David Linthicum's weekly question in the forum was the one that inspired this post.
When all you have are transactions, there is no "Business Intelligence". But when you have abstracted Business Intelligence, many users want to drilldown into its constituent details.
If say Sales in region A is off of last year's achievements, then users want to drill down and see which zones, cities and towns contributed to this fall.
Data abstraction is important to draw the overall picture but that's what motivates someone to drilldown and see only those transactions that have shown to be an aberrant component in the overall picture.
These days there is also the element of alerts and time sensitivity of transactional as well as higher level, abstracted "Business Intelligence".
Operational BI highlights areas of worry as and when transactions are flowing.
You want to know the five customer deliveries that could be delayed far more than the promised dates, ahead of time. This is so that you can do something about these and make sure that they are not delayed.
The 95 other deliveries that were made on time may not invite accolades but the five that were missed will definitely lower customer satisfaction and resulting bad word of mouth!
So these days, it is not only higher level abstractions that are important but the ability to highlight aberrations on time so that they can be fixed in time!
Abstraction is a mental process we use when trying to discern what is essential or relevant to a problem; it does not require a belief in abstract entities. Tom G. Palmer
Leave a comment