Gaining Master Data Insight

A vital but often neglected area in master data management (MDM) at present is that of data governance, moving beyond the storage and integration of master data to be able to help automate the business processes associated with that data. For example, if a change has to be made to an international product hierarchy, several business people may be involved in the process of developing a draft hierarchy, reviewing and approving this, prior to making it live in a master data repository. The management of the life cycle of master data in this way has often been neglected by technology vendors, yet this is key to ensuring that a master data project becomes a sustainable embedded process rather than a one-off data clean-up exercise. With MDM you need to treat the disease, not just the symptoms.

An example of how vendors are improving their support for governance is the latest announcement by Siperian of its Insight Manager. This module builds on the business model rules which can already be defined in the Siperian business model, allowing the monitoring of data quality and the effectiveness of the data stewardship process in addition to gaining insight through ad-hoc analysis using business data dashboards. By hooking Siperian up to MicroStrategy, the Insight Manager can take advantage of the event management abilities of MicroStrategy to generate alerts based on certain business rules, and also to easily build a data quality scorecard.

An example would be in the world of insurance. An individual may have multiple insurance policies (home, travel, auto, etc) yet these may be dealt with in different operational systems. Insurance companies may implement a master data technology to better manage a single view of their customers. It would now be possible for an insurance company with the Siperian MDM Hub to define a business rule to prompt an agent to attempt to up-sell an auto insurance policy to a family whose child has reached the legal driving age. The rule would be stored within the Siperian metadata, but the alert and event management handled by MicroStrategy.

Another use would be to generate a MicroStrategy scorecard which shows statistics of data quality at its various process stages (cleanse/merge/match) and enable a company to quickly see which processes were becoming bottlenecks e.g. if data stewards took abnormally long in the approval process compared to the norm (Siperian’s relationship with MicroStrategy here is not exclusive; a similar link has been demonstrated with Cognos, and others may be expected to follow).

When building up the model of each particular company’s business processes, the flexibility of the MDM technology’s business model and the ease of updating it are very important. While certain basic processes may be susceptible to standardisation of a model, this is certainly not the case with data governance processes, which vary substantially from company to company. Hence technologies which have a flexible, easily configurable model will find it easier to adapt to specific data governance rules than products that have a fixed data model.

As master data implementations become more enterprise-wide, they will need to expand from single domains (such as “customer” and product”) into true multi-domain platforms. Moreover they will need to address not just system to system merging and matching but need to support the complex, and highly tailored, processes that companies use to automate the full life cycle of master data.