Data Profiling 101

December 28, 2015 Kristin Mager

By definition, data profiling measures the data one has available in an existing source and determines the accuracy, completeness, and other statistics and information relating to one’s data.  The importance of measuring this is to learn if existing data can be easily used, to provide metrics on the quality, and to assess the risk involved in integrating the data.  With data profiling, one can understand the challenges early within a project and avoid delays and increasing costs.  Data profiling also allows the user to have an overall view of the data, which helps in improving the quality for data management.

Why should you profile your data and why does it matter?

The most important reason for profiling is to help an organization understand its data.  Batch, structured and non-structured data, or any useful data asset can be profiled.  Profiling allows an organization to evaluate massive amounts of data quickly in an organized and repeatable process.  This analysis will help discover the data that is accessible and its characteristics.  By determining the validity of the data one will be able to determine its usefulness.

Data profiling is important to companies who are faced with litigation.  Poor quality data or inaccurate data can result in legal or regulatory implications.  By profiling data, one can better understand the overall quality of the data and reduce the risk of litigation issues in the future.  In short, if you don’t understand your data, there is no way to fix it.

Before beginning any data initiative, an organization should ask the following questions:

  1. Do you trust the quality of your data?
  2. Does the data meet the needs of your project?
  3. Can you access the data?
  4. Does the data conform to business rules?

Not understanding the answers to these questions can result in project failures and budget overages.  Data problems can include, but are not limited to: data anomalies, data inconsistencies and duplicated data.  These issues can hinder the bottom line and the amount of time it takes to complete a project.

Information Technology departments are usually tasked with the job of profiling data because they understand the structures and can identify the unique key fields.  Having input from business users or subject matter experts though is also very beneficial because they can provide insight about the content and accuracy of the data.  Additionally, they can determine the most useful data and where it may be stored.  Having a collaboration from both sides will provide the most accuracy for an organization.

Before the next data undertaking, why not take the guess work out of the project from the very beginning?  Profiling can help to plan and scope a project, assess data quality and design new systems.  It will determine what data and resources will be involved and how accurate the data is in the system.  Knowing the quality of data upfront will help to keep accurate project time lines and eliminate surprises when the data at hand is incomplete.

The post Data Profiling 101 appeared first on Sherpa Software.

Read more...

Previous Article
Sherpa’s Top 10 Highlights of 2015

A new year is upon us which means it’s time to reflect on the year that was. For Sherpa Software, 2015 was ...

Next Article
November 2015 Newsletter

The post November 2015 Newsletter appeared first on Sherpa Software.