Periodic Rationalization & Prioritization of Information has multiple benefits

Many of these reports and information requirements, escape the centralized meta-data management process (which in many cases do not exist) as users create their queries through front-end tools, which are outside the meta data management frame-work.

This leads to the information/reports repositories, which are:

• Duplicate or super-set/sub-sets of other members.
• Non-standard nomenclature- with same reports and same data elements carrying different labels (for example Revenue data carrying the name of the labels- 'Gross Revenue','Booked Revenue'..)
• Reverse of the above- Different reports and data elements, carrying identical labels (For example Gross revenue and Net Revenue, both labeled as Revenue)
• Reports, which are not required but still being published.
• Reports being generated with a higher frequency than needed.
• Different reports set by different departments
• Uncontrolled change management process.

This results in:

• Different functions contesting each-other's numbers in the board-room.
• Overload in the system, with system not being able to deliver to information load.
• No one owns a report, and cannot take the decision around it.
• The end-users getting confused on different reports showing different numbers (this may not be only due to ms-matched formulas, but the cut-off for firing these reports might be different)
• Too much information leads to 'no information', due to lack of trust on the data or ability to absorb the data.

This field tip is more tactical and a quick-fix, which will also promote a structural fix.

Every month or at the end of certain period, take stock of your entire information inventory, and rationalize/normalize it on the following tracks:

Ist -PASS

• Delete the reports, which have not been accessed at all, after sending a pre-notification.
• De-activate (though not delete), the reports, which have not been accesses in last 2-3 cycles of generation.
• By looking at the description of the reports and their meta-data, look at reports, which are super-set to the other reports or they can be made super-set by minor changes.
• Identify the reports, which are showing different values under the same label. Correct the labels or the formulae or the cut-off times of firing of these reports.
• Look at the reports, which are showing the same values under the different labels, and fix them on the same lines as above.
• Check the frequency, and given the purpose of the report, reduce the frequency, after pre-notification.

2nd Pass- on the remaining reports

• Have a quick check with the stated owners on the purpose and need.
• Where-ever there is not stated owner, deactivate, after pre-notification and wait.

It is not surprising, for a medium sized organization to have thousands of reports and thousands more of analytics views.

It's a myth that rationalization and normalization takes a lot of time. If done on the regular basis, you will find the people will develop good expertise to do the Ist and 2nd pass with in a matter of a single day, for a given function. Enterprise level normalization may take few days. Moreover, you will find that as you go through the exercise, there will be general awareness coming in the users on this subject, and exponential growth in information requirement will slow down. One also has to see that the rationalization & normalization of reports is not an exercise in perfection. As long as you are able to achieve 70-80% effectiveness, you can be considered successful. As we move towards higher %ages, there are diminishing returns and solutions become more expensive.