Predicting High-Spending Customers with Logistic Regression

Executive Summary

This case study examines how retail businesses can use predictive modeling to identify high-spending customers. Using a logistic regression model, the study analyzes customer demographics and purchasing patterns to predict which customers are likely to spend over $1,000. The findings offer actionable insights for targeting high-value customers through personalized promotions, loyalty programs, and marketing campaigns.

Problem Statement

Retailers face challenges in identifying high-spending customers who can be targeted for promotions and loyalty programs. This case study aims to predict high spenders using customer demographic data and purchasing behavior, helping businesses optimize their marketing strategies and improve revenue.

Approach

Dataset: The dataset, titled "Retail Sales and Customer Demographics Dataset," includes attributes like age, gender, product category, and total amount spent.

Analytic Framework:

• Logistic regression was used to predict high spenders based on age, gender, and product preferences. SQL queries were employed to analyze spending behavior.
• The logistic regression model was trained using customer demographic and purchasing data, achieving high precision and accuracy. The model predicted whether a customer would spend more than $1,000 based on input features such as gender, age group, and product category.

Results

High Spender Breakdown by Gender

• Male: 15.71% of male customers were predicted to be high spenders.
• Female: 14.9% of female customers were predicted to be high spenders.

Insight: Both genders have a similar likelihood of being high spenders, suggesting businesses can target both groups effectively.

High Spender Breakdown by Age Group

• Millennials: Lead with 57 predicted high spenders.
• Gen X: Follow closely with 45 high spenders.

Insight: Millennials and Gen X are the most likely age groups to be high spenders. Businesses should focus on these groups for promotions.

High Spender Breakdown by Product Category

• Electronics: 53 high spenders.
• Clothing and Beauty: 52 and 48 high spenders, respectively.

Insight: Electronics has the highest number of high spenders, making it a key category for high-value marketing efforts.

View the full visualization here:

Key Insights

• Demographic Targeting: Millennials and Gen X customers, especially males shopping for electronics, are prime candidates for high-value promotions.
• Product Category Focus: Electronics, Beauty, and Clothing should be prioritized for high-spending customers, with Millennials showing a particular preference for Clothing and Beauty.
• Gender Balance: Males and females are almost equally likely to be high spenders, suggesting a balanced marketing approach.

Recommendations

• Target Millennials and Gen X for Promotions: Businesses should focus their marketing strategies on Millennials and Gen X, as they are most likely to be high spenders. Customized promotions based on their product preferences (Electronics, Beauty, Clothing) will drive better engagement.
• Leverage Product-Specific Promotions: Electronics leads in high spenders, making it a lucrative category for high-ticket promotions. Similarly, targeted campaigns in Beauty and Clothing can also capture a significant share of high-spending customers.
• Develop Personalized Campaigns for Both Genders: With high spender rates almost equal between males and females, businesses should develop gender-neutral promotions while slightly prioritizing male customers for high-value campaigns.

Conclusion

By using logistic regression to predict high spenders, businesses can strategically target their most valuable customers. The model's high accuracy and clear insights into customer demographics and preferences make it an essential tool for developing personalized marketing campaigns. Focused efforts on Millennials, Gen X, and product categories like Electronics, Beauty, and Clothing will yield the best results for boosting revenue.