Master Data Management (MDM) is the discipline that looks after the management of lists of large and complex concepts that are shared across an organisation. MDM helps ensure that these lists are all in sync, thereby providing a single source of the truth for a given concept – on an ongoing basis.
Bingo! So MDM is precisely what's needed to deliver a single view. And there are numerous tools on the market to aid MDM, which typically include:
- The ability to integrate with source data systems
- Data Quality components
- A repository to hold the single view
- Mechanisms to deliver the single view to consuming systems, such as Business Intelligence tools
- Synchronisation facilities
- Controls and reporting
But don't just rush off to an MDM tool vendor. All too many people have failed, believing that an MDM tool is all they need to compile and maintain their single view.
In truth, it's just one part of the solution, and must be supported by effective Data Governance and rolled out using a robust Master Data framework. Without the latter two, the MDM tool will fail to deliver the promised benefits.
Most importantly, to be successful, a single view solution needs all those who have an interest in a given list to work together, in partnership with any technology.
This cross-departmental collaboration will likely be a change to the way people are used to working, and probably to the way organisational processes are set up.
An established, tried-and-trusted Master Data Framework will guide organisations through the required changes to people and processes in a controlled way.
Data Governance is then the ongoing practice that looks after the relevant data and information, by coordinating collaboration between stakeholders to resolve the issues that will invariably crop up as datasets and domains expand over time.
Master Data framework
The Master Data framework addresses the elements an organisation needs to put in place if it is to implement effective Master Data Management (and, by extension, its single view). It needs to cover people, processes and information, as well as technology.
Central to the framework will be a Data Asset Plan, which looks at the overall landscape of the organisation, and takes into account the wishes of key stakeholders.
This is absolutely essential to the success of the single view initiative, because the project will require changes that reach right across an organisation, so it's important to get buy-in from the very beginning – and that includes the top-level management. Everyone needs to be pulling in the same direction if the single view is to be a success.
The Data Asset Plan also looks at how the transition from the as-is to the to-be state will be managed, and takes into account ongoing business-as-usual operations and other complementary and competing change programmes that are taking place – the organisation needs to continue to operate during the Master Data rollout, after all.
The Data Asset Plan will sit alongside frameworks to design, build, deliver and operate the Master Data Solution that underpins the single view. And once all of this is in place, you'll need effective Data Governance to ensure the single view remains accurate.
Data Governance
As touched on above, Data Governance is about ensuring data and information are properly looked after across an organisation.
This is important because the lists of data that make up the single view will be produced and consumed by more than one department, each for a different purpose.
Stakeholders need to get together to agree on issues around synchronisation (what data needs to be synchronised?
How often does this need to happen?), terminology ('car' or 'vehicle' may be used interchangeably in one list, but refer to different things in another), data formats (is the date displayed as dd/mm/yy or dd/mm/yyyy?) and the like.
These discussions will lead to a number of things, all of which make up holistic Data Governance, notably:
- A defined organisational structure
- A description and understanding of roles and responsibilities
- A set of principles, policies, processes, procedures and standards
- Appropriate reporting and assurance
To support this, there needs to be total top-level buy-in to ensure that once a set of principles are agreed, these are stuck to across the organisation. Any changes to the above need to be controlled and agreed centrally, to avoid the situation where things get out of sync again.
Earlier in this guide, we touched on the example of where a customer opts out of communications from an organisation, but the change is only made in one department's list, meaning that another department continues to send unwanted (and illegal) mail.
Effective Data Governance would help overcome this, because the staff member dealing with the opt-out request would know (through their understanding of their roles and responsibilities) which procedures and processes to follow to ensure the opt-out is reflected in all lists and, if necessary, how to escalate the issue (through the reporting and assurance structures).
Key considerations
So Master Data Management is the best way to deliver a single view, and an appropriate MDM tool can help you make this a reality.
However, remember that the tool is just one of three components required to successfully deliver a single view: without the right Master Data framework and robust Data Governance, your single view implementation will fail.
IPL has produced a best practice guide on creating a single view, which is a free download.
- Richard Shreeve is IPL's Consultancy Director. He has worked across the public and private sectors to help clients exploit their data.
