The tolerance for bad data drops as the importance of the data rises.
Posted Jun 14, 2004
Fully 25 percent of all critical data at large enterprises is flawed, however the reliance on this data for customer-facing applications is on the upswing, according to recent findings from two separate research firms. The resulting disconnect, analysts say, may lead to complications in companies' CRM efforts.
The conclusion regarding data quality is from the report "Using Business Intelligence to Gain a Competitive Edge," issued by Gartner Research. The author of the report, Ted Friedman, a principal analyst at Gartner Research, says that the 25 percent figure "is kind of a scary statistic," but that it doesn't come as a surprise. "Most organizations I run into are really, really poor on executing on this data-quality topic," he says.
CRM efforts in particular are essentially doomed by bad data, Friedman says. "People continue to fail in their CRM effort because they only care about getting the data all in one place, or making it accessible...they don't put any thought into the quality of that data," he says. But, he says, some companies are beginning to see the light: "They're saying, 'We finally realize that the reason we are failing is that our customer data is just [junk].'"
Another new study, "Data Quality: A Shift In Perceptions," conducted by NOP World Business & Technology and commissioned by data quality vendor QAS, found that companies that use CRM applications are more likely to have a data accuracy policy in place than those that don't. Fifty-six percent of companies that use CRM applications have data accuracy policies, but only 32 percent of companies without CRM have such policies. Overall, 80 percent of organizations believe that inaccurate data costs their business money.
The awareness of data quality as a critical concern varies from industry to industry, and even within divisions of a given organization. The more information-intensive the business model, the more likely it is that data accuracy has already been identified as an area of concern. Friedman says "companies [that] treat data as a competitive asset as opposed to an IT evil...do better about keeping data quality a priority." The financial services, insurance, and retail verticals, according to Friedman, are relatively progressive in this regard, with healthcare coming along. He places manufacturing and government among the laggards.
Measuring the accuracy of internal data is a challenge. According to the NOP survey, job function alters the perception as well. Ninety percent of data management personnel and 81 percent of IT personnel expressed satisfaction with their companies' data accuracy, while members of the marketing and sales departments were only 60 percent and 55 percent satisfied, respectively.
In other words, the tolerance for bad data drops as the importance of the data rises. "Your internal people can't do their jobs if the data's wrong," Friedman says. But the problems, he says, begin higher up the food chain. "Data quality is a business issue, not an IT issue," Friedman says. "If you live in a world where your businesspeople don't get that, you've got to do some bootstrapping to get them there."
The NOP survey, which contacted 250 organizations in the United Kingdom and compared the results against a similar survey conducted in 2002, found that only 48 percent of companies--less than half--undertake database cleansing at the point of input. A quarter of respondents said they cleanse their databases no more than annually, and more than half of those companies said they never engage in any routine data cleansing at all.
Friedman says that the mantra he keeps hearing is, "It's all about the data, stupid," and he says that awareness of the issue, at least, is on the rise.
Two years ago, lack of senior management buy-in was considered an obstacle to data-quality improvement by 22 percent of respondents to the first NOP World survey. In the most recent edition, only 8 percent said it was a problem.
But awareness is just the beginning, according to Jonathan Hulford-Funnell, COO of QAS. The next step, he says, is to gauge the level of attention that managers can spare to address the problem. "CRM takes up heaps of time from IT to begin with," he says, "so there tends to be very little capacity to do any more."
When it comes to serving customers, however, progress has to be made. "Quality of customer data is critical in CRM," Friedman says. "CRM is outward-facing, and your customer data is on display for the world to see. We'll see improvement there."
How to...ensure CRM data quality
What Will Wake You Up at 3:00 A.M.?
Data Quality Market to Grow at 30 Percent
Sponsored By: Marketo and Real Magnet
Sponsored By: Informatica