Ultimately the architecture chosen will not define success or failure. In fact. like any large project involving multiple stakeholders, one key issue is the ability to manage consistency between different teams, and prioritise and manage the politics that are inherent in enterprise projects. Like any large project, both IT and business commitment and involvement are critical to success.
In my opinion, these are critical objectives of any competent data governance program. Most successful MDM projects have had a strong, pragmatic data governance program running as a core stream. A pragmatic data governance program needs to be run bottom up, needs to be focussed, and needs to enable decision making.
Why bottom up? Master data will be reused across the enterprise. In order to properly manage consistency between core teams the data governance organisation (DGO) needs to understand how different areas within the business will use the data. Very often, data is used for very different purposes by different areas. If one team makes assumptions to adapt data elements to suit their own needs this can create havoc elsewhere. Only operational staff that are actively using data are able to understand the impact of changes in their areas. The DGO needs this knowledge in order to build data quality rules, assess the impact of changes to data (cleansing or enriching), build metadata (a business glossary of definitions) about data, etc.
Similarly, the DGO needs to be focussed. It is important to address the needs of all intended users of the MDM system – the DGO needs to fulfil an enterprise role rather than a departmental one – so focus does not mean addressing only one need. Our methodology enables the identification and prioritisation of core attributes across multiple areas and it makes sense to focus on these core attributes – rather than on all the data. Initially, the DGO should focus on data that is necessary to meet the business needs of the MDM project – although this role may expand with time.
Data Governance needs to enable decision making. This is where top down support becomes important. The MDM/DGO Steering Committee should be presented with key finding for decision making purposes. Key decisions could include approvals of budgets, agreeing priorities or approving key changes to existing systems.
Finally, and hopefully obviously, the DGO needs to be business driven and involve both business and IT role players. If this is achieved it will be the key forum to identify business requirements, manage conflicting priorities and manage risk. It will also be the key forum to drive data quality – the other key success factor for MDM. I will discuss that further in my next post.