Contradicting sales reports, inaccurate and outdated customer information, incomplete transaction records and redundant data are all telltale signs that you have dirty data on your hands.
Poor data quality, or dirty data, can affect performance and productivity, especially if business decisions are being based on it. Industry experts often say that data is a company’s biggest asset. Therefore, there is significant business value associated with keeping corporate data clean.
Up to 75 per cent of information workers admitted they have made wrong business decisions because of inaccurate, incorrect or incomplete corporate information, according to a recent survey by Harris Interactive. It polled employees from the U.S., Great Britain, France and Germany.
“Information workers”, as defined by the surveyors, are employees who use data from various applications, such as spreadsheets, business intelligence reports and executive dashboards, to make decisions.
Often, customers’ perceptions can be key indicators of the level of a firm’s data quality.
“In a customer-facing type of business, there tends to be a general feeling among customers that the enterprise (they’re dealing with) doesn’t know them or understand them,” explained Royce Bell, CEO of Accenture Information Management Services in London.
Consumer-focused enterprises today are always trying to find ways to intelligently predict the behaviours of specific group of customers based on their past activities, said Bell. This can prove difficult when working with unclean data and may lead to misleading information about the customers.
Poor data quality also has a significant impact on employee productivity, according to the survey. Information workers in the U.S. spend an average of 12 hours per week, or 30 per cent of their workweek, verifying the accuracy of the data needed to make business decisions.
Another sign of poor data quality is when employees are constantly raising questions and concerns over the trustworthiness of their data, said Kristin McMahon, audit marketing manager for enterprise information management at business intelligence software developer Business Objects in Vancouver.
“There’s nothing worse than going to a meeting with your vice-president of sales where you’re showing you have $45 million for the quarter and he has a report showing $65 million for the quarter. Whose report is right?” McMahon said.
Examples of dirty data can range from bad addresses, non-standard firm names, multi-format person names, incorrect account value, missing data and duplicate records. For instance, the different format for entering dates – with dashes, slashes or periods – may result in confusing information when merging data from various data sources, explained the Business Objects executive.
In resolving data quality issues, organizations should be careful not to address all their problems in one shot, cautioned Accenture’s Bell. “If [companies] are clever about it, [they] will sit back and think a minute before diving in and saying, ‘Clean everything.’”
All corporate data is not created equal, Bell pointed out. Certain information can be so vital to the company’s survival and competitiveness that they always need to be accurate and clean, he said.
On the other hand, there are those that simply don’t matter if they are dirty and the cost of cleaning them may not be worth the return, added Bell.
Deciding which data is crucial largely depends on the industry a company is in and its business objectives. “It goes back to what is [the company’s] strategic intent. If it’s all about increasing profitability per customer (for example), then you need to have the customer records spot on,” Bell pointed out.
Once there is an understanding of what data needs to be cleaned, the company should break up the project into manageable pieces, advised the Accenture executive. It is also important to show tangible successes and returns on the initial implementation before moving on to the next cleanup phase, Bell added.
Business Objects’ McMahon also suggested appointing a “data steward,” who knows the data intimately, understands the problem, and can lead a small data quality project.
Bell and McMahon agreed that executive buy-in is essential to maintain support for the data quality initiative. Part of the reason is that data quality is a continuing effort.
“Without a long-term commitment to maintaining data quality and appropriate usage, a company faces covering the same ground again and again,” wrote Bell in an article entitled, Is Your Data Dirty?, which he co-authored with Frank Dravis, a vice-president at Business Objects.
Data quality is a vital component of Customer Portfolios’ business. Customer Portfolio is a marketing services firm based in Boston that uses Business Objects’ IQ8 suite for its data quality integration. It provides clients with a full view of their customers with analytics that help drive targeted marketing programs.
A big part of its business is the ability to merge and clean data coming from the customer’s different data silos, such as call centre applications, e-commerce portals and catalogue selling activities, explained Mark Janowicz, founding partner at Customer Portfolios.
“Our endgame is to deliver high-end analytics for our clients, but the fundamental underlying data needs to be correct,” said Janowicz.
For more on Canada’s IT job market don’t miss our IT Labour Market Special: What’s hot, what’s not we’ve put together especially for you.