Fragmented Data Ownership Worries Cos The Most

S. R. Balasubramanian, Former Executive Vice President, Godfrey Phillips India talks to Biztech2.com about the nuances of data management and advises CIOs on the best practices to deal with data.

What can CIOs do about ‘fragmented data ownership’ to ensure a successful information management programme?

This is the single largest problem that plagues many enterprises. In the absence of a well defined policy of data management, users tend to keep data separately thus creating several islands of information. As there is lack of coordination, it leads to data duplication, lack of security and safety.

CIOs should take a lead in defining a data management policy by classifying all information and data assets in order of its importance and then discuss with various business heads about the access policy and the criticality. All these policies need to be embedded into the system to enforce access rights, safety considerations, etc. Once automated, data governance gets so much easier.

More from Biztech

Future Group - Reliance Retail Deal approved by CCI RBI ban on cryptocurrencies takes effect; prohibition could force investors to tap the black market

In the last organisation that I was with, we implemented single sign-on and identity management. All access rights were defined centrally with proper approvals and all changes were documented and this was related not only to the applications and databases but also to unstructured data in the form of notes and documents, which lay in the document management systems.

Can you highlight some best practices an enterprise should follow to effectively deal with ‘bad’ data?

Best practices suggest laying down a clear policy with respect to data generation, storage and destruction. If data is centrally stored and controlled, it is easy to manage and ensure that approved practices are followed. Principles of information lifecycle management need to be applied and enforced.

How has the importance of enterprise data warehouses grown with increasing dependence on analytics?

Enterprise data warehouse has really become important with the increased emphasis on data analytics. I notice many projects on business intelligence being initiated by companies where data cleansing and availability of the right data assumes great importance. Since the data volume swells over a period of time, its management becomes critical.

What challenges do legacy systems pose to data management?

In many cases, data of legacy systems reside in separate disks and have to be managed separately. In other cases even if the data resides on a central storage in a SAN/NAS environment, it may reside in several separate databases and hence need to be monitored separately. For MIS purposes it becomes a challenge to draw out data from several sources. Maintenance of databases for the purposes of back-up and recovery also poses a problem.

How important it is to have policies for data governance in place? How effective are they in data management systems?

The business environment today demands a proper corporate governance model to be in place. One of the measures necessary is a comprehensive data governance model. Clearly laid down policies specify a set of rules for the creation, storage, access rights, management and destruction of data. These rules need to be followed though in some cases these rules look good on paper but are not seriously followed.

What key features can CIOs introduce in their data management strategies so that they are able to draw maximum benefit from BI tools ?

CIOs can assign disk volumes for storage space so that intensive BI work does not affect performance of the other systems. Being business critical systems, CIOs would do well to provide for proper fall back mechanisms so as to recover in case of any fault of the systems or storage.