Not all business intelligence (BI) software buyers – or sellers, for that matter -- are created equal.
By submitting your personal information, you agree that TechTarget and its partners may contact you regarding relevant content, products and special offers.
So says Bill Hostmann, an analyst and research vice president at Stamford, Conn.-based Gartner Inc. In a recent report, Hostmann identified a number of distinct types of BI buyers and warned potential BI customers to identify which category they fall into or risk the consequences.
"Those who fail to recognize which segment they are in -- and the associated differences between vendors, products and best practices -- will expose themselves to additional costs relating to procurement, integration, governance and, potentially, failed initiatives," Hostmann wrote.
The first group he identified is buyers of BI software for departmental-level deployments. They are most often looking for analytical applications, usually from data visualization specialists or SaaS-based BI vendors, for a narrowly defined set of requirements such as sales tracking or supply chain analysis.
Hostmann said that it is important for these buyers to recognize that the tools used for departmental deployments require a high level of sophistication among end users. "It's going to be very much a power-user skill set that you're going to need," he said. Training is therefore necessary if the target end users lack the skill to take advantage of these types of applications.
Departmental BI deployments also perpetuate the data silo problem, of course, but Hostmann doesn't think this is any reason to avoid them. Departmental deployments can be extremely useful at targeted functions, he said, so "don't stand in the way of giving somebody the tools they need to do the job." He said, however, that IT should educate users that significant data governance and data quality work will need to be done if departmental analysis is ever shared with the rest of the organization.
"It is important to encourage their use, but it is also important to encourage best practices," Hostmann said. It is important for users to understand what it means to have inconsistent definitions and to develop some method for annotating analysis that explains data lineage and data definitions for consumption outside the department, he said. "Things like these that you might find automatically with a more comprehensive BI platform, you're going to have to do more manually when you're publishing results from one of these departmental-level applications."
The next type is the buyer that has deployed BI and data management technologies from multiple best-of-breed vendors throughout the enterprise. In most cases, this type of BI customer has already agreed on definitions for technology-independent business models, rules and services for BI applications, according to Hostmann.
"Recognize that, given the complexity of multiple vendors and technologies, failure to define and provide the right mix and levels of sponsorship, governance, and program and demand management will seriously limit the success of BI investments, deployments and adoption in this segment," Hostmann advises.
He also suggests that buyers of BI technology from multiple vendors establish BI competency centers, also called BI centers of excellence, to plan the organization's BI architecture and manage the development of standardized BI technologies throughout the enterprise.
Hostmann identified a third group of BI buyers: those that decide to purchase all their enterprise applications, including BI technology, from a single vendor. That limits buying options, for the most part, to one of the so-called mega-vendors – IBM, which acquired Cognos; SAP, which bought Business Objects; and Oracle, which purchased Siebel and Hyperion.
"The general feeling among this segment of buyers is that 'a single vendor will simplify and reduce the IT costs of vendor management and the development and deployment of applications,'" Hostmann wrote. That feeling, however, is not always correct.
In some cases, he said, while the mega-vendors are busy integrating their newly acquired technologies, not all applications from the same mega-vendor can yet communicate with one another on the metadata level.
"The way that the users navigate through the information is through the metadata breadcrumb trails, if you will," Hostmann said. "And if those breadcrumb trails don't connect together, then the users are not going to be able to navigate through that data, and you're going to have a real serious problem."
Potential BI buyers in this group should ask vendors about their current level of metadata integration, as well as about their product roadmap plans for future integration, according to Hostmann. He also urged buyers to do a thorough proof of concept and talk to reference customers with similar requirements.
In addition, they should "evaluate the overall product line and partner 'ecosystem,' not just its individual tools," he wrote. "They should also recognize that their negotiating position for subsequent purchases may be weaker if they commit themselves to a single strategic vendor."
Dig Deeper on Business intelligence best practices