Payroll Integrity Analytics ("Net Savings" Initiative)

Overview

An end-to-end analytics initiative that quantified, for the first time, how much of an internal cost-saving measure was actually sticking.

Business Problem

The company was deducting payroll for improperly-approved absences, but a large share of those deductions were later reversed — nobody was measuring how much of the "savings" actually stuck.

Objective

Build a reliable, repeatable way to measure the true net effect of payroll deductions, and identify why so many were being reversed.

Solution

Built a consolidated PySpark pipeline joining multiple internal data sources to calculate, for the first time, a "Net Savings" metric (Initial Savings − Reversals). Defined the identification methodology, an effectiveness funnel, and a root-cause analysis of the reversals.

Business Impact

Found that roughly three-quarters of an identified six-figure payroll deduction was being reversed the following month, with a confirmed net retention rate of 23.6%. Delivered a root-cause analysis and a 30/60/90-day action plan to leadership. A follow-up extension found that flagged employees had a meaningfully elevated termination rate, turning the tool into an early risk signal as well as a cost-recovery one.

Technologies

• PySpark
• Databricks