Information management is described as ‘the conscious process by which information is gathered
and used to assist decision-making at all levels of the organisation’.
It sounds easy but how do you get to this Utopian place? Starting at the basic level every organisation has data: individual building blocks of information which convey little meaning (e.g. strings of textual or numeric characters).Most organisations have too much data floating around their organisations these days, the important thing is to create information from it.
Information is data that has been organised in some useful way in order so that meaning can be extracted from it. Once it has meaning and context it’s much easier to start the steps to providing “business intelligence”.
The key here is to have standards for data definitions – recently called Master Data Management and now lately Data Relationship Management. What is the definition of MDM/DRM: Master Data is the reference data, which we see in General Ledger hierarchies and reporting structures.
At a simple level having MDM/DRM in place means having a single version of hierarchies and definitions, so that all applications and users can reference the same definitions, to allow consistent reporting and analysis. Otherwise with separate applications running inconsistent hierarchies it means users are not comparing like with like.
This thing is too big! Yes, companies can have multiple systems and multiple hierarchies and multiple definitions, which at first glance can seem to big to manage let alone trying to get the systems to speak to each other. You need to have a grand vision to include all systems and understand dipendancies however it’s best if you limit the scope of the initial deployment.
Getting buy in from the business/IT: You need to identify the stakeholders and help them understand the benefits, such as reduced downtime and less errors/fixes. It is very difficult to centralise this
process as most users don’t want to lose control of their own MDM. One way to influence this is to focus on the end state and reduced maintenance and quicker updates that this model can provide.
Creating Data Governance Policies and Processes: Once you have buy-in the next tricky part is creating a process that aligns with everyones calendar and timetables. It’s important to note, that you don’t need to immediately build everything to be “Enterprise Ready”, as this can create extra layers and distrust with the system, make sure it’s fit for the immediate purpose and has room to scale.
The key to progressing at speed is to have single data definitions and a master repository of hierarchies. Not only does this allow production systems to be updated but it also improves the
development systems – allowing alternative hierarchies to be tested and easily rolled back.
Oracle has updated and enhanced their Hyperion MDM tool, which is now called Data Relationship Management.
We have long talked about MDM and the benefits, although managing to get all users: Finance, IT and the business to agree to relinquish control has been harder.
We have at this stage limited our MDM activities to Management Reporting systems, which has been easier to control through the use of standard hierarchies, naming conventions and data definitions. Currently on Hyperion version 9.3.1, we have utilised skeleton cubes to manage the hierarchies to simplify the maintenance and create a single repository for management reporting changes.
Looking ahead, Version 11.1, can manage the approval and changes in hierarchies much better, through workflow with the DRM tool and also integration and synchronisation with multiple systems and data warehouses. The other important aspect here is this would also simplify development and testing of alternative hierarchies, being able to easily test and roll out to other systems.
Anything that can simplify back-end processes and add business value is always welcome, I’m looking forward to the upgrade 😉