TK
Menu
← Research

Journal

A Reproducible Data-Quality and Exception-Management Framework for Resource-Constrained Organizations

A practical framework for helping small businesses and nonprofit organizations identify data-quality issues, prioritize high-risk exceptions, and manage corrective action with limited resources.

Allen Teerahumba, Emmanuel Hagan, Flora Phiri, Trevor Kauyu, Munashe Naphtali Mupa · September 17, 2026 · IRE Journals

Overview

Reliable data is essential for operational reporting, but small businesses and nonprofit organizations often lack dedicated data-quality teams and enterprise monitoring tools. This article introduces a Reproducible Data-Quality and Exception-Management Framework designed for organizations working within these constraints.

How the Framework Works

The framework connects data-quality checks with a structured process for:

  • Detecting issues
  • Assessing risk
  • Identifying root causes
  • Assigning responsibility
  • Managing remediation
  • Verifying closure

Its supporting tools include an exception register, a risk-scoring model, a minimum viable dashboard, and a closure playbook.

Study and Findings

The study demonstrates the approach using the Brazilian E-Commerce Public Dataset by Olist, covering 99,441 orders. Twelve order-level controls, supported by additional table-level tests, assess issues such as missing information, duplicate records, inconsistent dates, reconciliation differences, and unusual values.

Overall, 21.84% of orders triggered at least one rule, while 6.57% reached a risk score of four or more. Flagged anomalies were treated as candidates for review rather than automatically classified as errors.

Simulation and Limitations

A queue simulation also explored how prioritizing higher-risk exceptions could shorten their completion times without reducing the total workload. These results were based on stated service-time assumptions, not observed organizational performance; real-world pilots are needed to establish remediation benefits.

Practical Contribution

The article provides a reproducible starting point for organizations seeking a more structured, auditable approach to data quality and exception management without relying on extensive specialist resources.

Read the full article on IRE Journals