Intelligent Enterprise

Last year, in our Business Intelligence Pipeline Voting Booth, we asked the readers to rate the overall quality of their data for analytical purposes: excellent, good, fair or poor. Respondents leaned toward the bad news end of the spectrum. This year asked the same question again, and it looks like things have, in fact, gotten worse.

In the first poll, from October of 2004, a surprising 19 percent of readers rated their data quality as "excellent." This year that number plunged to four percent.

In the middle range -- "good" and "fair" data ratings -- things didn't change all that much. The number of respondents who rated their data quality as "good" nearly doubled from last year, to 23 percent. The portion who consider their data of fair quality slipped a bit, to 43 percent from slightly more than half of respondents last year.

The worst news came in the portion of Business Intelligence Pipeline readers who rate their data as poor, which we defined as "often worthless for BI purposes." Last year only 19 percent chose this description for their organization's enterprise data. In our most recent poll, that number swelled to 30 percent.

This marks the first time we've replicated a poll to gauge how responses have changed from year to year. We're going to be doing more of it, so keep an eye out for new Voting Booth pages. You can find our new poll, on which improvements you'd like most to see from vendors' analytical products, here or on our home page.

Have any thoughts on why data quality is slipping? Or is yours good enough as is? Share with us! Add a little insight to the comment box on our blog page.