While Software as a Service technology can help streamline rollouts of business intelligence tools, industry analysts and SaaS BI users caution that one of the biggest impediments to success is the misperception that the cloud-based delivery model eliminates the inherent complexity of BI projects.
“It’s important not to underestimate the work involved,” said William McKnight, president of McKnight Consulting Group LLC, a Plano, Texas-based consulting firm that focuses on BI, data warehousing and master data management. “Most of [deploying] business intelligence is integration, it’s modeling, it’s data quality. Wherever you happen to host any of the layers of BI, you still have to attend to those requirements.”
To ensure a smooth deployment, McKnight suggests taking a staged approach to developing a SaaS business intelligence system. “Put the less mission-critical elements in the cloud to begin with to understand the model, and escalate from there,” he said. Organizations should also consider the potential security advantages of a “hybrid cloud” approach that includes private clouds set up and managed internally, McKnight advised. Applications involving sensitive data -- financial or health care information, for example -- “don’t fit the public cloud model,” he noted.
Here are several more best-practices guidelines that McKnight and others recommended to help pave the way for successful implementations of SaaS BI tools:
Fully vet your vendor. Clearly, that’s a prerequisite whether you’re doing business intelligence in the cloud or using on-premises BI applications. But there are some special requirements to consider when you’re moving BI data to the cloud, according to analysts. To start with, they said, you should know exactly what kind of SaaS model you’re buying into and whether that could have any regulatory-compliance ramifications for your organization.
Questions to ask include whether a vendor is offering a multi-tenant approach in which your data would be commingled with information from other companies on servers that could be located in various facilities, or a hosting model in which the data would reside on dedicated servers in a specific location. Getting a full explanation of the vendor’s data security and integrity controls is critical as well, especially for organizations facing stringent compliance requirements on security and data privacy, said Fern Halper, a partner at consulting firm Hurwitz & Associates LLC in Needham, Mass.
It’s also important to drill down into how a SaaS business intelligence vendor handles service-level agreements as well as loss of data and business continuity issues in the event of server or data center failures, Halper said. In addition, be sure to quiz vendors on issues such as how quickly they can scale up systems, their systems management processes, the staffing resources they have available and how they handle technical support tickets, she advised.
Start small and demystify the technology. Just because you can get up and running in a relatively short amount of time with cloud BI tools doesn’t necessarily mean you should aim for a big-bang deployment. Whereas McKnight suggested starting with less-critical data, other analysts and users recommended that BI project managers work with business representatives to identify potential SaaS BI uses that could have a positive impact on an organization’s operations or bottom line and then play up early successes to help foster broader user buy-in.
Make a splash at the top. Andrew Bartels, IT director at PSA Insurance & Financial Services Inc. in Hunt Valley, Md., recommends putting SaaS BI data directly into the hands of top executives via easy-to-use dashboards, so that they’ll start relying on the data in their daily decision making. If a BI team can get senior management to use the cloud-based tools and speak the language of BI internally, “that would translate throughout the organization,” Bartels said. “People are going to key into that very quickly and say, ‘How can I get my hands on that data?’”
It also helps to make the data accessible in a way that business users are already familiar and comfortable with, according to Bartels. For example, PSA put dashboards with data from its SaaS BI system on the home page of its corporate intranet, where users already went on a regular basis to access information such as internal announcements and the company’s contact list. “Giving them analytics within that environment meant they didn’t have to learn a new place to go,” Bartels said.
Involve the business up front. Building reports and dashboards might not be technically difficult with SaaS BI tools, but knowing what you can do with the data that’s in an organization’s systems isn’t always readily apparent to BI or IT professionals, said Jamie Buck, manager of special projects at Foundation Source Philanthropic Services Inc., a Fairfield, Conn.-based company that provides management and advisory services to private foundations.
Foundation Source began using a SaaS BI system a year ago to enable more of its business users to build their own reports. To set up the report templates, “you really want someone from the business in the discussions from the beginning, to know what [data] is available and what KPIs make sense,” Buck said, referring to key performance indicators.
Don’t forget an exit strategy. There may come a time when you want to switch SaaS BI vendors or shift to an on-premises BI system. With that in mind, analysts said, it’s important during contract negotiations to pay attention to the clauses around termination of a SaaS business intelligence deal -- for example, by making it clear who owns the data, how the data can be moved out of the cloud-based BI system and whether there are any hidden fees or penalties associated with switching to another vendor.
Beth Stackpole is a freelance writer who has been covering the intersection of technology and business for 25-plus years for a variety of trade and business publications and websites.