Blog

Jan
05
CLOUD BI: THE CASE FOR THE HYBRID APPROACH

As more and more information services move to a cloud-based delivery model, it was only a matter of time before business intelligence applications followed suit. New companies like Birst and GoodData, and even major plaers like SAP with their recently-unveiled Cloudfor Analytics offering, are moving the management and delivery of BI solutions to cloud services. These offerings make enterprise-wide BI delivery much more manageable, as users can access analytical applications in any setting without high initial costs or complex deployment.


These solutions are especially appealing to small and mid-sized business who want to avoid the cost of on-premises servers and infrastructure. And while the seamless delivery of reports, visualizations and analytics across the enterprise is appealing, any IT department considering such an approach must ask where the data will reside with such a deployment. Will the data live in the cloud along with the hosted BI applications? Will it still be populated and managed on local servers? Or will a hybrid approach be used, and if so, what determines which data is local and which is hosted?


The cost savings that drives the business decisions about cloud BI may be at odds with IT and their reluctance to “outsource” the data to off-premises servers. Security concerns are always foremost in the mind of IT’s management (and often the business side as well). What the business needs to bear in mind is that the cost of on-premises security can often exceed that of a hosted cloud service (which often includes in in their base pricing). However, IT leadership may not be ready to relinquish control of their security environment to a third party, especially one whose focus is on data presentation rather than data protection.


This is why a hybrid approach is often favored by enterprises looking to dip their toes into the cloud BI pond. The hybrid architecture is especially attractive to cloud BI (as opposed to other cloud services) because it is easier to divide the platform between data access and presentation and data management and integration. IT can still retain local control over their databases (including the staff that’s involved in database administration, data architecture and ETL), while business users can access and interact with reports, analytics and visualizations anywhere and anytime via a cloud-delivered presentation platform. This approach is especially attractive to organizations who already have large legacy databases that are the result of major resource investment.


The new SAP Cloud for Analytics platform presents a great case for existing SAP and BOBJ clients to embark on a hybrid approach to cloud BI. Enterprises can “offload” much of their management of reporting and analytics to an integrated cloud environment, while still locally managing and maintaining SAP source systems, databases, and even BOBJ universes. Over time, as new data sources need to be integrated and analyzed, business and IT can decide how to incorporate those into a hybrid architecture, either continuing to manage them locally or as part of a cloud-hosted data platform. In the end, hybrid cloud BI creates balance between the business and IT in terms of value delivered to users and control over the source of that value.


In Part Two of the Cloud BI series we will examine the technical particulars of managing both local and cloud-hosted data and how that is incorporated into cloud-delivered reports and analytics.