The cloud provides a vision of an IT infrastructure that defines both the current model of IT services and the current state of data center evolution. It offers a number of advantages to BI applications, including capability to apply enormous resources and numerous processors on an ad-hoc basis, which may be used for solving prediction problems in analysis. It also reduces the need for in-house infrastructure. A movement toward cloud-based BI is part of a larger range of developments in BI, which must now cope with an expanding set of demands.
Market conditions and technology have created a need for changes in the way that BI is delivered, and cloud-based services occupy a central territory—both in describing an infrastructure for internal deployment and in defining an array of service offerings that play to these trends. Public cloud-based BI has now been made available by a number of vendors, including a significant offering from SAP in the form of its Business Objects platform. These SaaS BI solutions are of greatest importance to smaller businesses and departments of larger enterprises. Private cloud infrastructure is also being promoted as a means of improving BI operations within internal infrastructure.
Trends in the evolution of BI and Analytics favoring cloud-based approach include a move toward real time analysis, an increasing requirement for business process analysis to improve ongoing processes, a move toward sophisticated predictive analysis, and a need to extend analytic capabilities toward better usability in the front office . These trends put pressure on older Data Warehousing solutions based on batch processing. Although these older systems are still necessary to handle the diversity of data available in the enterprise, they must coexist with real time analytics. Together, the trends favor a cloud approach, where resources can be allocated on an as-needed basis and access is through a flexible, browser-based interface..
Cloud-based BI is not without its own issues, many of which remain as yet unresolved. Data storage, data transfer, and integration rank among the top infrastructure issues, while security remains a prominent concern for early adopters, particularly of public cloud or SaaS BI solutions. There are also the usual teething problems of new technology, such as uncertainty as to the stability of newer providers, and lack of established best practices.
Use of BI from the cloud imposes new architectural concepts that need to be considered to ensure that accurate analytic results are obtained. With the private cloud, these issues remain the same as with previous solutions, in that data marts are likely to continue to proliferate and users will increasingly require integration of data from beyond the firewall. With public and hybrid cloud solutions, the impact is likely to be considerably greater, and very careful planning will certainly be required.
I have recently published a major report on this topic that is now available from Cutter Consortium (http://cutter.com).
