BI trends in retrospect: first call to board the EIM train
By Julian Field, Director at Knowledge Integration Dynamics (KID).
[Johannesburg, 11 Dec 2012]
At the beginning of 2012, the research firms and pundits proposed numerous trends for the business intelligence (BI) world. Statistical analytics, BI in the cloud, mobile BI, in-memory analytics, agile BI, and big data were among the common technologies, methodologies and processes purportedly top of mind.
In essence all came true in South Africa, where many business executives and IT leaders did investigate some or even all of the top issues, but not all were deployed, says Julian Field, Director at Knowledge Integration Dynamics. Investigations into employment highlighted some major issues that have held the progress of BI in check and left some leaders querying suitability and just how to go about implementing the technologies. Analytics and big data were seen as tools to mitigate risk, help businesses develop new products and predict customer behaviour. Embedded BI was a related component seen to help cross-sell, support lean supply chains, co-ordinate activities across departments and manage distribution.
The BI software vendors responded as exemplified by SAP, which upgraded its BI products by improving BI for mobile platforms, adding social media features to support big data analytics, and more analytics tools for business users in SAP Visual Intelligence. In the new Business Objects release, mobile workers gained access to BI analytics on mobile devices.
Big data hasn't really had the impact many predicted, certainly not in South Africa. Two of the major concerns are cost and perceived benefit. However, the sooner businesses board the big data train – an inevitability in any event – the sooner they can begin to mitigate the cost of gleaning long-term knowledge. Organisations and consumers have no choice but to concede to using the often imposing resources that fall under the big data banner. As more discussions take place, it is becoming clear that there are very few remaining compelling reasons why they should not use the cloud and the big data it contains. Big data is here to stay and permeates, or can add value to, every business process. It is a fact highlighted by some maturity models that the value of information increases as people become more familiar with applying and using it. It is only natural that poor or raw patterns and requirements will prevail initially; their search patterns for information and discoveries become more economical and direct as users and consumers become more familiar with the uses of big data and learn what they want from it.
One of the challenges for BI in the cloud that still remains, however, is ensuring good quality data when it resides in off-premise, cloud-based systems. The hurdle has always been that organisations storing data in the cloud have been subject to service level agreements (SLAs) with their service providers that focus on access availability, speed of delivery, data recovery and security, but never on maintenance or watchful and responsible care in accordance with enterprise processes, procedures and practices. The result is that the data can never be fully trusted. Integration and quality concerns, which underscore the difference between trusted or not, require both cloud and on-premise data, information and content be subject to the same standards. They must endure the same rigours, the same exchange protocols, integration and quality processes, domain-respective business rules and so on, to ensure that all information is uniformly managed in a standardised manner. However, there will always exist a growing need to expand on existing on-premise data management processes and capacity to accommodate external cloud data, information and content.
Virtualisation is a trend that not many of the major research firms predicted upfront, yet has gained a fair amount of mindshare in South Africa. Yet, for all the intended benefits, the analytical, persisted database retains strong benefits in a warehouse architecture, with positive long-term cost benefits. The traditional data warehouse architecture, with costs calculated in a standard five-year period, remains far more cost-effective than a virtualised data environment with its additional processing requirements. Many people, however, overlook the ongoing operational costs because this tends to be less visible. The trend is to focus on the capital costs because this can be easily ascertained, but doing so will result in higher costs for the organisation.
All of these technologies ultimately support a move towards enterprise information management (EIM), which has yet to achieve widespread maturity in South Africa. However, in line with Gartner's EIM maturity model, local businesses continue to move in the right direction.
In plain English, EIM seeks to get the right information to the right person at the right time. It sounds deceptively simple, yet is an extremely complex undertaking because most businesses have grown in operational silos and those silos contain information about various aspects of the business, from products, services and customers to costs, employees, partners and more. The problem is that not all of it is contained in the fields of databases; much of it is in e-mail, photographs, word processing documents, spreadsheets, presentations, PDF documents, HTML pages, on paper, and every other manner of information storage.
That information can be extremely useful to gaining operational efficiencies, cross-selling, up-selling, customer management, rationalisation, growth, and almost every aspect of business – if it is correctly harnessed. EIM essentially combines structured and unstructured data and information, a conjoining of enterprise content management (ECM), or unstructured data, with business intelligence (BI), or structured data, and seeks to transcend traditional IT barriers to deliver an enterprise-wide view of data and information that supports decision-making.
It is a substantial field. It covers data warehousing, BI, data quality, data governance, data analytics, master data management, mobile data, cloud data, data security, extraction, transformation and loading – anything and everything to do with data, how it is collected, where it is stored, how it is changed, manipulated, used and destroyed.
The broadness of EIM requires a sustained effort over several years to achieve, requires senior executive commitment and support alongside the support of business unit and IT leaders. There are rare examples of EIM deployments in South Africa, most notably in healthcare, and so the benefits and competitive advantage still escape many.