Customer Behavior Analysis

An end-to-end analysis of customer shopping behavior using transactional data across product categories, surfacing insights into spending patterns, customer segments, product preferences, and subscription behavior to guide strategic business decisions.

The brief.

The business needed to understand who their best customers were, how discounts and subscriptions influenced spend, and which product categories deserved investment. The brief called for a clean dataset, sharp segmentation, and SQL-driven answers to ten focused business questions.

How it was built.

1. STEP 01Imported and cleaned the dataset in Python with pandas; filled null review ratings using the median rating per product.
2. STEP 02Engineered new features: age_group buckets and purchase_frequency_days; dropped redundant promo_code_used (duplicate of discount_applied).
3. STEP 03Loaded the prepared data into SQL Server for structured analysis.
4. STEP 04Answered ten business questions with SQL, covering revenue by gender, discount-driven spend, top-rated products, subscriber spend, discount penetration, customer segmentation (New / Returning / Loyal), top products per category, repeat-buyer subscription likelihood, and revenue by age group.

What it moved.

• →Quantified revenue contribution across gender and age cohorts, exposing the highest-value segments.
• →Showed that subscribers spend meaningfully more on average than non-subscribers, validating the program.
• →Segmented customers into New, Returning, and Loyal groups with clear counts to drive lifecycle marketing.
• →Identified the top 3 products per category and the top 5 by review rating to focus merchandising.

Tools used.