With the ever-increasing demand from the business to make data-driven decisions, IT organizations are struggling to ensure quality in the large volumes and variety of data. IT consistently hears that reports the business relies on are not accurate, and they are required to perform time-intensive and costly manual fixes to the organization’s data. In such cases, business users often bypass IT and establish their own data silos to manage quality, but these workarounds compound systemic data quality issues by further fragmenting data consistency and governance process within multiple silos. The consequences of allowing this trend to continue are severe, and include:

  • Unwillingness to respond quickly to changing market conditions due to lack of confidence in supporting data
  • Acting based on supporting data which is either incomplete or incorrect
  • Regulatory fines or expensive refiling of financials
  • Inability to implement and utilize big data analytics tools

Prevention is Cheaper than Treatment

The biggest mistake an organization can make in addressing poor data quality is to approach the problem as something that can be solved completely through the application of a series of small fixes. The first issue with such an approach is the concept that this problem can be solved completely in the first place. The fact of the matter is: 100% pure data quality does not exist. Instead of striving for an unattainable goal, the focus should be on reaching a level quality that is satisfactory to the business. By establishing this benchmark for success prior to embarking on a data cleansing initiative, the entirety of the program will feel more approachable and attainable. This in turn will increase buy-in, allow for more targeted allocation of budget, and reduce the risk of the project not even getting off the ground because it feels so impossible.

The second issue with this approach is the notion that this is something that can be permanently resolved with the application of a series of small fixes. While this might seem like the cheaper and easier approach, the reality is that treating the issue is different from curing it, and over the long-term prevention is multitudes cheaper than remediation. The first steps to achieving this foundational shift if an organizations data strategy include:

  • Address and resolve data quality controls at the point of ingestion and at the process level to ensure sustainable improvement
  • Embed a data governance and controls strategy into the enterprise IT data governance framework
  • Align data quality initiatives with overall business objectives
  • Develop a prioritized data quality improvement project roadmap