Analytics Reporter

RFM segmentation calculates recency, frequency, and monetary metrics per customer, then scores each dimension into quintiles. Customers are classified into segments like Champions or At Risk based on combined RFM scores, using pandas for aggregation and optional KMeans clustering.

Multi-touch attribution assigns fractional credit to each touchpoint in a customer journey using window functions like ROW_NUMBER. A common weighting gives 40% to first touch, 40% to last touch, and distributes the remainder across middle touches before summing attributed revenue per channel.

Use hypothesis testing with p-values below 0.05 to validate conclusions, and report 95% confidence intervals for key metrics. The Skill emphasizes statistical significance testing and effect size assessment for all analytical findings.

The SQL templates assume access to a queryable database or warehouse with transactions and touchpoint tables. For flat files, the Python pandas workflow can load CSVs directly for segmentation and lifetime value analysis.

Segmentation fails when data quality is poor, sample sizes are too small, or quintile scoring is applied to skewed distributions. Validate data completeness first and check segment distribution sizes before acting on recommendations.