“Knowledge is of two kinds. We know a subject ourselves, or we know where we can find information upon it,” according to Dr. Samuel Johnson.
Metadata indeed provides businesses with the type of knowledge to find the required information and to better understand enterprise data assets, manage them more efficiently, communicate and share them across different user groups, extend and rationalise their architecture and control their access and security.
Metadata management is the cornerstone of any flexible and scalable enterprise information architecture and a key data integration enabler. (Try implementing SOA or Enterprise Search without proper metadata management!) However, it is often only considered a good architectural principle.
In my experience, I have seen that investments in metadata management are often postponed to better times. During these challenging times, enterprises struggle to receive the required business sponsorship and to present a credible business case around value drivers such as:
- Time saved in communication, investigation and reconciliation during data management related activities
- Increased confidence in data and reports, switching company focus to more core business activities
- Enhanced efficiency in the full project lifecycle and assessing impact of post-go live changes
In some organisations, metadata management is considered a pure IT tool implementation project, but this should not be the case. Deployment of metadata management must include change management, fostering a culture of data governance and data sharing to maximise metadata quality and the advantages of metadata reuse across the enterprise.
Accenture has found that high performing businesses are taking advantage of the benefits of metadata management investment to help them stay competitive during these challenging economic times. These companies consider a value based phased approach towards metadata integration and they start from tactical suboptimal solutions - integrating only a defined type of metadata and supporting targeted user groups and process needs, for example.
Most importantly, enterprises looking to start a metadata project should keep the following in mind when dealing with metadata:
- Metadata does not necessarily have anything to do with IT
- Metadata is potentially infinite and new categories are developing everyday
- Prioritise value-add metadata, implement gradually and stop when cost greater than incremental benefit
- Metadata faces the same issues as any other type of data (e.g. quality control, data governance and configuration)
- Storing and integrating metadata is only the first challenging step. The real value comes with industrialised metadata collection and wide usage in project life-cycle and business processes
- Addressing metadata management together with master data management and data quality can appear more challenging, but it can provide several synergies
- Think about it from the start of the information architecture design
- Evaluate accurately buy vs. build solutions
- A single repository design cannot accommodate all metadata
- Effective metadata management relies more on a culture of knowledge sharing across the enterprise than on a complex technical solution.