You may think you know what business intelligence (BI) or data analytics in the cloud means, but think again.
Analytics in the cloud doesn’t refer to just a SaaS-based BI application or a hosted
Failure to understand the details of cloud analytics could mean disaster for future projects, said Gartner analyst Bill Gassman, co-author of the report.
“Enterprises will increasingly pursue cloud analytics, but projects may falter unless all members of the project team have the same idea of what it means, in order to avoid working at cross-purposes,” Gassman and fellow analyst Rita Knox wrote in the report.
According to Gartner’s definition, “cloud analytics refers to any analytics effort in which one or more of these elements is implemented in the cloud, be it public or privately owned.”
The six elements are data sources, data models, processing applications, computing power, analytic models, and sharing or storing of results. “We see any number of those combinations being built,” Gassman said.
The problem arises, he said, when an organization decides to pursue cloud analytics but different divisions and departments define the term differently.
A marketing department, for example, may decide to begin using a SaaS BI application without understanding that the organization is still responsible for creating the data models or housing the data in a data warehouse. Meanwhile, the IT department may not even know what marketing is up to.
“My biggest concern is that a marketing organization is going to get deep into looking at these [cloud analytics] vendors without realizing IT and others are going to have to supply a lot of things,” Gassman said.
In fact, much of the confusion around cloud analytics is the result of marketing hype, he said, adding that all types of vendors that offer one or another of the six elements call themselves cloud analytics companies.
SaaS BI vendors like GoodData, for example, refer to themselves as cloud analytics vendors, as do social networking analytics companies like Coremetrics. But they offer very different products. Then there are data warehouse vendors -- Kognitio and Teradata, for example, which are also sometimes referred to as cloud analytics companies -- which offer their analytic database in the cloud or in a hosted environment.
“The danger is people will go down this road and not understand the scope,” Gassman said.
As a result, customers don’t always understand what they’re getting. And in some cases, especially with SaaS BI, he said, vendors are targeting marketing and sales departments with the express intent of avoiding the more tech-savvy IT departments.
So while significant opportunities to create business value via cloud analytics exist, organizations must understand what the term truly means and pick a combination of the six analytic elements deployed in the cloud to best meet their needs, Gassman said.
A good place to start, he said, is with vendors that collect and analyze data that already resides in the cloud. These companies often collect data from various sources and run high-level analytics on them to help customers identify industry-wide trends, for example.
Non-analytic SaaS applications that create analytic artifacts, as Gartner calls them, represent another potential quick cloud analytics win.
“Coremetrics provides a cloud analytics application for website usage,” Gassman wrote. “Because it sees millions of website visitors across its many customers, it can analyze this behavior in aggregate and offer it as benchmark data to its customers.”
But whatever combination of cloud analytics a company chooses, agreeing beforehand on a definition for the term is essential.
“Cloud analytics isn't a single, uniform product or technology category,” the report states. “To gain any business advantage from it, enterprises must have a clear idea of what they mean when they use the term.”