Retail & Marketing Analytics
Which customers, categories, and regions should a retailer prioritize, and is discounting associated with stronger sales?
01 / THE QUESTION
Which customers, categories, and regions should a retailer prioritize, and is discounting associated with stronger sales?
02 / THE APPROACH
Clean and analyze 10,000 retail orders using Python / Pandas, RFM scores, K-Means clustering, customer lifetime value, and cohort retention. Present category and customer patterns in Power BI.
03 / THE RESULT
A documented set of customer win-back, regional investigation, and discount-testing recommendations grounded in the project dataset.
[ Project insights ]
EVIDENCE / FINDINGSTwo categories carry the sales
Electronics ($328K) and Office Supplies ($321K) account for roughly 60% of revenue in the analyzed dataset, highlighting category concentration.
A clear win-back audience
278 At Risk customers — 14% of the customer base — have not ordered in 126+ days. Their roughly $194K in past revenue supports prioritizing a win-back test, not claiming guaranteed future sales.
Investigate before spending more
The South earns 33% less than the East, with fewer customers and lower spend. Discounts correlate only 0.00–0.02 with sales or profit; that association alone does not prove their causal impact.
CONTEXT / LIMITATIONS
These findings describe the repository’s sample dataset, not verified commercial outcomes. Regional campaigns and discount changes are recommendations; no campaign uplift or causal experiment is documented.