In-Depth

Survey: Executives Still Don't Take Data Quality Seriously Enough

In a new survey from PricewaterhouseCoopers, the company found that data quality is still a serious issue to which executives still aren't paying enough attention.

• By Stephen Swoyer
• 11/17/2004

If effective decision-making is empowered by access to valuable, reliable information, then many executives aren’t making terribly effective decisions.

That’s the upshot of a recent survey from PricewaterhouseCoopers (PwC), which found that almost half of respondents don’t believe senior management takes data quality seriously enough.

PwC asked 452 companies in the U.S., U.K., and Australia about their data-quality strategies. Surprisingly, it found that board-level involvement in data-management issues has actually decreased over the last three years. For example, two-thirds (67 percent) of organizations said they had a company-wide data-quality strategy in place, but only 40 percent were certain that their data-quality strategies have board approval. If that’s not bad enough, 41 percent of respondents don’t believe senior management takes data quality seriously enough.

The simple upshot, says PwC partner George Marinos, is that organizations still aren’t aligning their data-quality strategies with their core business objectives. “You really want to make sure that you know what your objectives are and then build your data-quality process out of that. But when we compared the results from this year with an earlier survey, we found that executives are still paying insufficient attention to the quality of this data,” he comments. “The proportion of businesses with a formalized company-wide data-quality strategy only rose marginally, which is really surprising.”

It’s surprising, Marinos says, because data quality should be a front-and-center issue for many companies. For example, some organizations are now tackling performance- and quality-improvement initiatives such as business performance management (BPM) and business activity monitoring (BAM). Both strategies purport to empower executives to make better decisions. Both require access to reliable, time-critical information.

More to the point, Marinos says, even if companies aren’t investing in BPM, BAM, or any other quality improvement acronym, they’re almost certainly dealing with new regulatory compliance requirements—which also involve significant data-quality investments.

“About 63 percent of respondents indicated that regulatory compliance is a major business driver for them, which raises the question: if it’s such a major business driver, why hasn’t this resulted in board-approved data-quality strategies?” Marinos asks.

He’s not altogether pessimistic, however: “My sense is this is really ready to take off. Executives haven’t really made a strong linkage between the control objectives they’re looking at now and a formal data-quality approach. They are looking at control objectives around systems [and] around processes, but they’re not getting down the next level into data, into data-quality issues for operational data. My sense is that the controls they’ve put in place are working, but they still have poor data. So the next move is to put in the additional controls in around data quality.”

There’s some (albeit paradoxical) evidence this is already happening, says Marinos: Three years ago, 37 percent of respondents said they were “very confident” in the quality of their data; in 2004, just 34 percent said the same. The difference, Marinos suggests, reflects a gradually heightening awareness of the problem. “The types of discussions I’m having now aren’t ‘What is [data quality]?’ but ‘I know I’ve got a problem and how do I fix it?’” he explains.

Marinos believes that data-quality issues will capture executive mindshare only gradually. He describes this as an “evolutionary” approach to the problem and contrasts it with the almost overnight revolution in terms of how executives view security.

“Ten to 15 years ago, security wasn’t making front-page news, but in the last seven years it’s been in the forefront of risk-management concerns that companies have. I feel data-quality concerns will reach that point,” he says.

The security revolution was prompted in large part by sensational exploits. Marinos doesn’t think that the same thing will happen with data quality, even though he believes that data-quality issues are at the root of many business problems today. “We’re seeing these [sensational data-quality failures] already, but we’re not calling them data-quality issues. I think you can pick up any newspaper with any sensational headline, and at the root of it in many cases, it’s a data-quality issue,” he concludes. “It’s not having the proper governance around your data. And some of the cases we’ve been seeing in the news could in fact have been data-quality issues.”

Elsewhere, over half of survey respondents said that they’re likely to exchange more data with third parties over the next three years. At the same time, however, only 18 percent of respondents whose organizations share data with third parties say that they’re “very confident” in the quality of that data. What’s more, only 24 percent of these companies are actually measuring the quality of that data. Almost one-quarter (24 percent) actually express little to no confidence in the data they get from third parties.