Nobody likes audits. But ensuring data quality is foundational to all business intelligence endeavors. How can a business make quality decisions with poor quality data?
A tribe member in one of my weekly #BIWisdom tweetchats mentioned he read an article about a survey finding that 80 percent of companies claim they deal with poor data quality, yet less than 52 percent of those companies consider doing a data quality audit.
Those statistics sparked a boisterous bunch of tweets with opinions and questions from BI users, consultants and vendors including:
- “How could the respondents in that study claim poor quality if they didn’t audit it? Is quality simply a matter of perception?”
- “No, it’s not perception. Data quality has been proven to be a dependency for success or failure in BI initiatives – whether or not it’s called out as the underlying factor.”
- “So shouldn’t the precursor to any BI initiative be a data quality audit? Otherwise, the outcomes may be flawed.”
- “Data quality is important, but it shouldn’t impede the initial progress of self-service data discovery. Then when you have potentially interesting through self-service data discovery, then you add data quality governance.”
- “It’s impossible to always have good data on the front end; there are too many exceptions and shortcut coding.”
- “But you can’t edit all the data. What about unstructured data and also external data?”
- “The idea is to identify the Key Data Elements and focus on those during the audit.”
- “The necessity for data quality should depend on the impact of the potential decision.”
- “Who is responsible for pushing the audit – IT, line of business, top executives?”
- “LOBs need to be able to see where strategic failures come from. If data quality is suspect, then they should demand an audit.”
On the other hand, several in the group pointed out that poor data quality is different from using wrong data for decision making. One tweeted that discovering low data quality can be insightful in itself. Another added that it can be helpful in identifying gaps in process control.
I pictured their heads shaking in despair as the #BIWisdom tweetchat hour ended and they all agreed on one point: For the audit to be effective, an organization needs agreed definitions of data quality. But how many businesses have that?
Is an effective data governance program in place in your organization?
Bottom line: In our Wisdom of Crowds® Business Intelligence Market Study published in May 2014, seven percent of the surveyed respondents reported the state of data in their organization has “inconsistent data sources and information is generally unreliable and distrusted.” Only 31 percent reported their data as “truth” with common filters, rules and semantics. How does your organization compare?
Correcting data after it has been created is more costly than implementing upstream controls at the point of data entry – some estimate it costs 10x more to deal with poor quality data downstream.
Consequently, some large organizations are appointing chief data officers, especially in banks, healthcare and government, which are saddled with regulations governing patient and customer data. But the CDO is a new leadership role, and it will take some time for it to make a significant impact. And success with this strategy requires that the CDO be cross-functional and an expert at data integration.
My opinion: A Business Intelligence Competency Center (BICC) can assist with data governance and help bridge the gap when there is conflicting data among departments or business units.
Howard Dresner is president, founder and chief research officer at Dresner Advisory Services, LLC, an independent advisory firm. He is one of the foremost thought leaders in Business Intelligence and Performance Management, having coined the term “Business Intelligence” in 1989. He has published two books on the subject, The Performance Management Revolution — Business Results through Insight and Action, and Profiles in Performance — Business Intelligence Journeys and the Roadmap for Change. He hosts a weekly tweet chat (#BIWisdom) on Twitter each Friday. Prior to Dresner Advisory Services, Howard served as chief strategy officer at Hyperion Solutions and was a research fellow at Gartner, where he led its Business Intelligence research practice for 13 years.