There are companies that have no data quality initiative, and truly do believe that if they see no data problem. In effect, they say that if it does not interfere with day-to-day business, then there is no data quality problem. From what I have seen in my consulting experience, it usually doesn't take long for the data quality issues to rise to surface, and cause undue chaos and rework.
So, assuming a company has a data initiative that has profiled and corrected (where appropriate) the data, creating on-going data quality is the next step to true data quality management. (Read more about how to employ a data quality strategy in this white paper).
On-going data quality is the activity of tracking, managing, and monitoring data quality for the organization. Most purchase a tool or extension to an existing data quality tool to accomplish on-going data quality.
Some of the characteristics of on-going data quality would include:
- Ability to set a target for measuring data quality as it comes into and move around the organization applications (the data lifecycle). This involves checking for completeness, accuracy, and integrity of the data
- Reporting and/or a dashboard to communicate data quality metrics over time. Consider improvement overtime of certain data quality metrics (for example – a scorecard showing address correction and verification or name standardization over time)
If you do not have an on-going data quality initiative or a data quality initiative as part of data management, consider the following:
- Upper management backing is mandatory. This requires education of upper management so they know the gains that can be achieved through a data quality initiative. Sometimes one can only count what it has cost the corporation, in the past, for inaccurate data and the issue it caused in a process. For example, how much rework was caused and the cost of that rework based on a specific data quality issue.
- Software for data quality and on-going data quality.
- Team that includes input from business users.
- Metrics to measure.
