As we are coming to understand, BI usage patterns have begun to segment into partitions defined by the types of users and the scope of the solution deployment, as well as the purpose of the analytical activity. This is finally the toehold we’ve been looking for in understanding how to create a BI landscape where strategies are increasingly drawn toward the capability of vendors. And these vendors offer more and more complete SaaS/PaaS solution strategies.
For companies interested in offloading the risk and technical complexity of operating BI solutions, we have seen how easily some vendors attract customers. They do this by offering analytics solutions in exchange for higher OPEX costs. The proposition is deceptively simple: you give us your data and your money, we host it and serve up analytics to you available anywhere: desktop, web, smartphone, tablet, you name it.
This has driven some companies understandably to put some components of their BI landscape in the cloud. And for many organisations, there is a sweet spot of analytics service delivery that makes sense to offer in this manner. But what is not quite obvious is which aspects of the BI landscape are truly suited to deliver on all of the organisation’s various needs: TCO mitigation, risk avoidance, and service satisfaction. Some companies are taking BI and analytics functions formerly delivered through SaaS partnerships and moving them back into the data center, while others are finally taking components that have been ‘in the house’ all this time and migrating them to the cloud.
So while we note that analytics solutions are migrating in both directions in the industry, what is not at all obvious is what criteria generally guide the most effective decisions. This is where our user/usage partition comes into play.
Segment 1: Focused Solution Drivers
This is the group of sophisticated, data-savvy users who need powerful analytics tools to answer questions the business doesnít even know about yet. They need data mining and discovery, and they need the tools they use to be easily adapted to their needs (extensible, quickly deployed, and adaptable). Note thatís not necessarily simple or easy-to-use, because these users are tolerant of higher learning curves and BI/analytics is a big potion of their daily work. This segment of a company’s BI landscape needs to stay in-house, because the volumes of data involved are often very high, and the analytics produced (with the transformative business decisions they enable) are among the most valuable informational commodities produced from the company’s entire data wealth.
Segment 2: Department Dashboards
These are users with modest familiarity with their data and with the tools to process it. They are often operating on OLAP/ROLAP data sets and tools and preparing data for consumption by larger constituencies (so-called self-service BI falls into this category). They are answering hard business questions, but not necessarily opening new gaps in the organisation’s capabilities. They often process department-sized data sets, which makes this segment a good candidate to put in the cloud, especially when the user communities consuming this data need to push analytical capabilities to a larger or more mobile audience. Yet the analytics themselves are not so sensitive and transformative that they cannot be risked to deploy outside the data centre.
Segment 3: Enterprise Reporting
For casual reporting needs, presentation-quality reporting to go in front of external customers, and parameterised or canned reports that can be distributed widely to a broad workforce, enterprise reporting forms another segment of solutions and tools. These are the best fit to consider moving to SaaS vendor offerings, as they seldom contain transformative analytics or proprietary decision-making capability. With the security of the data itself decreasing as a factor against SaaS solutions, the organisation’s most routine reporting needs can often be moved outside the house. Typical results are a TCO savings and service delivery improvements without introducing unacceptable risks.
All this planning and effective BI strategy is just the first step in creating an effective business intelligence landscape. In upcoming blog posts, we will delve deeper into what it takes to manage such a landscape, when pieces of the organisation’s portfolio are in the data centre and portions are in the cloud.
Mr. Briggs has been active in the fields of Data Warehousing and Business Intelligence for the entirety of his 17-year career. He was responsible for the early adoption and promulgation of BI at one of the world’s largest consumer product companies and developed their initial BI competency centre. He has consulted with numerous other companies about effective BI practices. He holds a Master of Science degree in Computer Science from the University of Illinois at Urbana-Champaign and a Bachelor of Arts degree from Williams College (Mass).
View Linkedin Profile->
Other Articles by Douglas->