In an earlier post (The Business Intelligence Chronicles Part 16: Is Michael Jackson still alive? (aka Doubting Thomas/BI Skeptic) I had talked about the need to allay the fears of data skeptics - the guys who question the veracity and accuracy of information presented to them via BI tools.
Picking up that thread again I wanted to talk about the close relationship between BI and Data Governance and the two distinct forms the relationship can take.
Approach 1: "A world through rose tinted glasses"
Filter out bad data from the presentation layer and place it in an "exception report" or club it into "unknown" or "orphans" or "all others" category in the presentation layer
Approach 2" : "show warts and all"
In the presentation layer show data as is - errors and all. Visibility to errors will drive action to get the source data rectified. As a fellow blogger commented "nothing quite gets data fixed as quickly as the bad entries showing up on something the CEO sees".
The decision on what approach to take leads to an interesting dilemma for the BI manager - "Will my BI system not appear less trustworthy if it shows all the warts?" or " Will I be able to get the focus and attention required to rectify data issues if Senior Management are not exposed to them"
This leads to an excellent opportunity for Leadership to couple the BI and Data Governance initiatives with BI becoming the trigger point of a feedback loop into the Data Governance mechanism which then initiates action to identify and rectify the issue (user error, lack of validation in a data input system, problem with one or more interface etc.). And voila! if all works well the next run of the report will show correct information.
Obviously all this holds true if senior management values the BI they receive and realize the importance of its accuracy. If that holds true than the quality of data can improve drastically within a few cycles. Nothing inspires action more in a corporate environment than a missive from the CXO.
Picking up that thread again I wanted to talk about the close relationship between BI and Data Governance and the two distinct forms the relationship can take.
Approach 1: "A world through rose tinted glasses"
Filter out bad data from the presentation layer and place it in an "exception report" or club it into "unknown" or "orphans" or "all others" category in the presentation layer
Approach 2" : "show warts and all"
In the presentation layer show data as is - errors and all. Visibility to errors will drive action to get the source data rectified. As a fellow blogger commented "nothing quite gets data fixed as quickly as the bad entries showing up on something the CEO sees".
The decision on what approach to take leads to an interesting dilemma for the BI manager - "Will my BI system not appear less trustworthy if it shows all the warts?" or " Will I be able to get the focus and attention required to rectify data issues if Senior Management are not exposed to them"
This leads to an excellent opportunity for Leadership to couple the BI and Data Governance initiatives with BI becoming the trigger point of a feedback loop into the Data Governance mechanism which then initiates action to identify and rectify the issue (user error, lack of validation in a data input system, problem with one or more interface etc.). And voila! if all works well the next run of the report will show correct information.
Obviously all this holds true if senior management values the BI they receive and realize the importance of its accuracy. If that holds true than the quality of data can improve drastically within a few cycles. Nothing inspires action more in a corporate environment than a missive from the CXO.
Mr. Lely, I desire you would use all your skill to paint my picture truly like me, and not flatter me at all; but remark all these roughnesses, pimples, warts, and everything as you see me, otherwise I will never pay a farthing for it.
-Oliver Cromwell
Posts
No comments:
Post a Comment