Perspectives & Strategy

Business Intelligence as a Service (BIaaS) is peeking many interests within the IT community. As I stated previously Gartner believes that 25% of all software will be delivered through an SaaS based service by 2011. At the same time Business Intelligence usage is expected to rise 130% by 2010. Many SaaS vendors are currently planning to incorporate BI in parallel to their applications, some are even partnering with existing BI vendors to make this happen.

According to Dhiran Gala of MAIA Intelligence ”BIaaS benefits address various business user and IT challenges” and can be “cost effective – ‘pay as you go’, scalable, quickly deployed”.

Does delivery of BI through the cloud represents an infrastructure revolution? Or are there still too many challenges to make this a viable solution?

For small and medium-sized enterprises the advantages are clear, implementation requires little or no capital expenditure, a great bonus. I accept that offering low-cost BI solutions does not mean inferior quality architecture or below par performance. According to Paradigm Analytics “what you will be giving up is the predictive capabilities and real time information access which is what true Business Intelligence is all about”. I believe that there are still concerns about data security that have yet to be fully answered by vendors.

I am and have always been concerned about the use of corporate data. The key challenge of BI is pulling together all relevant data into a single location, the Data Warehouse. When the corporate data exists in a variety of applications this is no less of a project challenge that it would be in a traditional environment. We still need to demolish data silos, even in the most analytically advanced corporations this is still a major challenge. Adding a new dimension of external storage for BI does not resolve this problem, on the contrary it may make things worse. Whilst on this subject it is important to note that BI solutions obtain their data from a variety of systems, not just the single ERP solution, so adding BI to the SaaS application is not sufficient to provide real intelligence, without the addition of other data elements that must come from other applications.

In my earlier article “The advantage of Good Business Intelligence” I stated that in all too many organisations data and information is not used to its best advantage; and ready access to information about an organisation’s business, products and customers is a key element in supporting decision making. The Data Warehouse is aimed at supporting decision making within the company and requires a broad spectrum of data sources in order to support the best analysis capacity. BI is about leveraging corporate-wide information, and I am yet to be convinced BI in the Cloud is able to support this.

Data quality has always been one of the key obstacles to good Business Intelligence. This challenge does not simply disappear by engaging on external service provider. Entry quality, process quality and other quality challenges still have to be overcome in order to provide a high quality solution.

One of the struggles the corporate IT department has faced has been the challenge of unifying the corporate IT architecture. Taking SaaS applications from multiple vendors can make for a more complex architecture. For anyone considering BI as a Services the same questions must be asked as they would be for a conventional BI solution in order to define a quality solution. I think there are still a great many challenges to overcome.