Scribd
Upload
3 views

Big Data Case Study (1)

This case study explores how an e-commerce analytics firm utilized big data to analyze user behavior on Facebook Marketplace, aiming to enhance the platform and assist sellers. By processing petabytes of user interaction data, the firm identified buying patterns, listing performance factors, and regional preferences, leading to actionable recommendations that resulted in a 30% increase in sales and a 20% boost in user satisfaction. Key challenges included managing large data volumes while ensuring user privacy and data quality.

Uploaded by

sahil.s210958101
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
3 views

Big Data Case Study (1)

This case study explores how an e-commerce analytics firm utilized big data to analyze user behavior on Facebook Marketplace, aiming to enhance the platform and assist sellers. By processing petabytes of user interaction data, the firm identified buying patterns, listing performance factors, and regional preferences, leading to actionable recommendations that resulted in a 30% increase in sales and a 20% boost in user satisfaction. Key challenges included managing large data volumes while ensuring user privacy and data quality.

Uploaded by

sahil.s210958101
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 2

Big Data Case Study: Optimizing Facebook Marketplace User Experience

1. Introduction

● Background: An e-commerce analytics firm aiming to understand user behavior on


Facebook Marketplace to help improve the platform and assist sellers in optimizing
their listings.
● Objective: Utilize big data analytics to analyze Facebook Marketplace user behavior,
buying patterns, and seller performance.

2. Problem Statement

● Challenges: The firm needs to process and analyze large volumes of structured and
unstructured data generated by millions of Facebook Marketplace users and listings.
● Impact: Difficulty in providing accurate, actionable insights for platform
improvements and seller strategies.

3. Data Collection

● Sources: User interaction data from Facebook Marketplace including views,


messages, purchases, listings, user feedback, and demographic information.
● Volume: Petabytes of data collected over two years, encompassing millions of listings
and user interactions.

4. Data Processing

● Technologies Used:
○ Storage: Amazon S3 for scalable cloud storage.
○ Processing: Apache Spark for distributed data processing.
○ Ingestion: AWS Kinesis for real-time data ingestion.
● Methods: Data extraction using Facebook Marketplace API, data cleaning, and
feature extraction (e.g., listing popularity, buyer behavior metrics, seller performance
indicators).

5. Data Analysis

● Techniques:
○ Descriptive Analytics: To understand general marketplace trends and user
behavior patterns.
○ Predictive Analytics: To forecast demand for different product categories and
price points.
○ Sentiment Analysis: To analyze user feedback and comments.
● Tools: Python (Pandas, NumPy, NLTK), TensorFlow for machine learning, Tableau
for visualization.

6. Findings

● User Buying Patterns: Identified peak times for purchases and high-activity
categories.
● Listing Performance: Determined factors that contribute to successful listings (e.g.,
quality of images, detailed descriptions).
● Geographic Insights: Found significant differences in product preferences and
pricing across different regions.
● User Segments: Categorized users into segments such as "frequent buyers,"
"occasional sellers," and "power sellers."

7. Implementation

● Actionable Steps:
○ Listing Optimization: Advised sellers on best practices for creating attractive
listings.
○ Pricing Strategy: Developed a dynamic pricing model based on demand and
regional preferences.
○ User Experience: Recommended platform improvements to enhance search
and discovery features.
● Challenges: Balancing user privacy with data analysis needs, ensuring fair
marketplace practices.

8. Results

● Outcomes:
○ Increased Sales: Sellers implementing the recommendations saw a 30%
increase in successful transactions.
○ Improved User Satisfaction: Platform changes resulted in a 20% increase in
user satisfaction scores.
○ Reduced Time-to-Sell: Average time to sell items decreased by 25% for
optimized listings.

9. Conclusion

● Success Factors: Importance of real-time analytics, regional customization, and user-


centric design.
● Lessons Learned: Initial challenges in handling the diverse nature of marketplace
listings and ensuring data quality across various product categories.

Example Visualizations and Reports

1. Heat Map: Displaying peak times and days for marketplace activity across different
product categories.
2. Scatter Plot: Showing the relationship between listing quality factors (e.g., number of
photos, description length) and time-to-sell.
3. Geographic Dashboard: Visualizing regional preferences for different product
categories and price points.

You might also like