Excellent BI Read on Data Quality
I stumbled across an interesting blog post this morning by Vishwas Agashe entitled “Don’t shoot the messenger, attack the message! How to avoid data quality issues from derailing BI initiatives”. Read it here. Vishwas makes a very compelling case for talking about data quality BEFORE you begin a business intelligence initiative. Good stuff.
One of the challenges I have seen in pushing forward BI (which is very much an overused term these days) is clearing some misconceptions around the end result. Specifically with data quality, there are two major (perception) hurdles:
1. “We can’t do a BI project because our data is a mess.”
2. “We will do this BI project because it will fix the data that is currently a mess.”
As Vishwas points out, data quality is a people (not systems) problem. It will always exist. Get over it. The goal of a corporate reporting initiative, therefore, should include steps for identifying, quantifying and monitoring data quality (from the first day until forever). Set expectations with the user community, support it with hard data (“here’s how good our data was six months ago; here’s how we look today”) and make it someone’s responsibility to care about.
Sure, data quality is a problem… but if treated properly it can also be part of the solution.
Check out Vishwas’s blog here.