The tightening of the global economy has forced organisations to scramble and reshape their operations to maximise existing processes in the face of slowing sales. Today’s CIOs are tasked with cutting costs, reducing spending on non-essential projects and, yet again, doing less with more. Legacy infrastructure upgrades have been slowed and the cheque book is closed on many web services computing projects, reports Dominic Micic, DataFlux Australia’s Principal Solutions Architect.
However, smart businesses also understand they have a strategic asset - their corporate data. Data can be used to drive the business forward, make better decisions and positively impact profitability. The chaotic dynamic of data management presents a real challenge. Data is now collected and saved from every conceivable source - internet applications, front-office and back-office systems, customers, partners and suppliers, social media - and this complexity requires companies to have a sophisticated, deliberate process for managing this vital information.
After all, data holds the key to sales, marketing, customer support, production and other initiatives. Companies that fail to build accurate data sets about their customers, products, materials, locations and assets will struggle to compete in today’s marketplace.
As with any corporate initiative, the best starting point is to set measurable business goals. After all, no one has ever managed, moved or standardised data for the pleasure of it. There must be a business reason to conduct any data management initiative. At the highest level, there are three primary reasons why organisations perform any business function: to stay out of trouble, to make money or to spend less money. Or in business terms: manage governance, risk and compliance (GRC), better business performance and for cost control.
The key for data management professionals is to tie data management efforts to each of these business objectives. Already, far-sighted corporations are appointing data stewards and data guardians who are tasked with ensuring the business gets the trusted information it needs to succeed.
However, there is another strategic goal that businesses and IT leaders can aim for - to create a single, accurate and unified view of its corporate data so the data will be accessible to the entire organisation. The technical phrase that acts as the umbrella for this approach is master data management, or MDM. By freeing application data from their silos of information and establishing a centralised repository hub - or a single source of trusted data - business users are able to access the right data and be productive in their decisions and analysis.
Many enterprises have built complex IT environments, resulting in the creation of disparate information systems across the organisation. By going down the MDM route, companies achieve a single version of the data, which is cleansed and validated using data quality software tools.
Whether the business is running a customer relationship management, enterprise resource planning system or even a call centre application, it should look to consolidate the single source of that data into a master repository that can be accessed quickly by the user.
The advantages of MDM can be seen in both the business and IT sides of an organisation. When considering freeing data from applications prior to an MDM initiative, there are some important things to consider:
- Data is forever. It has a lifespan beyond any application as demonstrated each time we migrate from one application to another. We need to treat the data with the respect it deserves. It is our key business asset.
- Data is owned by the business - not the application. Having each application owning its own copy of key information like customer, product or employee will ensure that the sought-after ‘single version of the truth’ will remain elusive.
- The quality of the data is never guaranteed without diligence. As the business is responsible for creating and modifying the data, it needs to take responsibility for its accuracy. IT can certainly assist with tools but fundamentally business needs to be responsible for data governance.
- MDM is fundamentally about creating an environment where the enterprise master data is offered as a service to applications. Applications act as clients to the MDM system and are subjected to centrally managed business rules that protect and secure the data. Sometimes the MDM system design will delegate management of certain entities to actual applications. Other times the MDM system may manage the information itself. These technical details should not obscure the key idea that the master data is decoupled from the application and managed beyond them.
So when do you need to implement an MDM system? If the business is frustrated by inaccurate, inconsistent data where each system has a different version of the truth then you have the need for MDM. Bear in mind though, businesses will not come asking for MDM. They will not approach IT to talk about the underlying technical architecture. They just want to know why they cannot get access to their data and why it appears wrong.
The CIO or IT director can take the lead in the boardroom and push for investment in a data-driven system that removes any questions about whether the technology division is supporting the rest of the executive team.
Placing data at the centre of the equation allows management to tackle the key data governance and compliance issues that all company directors approach with some trepidation. Losing, leaking or sending out the wrong data can cost businesses millions of dollars. Compliance is costing business a significant amount of capital spend today.
By enabling a phased approach, organisations can move from project-level data quality and data integration to full-scale MDM deployments, all using the same data management platform.
The business should identify the key reference data and work with IT on building a shared data repository. This repository houses cleansed, accurate information gleaned from the key enterprise applications and delivers an enterprise-wide consistent view that is fully governed, catalogued and accessible.
The most powerful outcome is that the business takes control of its destiny, as it is now equipped with a data-driven approach to IT.
There is another critical step before the creation of the master record. This relates to the value of the data itself, or the data quality. Some industry estimates show that data quality problems cost businesses in the US more than US$600 billion a year. The main reason for inconsistent or unusable data is human error. We all know how bad information, or disparate information, can lead to frustration and poor productivity.
Human error can be the result of poorly structured processes. For many companies, data quality issues arise when they try to integrate data from other sources. If one business unit or division uses one set of standards for data quality, while another unit employs different practice, then the integration of data will lead to a master data set filled with inconsistent information.
The six steps to data improvement can be defined as define, discover, design, execute, evaluate and control. This unified view of any type of data is achieved by implementing a data management platform, which provides the single view of the customer, your product range, your supplier information or your procurement data.
Once all of the key business information is in one logical place - the central repository hub - data governance rules can be properly applied. Even before this maturity level is reached, and the legal team is interested, there are many other executives within the organisation who would see benefits from the central single source of truth.
If data is the key asset in the corporation then it needs to be treated in a centralised way. It can’t be locked away in multiple places with inconsistencies. The master data management approach - where data in the application can point to the central hub, but not act as the single source - is the future of enterprise data management.
*Dominic Micic is the Principal Solutions Architect for DataFlux in Australia. He has worked in the technology industry for 14 years in a number of senior technical roles, focusing on data quality and master data management initiatives. Micic has led systems integration projects, data quality audits and master data management projects, and has a particular expertise in ensuring that data projects support key business unit requirements.