied business validation to:
      • minimum_nights
      • maximum_nights
    • Restricted unrealistic values above 365 days

      • Created calculated business metrics:

      • Minimum Potential Profit
      • Maximum Potential Profit

      using:

      • Price × Minimum Nights
      • Price × Maximum Nights
      FactReviews & FactBookings Enhancements
    • Merged booking-related availability data into review tables using Power Query Merge operations
    • Used listing_id as the joining key
    • Applied additional validations for booking duration fields

      • Binary Field Optimization

      Converted:

    • Yes/No
    • True/False

      • fields into:

    • 1/0

      • This improved:

    • Data model efficiency
    • KPI calculations
    • Compression and performance within Power BI

      • DimListingAmenities Creation

      Built a dedicated amenities dimension table for granular amenity analysis.

      Transformation Process
    • Duplicated the listing table in Power Query
    • Kept only:
      • listing_id
      • amenities
    • Split amenity values into separate rows using delimiter-based normalization
    • Expanded amenities into one row per listing/amenity combination

      • Performed additional cleaning:

    • Removed quotes and brackets
    • Applied trimming
    • Standardized formatting for consistency

      • This enabled advanced amenity-based analysis such as:

    • Listings with WiFi
    • Most common amenities
    • Premium property feature analysis

      • Calendar Table Development

      Created a dedicated Calendar table to support time intelligence and trend analysis.

    • Included:
      • Year
      • Month
      • Quarter
      • Day
      • Month Name
    • The calendar table enabled:
      • Monthly booking analysis
      • Seasonal trend analysis
      • Time-based DAX calculations
      • Year-over-year comparisons

      Data Modeling & Relationships

      Built a structured Power BI data model using Fact and Dimension tables.

      Model Enhancements Included:
    • Defining Primary/Key columns
    • Creating relationships between entities
    • Validating:
      • Cardinality
      • Cross-filter direction
    • Configuring data categories:
      • URLs
      • Geographic fields
    • Formatting measures and attributes appropriately
    • Adjusting summarization behavior for analytical accuracy

      • Resolved modeling challenges related to:

    • Multiple date fields
    • Calendar relationships
    • Different business contexts for dates
      • Host registration dates
      • Booking dates

      Cross Validation & Data Reliability

      Performed extensive cross-validation between:

    • Raw CSV/Excel data
    • Power BI model outputs

      • Validation checks included:

    • Record counts
    • Sums
    • Averages
    • Min/Max values

      • Used Excel formulas such as:

    • COUNT
    • SUM
    • AVERAGE
    • MIN/MAX

      • Compared results against Power BI visuals and measures to verify consistency and reliability.

      Additional validation included:

    • Identifying transformation errors
    • Detecting incorrect null handling
    • Validating merges and filters
    • Ensuring raw data fidelity

      • This process ensured the final analytical model was trustworthy and suitable for reporting and dashboard creation.

      Dashboard Design & User Experience

      Designed the report using a clean, executive-friendly layout focused on storytelling and usability.

      Key Design Elements
    • High-resolution report pages (1920×1080)
    • Airbnb-inspired color palette
    • Consistent typography and spacing
    • KPI cards positioned at the top of each page
    • Interactive slicers for:
      • Area
      • Room Type
      • Price
      • Host
    • Slicer synchronization across report pages
    • Conditional formatting and layout optimization
    • Visual interaction management to avoid unnecessary cross-filtering issues

      • DAX Calculations & Business Metrics

      Created multiple DAX calculated columns and measures.

      Calculated Columns

      Examples included:

    • Price category segmentation
    • Quality/rating categories

      • Measures

      Developed analytical KPIs such as:

    • Total listings
    • Total bookings
    • Total reviews
    • Average rental price
    • Average stay duration
    • Adjusted price trends
    • Year-over-year price comparisons
    • Host contribution metrics

      • Report Pages & Analysis

      1. Overview Dashboard

      Created an executive summary page displaying:

    • Total listings
    • Total hosts
    • Total bookings
    • Total reviews
    • Average rental price
    • High-level trends and KPIs

      • Focused on providing a fast and intuitive business overview.

      2. Host Analysis

      Analyzed:

    • Top 10 hosts by listings
    • Top 10 hosts by review scores
    • Host contribution to the platform
    • Host engagement and pricing behavior

      • Designed the page to resemble a “host profile” experience.

      3. Property Analysis

      Analyzed:

    • Room/property types
    • Average rental prices
    • Geographic distribution
    • Review performance
    • Property popularity and pricing relationships

      • Included interactive maps and property-focused visuals.

      4. Booking Analysis

      Analyzed:

    • Booking trends by month/year/quarter
    • Average stay duration
    • Seasonal demand
    • Adjusted pricing trends
    • Booking performance by location and host

      • Included detailed booking tables for operational analysis.

      Business Insights & Findings

      The dashboard enabled the identification of multiple strategic insights, including:

    • High-demand locations and booking periods
    • Top-performing hosts and premium properties
    • Pricing behavior across property categories
    • Relationships between review scores and pricing
    • Seasonal booking trends
    • Amenity-driven property performance
    • Opportunities for pricing optimization and business growth

      • The Outcome:

      This project resulted in a fully interactive and business-ready Power BI analytical solution combining:

    • Data preparation
    • Data cleaning
    • Transformation
    • Data modeling
    • Validation
    • Dashboard development
    • Business storytelling

      • The final solution successfully:

    • Transformed raw Airbnb data into a reliable analytical model
    • Applied Data Warehousing and Business Intelligence principles
    • Delivered interactive dashboards for business stakeholders
    • Improved data quality and reporting reliability
    • Enabled strategic business analysis through KPIs and trends
    • Created scalable foundations for future reporting and analytics initiatives

      • The project demonstrates practical experience in:

    • Power BI
    • Power Query
    • DAX
    • Data Cleaning & Transformation
    • Data Modeling
    • Fact & Dimension Architecture
    • Cross Validation
    • Business Intelligence
    • Dashboard Development
    • Data Visualization & Storytelling
    • KPI Development
    • Analytical Thinking

      • Overall, this project showcases the ability to build an end-to-end Business Intelligence solution similar to what a professional Data Analyst or BI Developer would create in a real business environment.

      Tools & Technologies

    • Microsoft Power BI
    • Power Query
    • DAX
    • Data Modeling
    • Data Cleaning & Transformation/li>
    • Business Intelligence Reporting
    • Interactive Dashboards
    • SQL-based Data Sources
    • Excel (Cross Validation)
    • Fact & Dimension Modeling
    • Time Intelligence Analysis