top of page

Driving the Price

In my Spring 2026 semester at USC, I took an Enterprise Data Analytics class led by a Principal Architect at Google and the Founder of Global Student Startup Foundation. For our final project, we were tasked with performing end-to-end data analysis on a large dataset, visualizing our data, conducting predictive analysis, and synthesizing our results into actionable business recommendations. Our final deliverable was a process document that demonstrated all of the skills we had honed in the class.​

​

I chose to analyze Uber and Lyft Rideshare data from Boston, MA. I leveraged the USC ADAM Methodology to clean, transform, and engineer features in Google Colab and Power BI. My thorough understanding of data analysis enabled me to analyze this large dataset and develop business model recommendations that addressed the guiding questions I conceptualized.

​​

Project Focus

End-to-End Data Analysis,

Data Visualization,

Strategic Storytelling

Role

Data Analyst,

Performance Analyst

Timeline

Spring 2026

Problem Space

​Pricing is one of the most unpredictable parts of the rideshare experience today. Fares fluctuate based on distance, time of day, weather, and surge demand, making it difficult for both consumers and platform operators to predict what a ride will actually cost. ​​

​

From a business analyst's perspective, I wanted to understand what was really driving these price differences and which platform, Uber

Screenshot 2026-06-17 at 5.14.00 PM.png

or Lyft, delivered better value to consumers in a major city. This project set out to answer that question through an end-to-end analysis of real-world ride data from Boston, Massachusetts.

Design Questions​

​To begin, I selected design questions to guide my methodology and to help create prompts and figures in Google Colab and Power BI that address my problem space directly.​

1. What ride tier has the highest average price, and how do Uber and Lyft Premium options compare?

​

2. How do ride prices fluctuate throughout the day, and do Uber or Lyft show stronger price variation by hour?

​

3. Is there a statistically significant price difference between Uber and Lyft for similar rides?

​

4. Which rideshare platform offers consumers the best value per mile traveled, Uber or Lyft?

​

5. Can we accurately predict ride prices based on distance, surge multipliers, and citywide temperatures in Boston, Massachusetts?

Analysis & Tools

With the dataset prepared, I used Python in Google Colab to explore pricing patterns, visualizing price distributions, hourly trends, and the relationship between distance and price. I built a linear regression model to test whether ride prices could be predicted from distance, surge multipliers, and weather. I then exported the cleaned data to Power BI to build an interactive dashboard, complete with custom DAX measures, KPI cards, and slicers that let users isolate and compare Uber and Lyft pricing side by side. ​

Screenshot 2026-06-17 at 9.10.30 PM.png

Key Findings

The analysis revealed clear, consistent differences between the two platforms:

  • Lyft dominated the premium tier, with its Lux Black XL averaging $31.04 per ride compared to Uber's top tier at $29.91

  • Uber held a pricing edge across nearly every hour of the day and at the economy end of the spectrum

  • Across the full dataset, Lyft's pricing also proved slightly more sensitive to distance (a correlation of 0.34 versus Uber's 0.32)

    • This means the price gap between platforms tends to widen on longer rides

  • The findings point to one consistent conclusion: Uber offers better overall value across car tiers, time of day, and distance traveled.

Screenshot 2026-06-17 at 5.16.16 PM.png

Reflection

This project pushed me to think like both a data analyst and a business strategist at once. Cleaning and engineering a 50,000-row dataset taught me how much real analysis goes into it before you ever build a chart. Building the Power BI dashboard forced me to think about how a non-technical stakeholder would actually use my insights and how these visuals aid in offering business recommendations. More than anything, this project reflects the kind of work I want to keep doing: using data to answer real business questions and turning those findings into recommendations that drive growth and increase profitability.

If you'd like to dive deeper, you can check out the full paper below, which walks through my complete methodology in Google Colab and Power BI, the key questions I set out to answer, and the final findings and business recommendations that came out of the analysis.

bottom of page