As a Business Analyst, you receive a dataset from a business unit with missing values and inconsistencies. How would you approach this to ensure reliable insights?

Situation

In a university project, I analyzed survey data for a local utility provider that had incomplete records and conflicting entries.

Task

I had to turn this unreliable dataset into a clean foundation for presenting service improvement recommendations.

Action

I first assessed the extent of missing and inconsistent data by counting gaps and checking for patterns. For missing values, I used median imputation for numeric fields and mode for categorical ones. For inconsistencies, I cross-referenced with source logs where possible and standardized formats. I documented each step so the business unit could replicate the process. Then I created a dashboard highlighting key trends with caveats about the data quality.

Result

My cleaned dataset let the team correctly identify billing issues as the top complaint, and the business unit implemented my recommended tracking system.