At a recent conference, I met a professional - I'll call her Michelle - who was looking for guidance about her organization's business intelligence (BI) program. She was facing a situation not uncommon to many organizations today. Over the last several years, her organization had amassed several BI products from seven different software vendors. This accumulation occurred as a result of a decentralized IT decision-making process and the assimilation of acquired companies due to M&A activity.
Although practitioners used the BI software products, supporting them had become time-consuming and resource-intensive. The CIO asked Michelle to identify efficiencies within the IT group in order to reduce costs of maintaining and supporting their current infrastructure. One area that was under consideration was the number of different BI products. Michelle posed a direct question to me, "How many BI products do we really need?"
From an IT perspective, the ideal number is "one." That is, one suite of BI products from one software vendor. A suite of BI products includes enterprise reporting, ad hoc query and analysis, OLAP, analytics, data mining and data visualization capabilities. The efficiencies are obvious: fewer software vendors to manage and fewer products with the same capabilities to support and maintain. By reducing the number of software vendors to one, IT professionals could focus on understanding and becoming experts on one BI product suite, instead of having limited knowledge about many different vendor products. When there is a choice of developing narrow and deep expertise or staying shallow and broad, I recommend developing the deep expertise. By standardizing on one BI suite of products, an entire host of efficiencies are possible. Just to name a few: in-house expertise, reusability and sharing of standard reports or queries, sharing of meta data and security features, a common infrastructure to leverage across the various BI products and, last but not least, user training.
From a BI user perspective, the most important aspects of a BI software product are the graphical user interface, ease of use and the ability it provides to access information. While most BI software vendors have adopted Microsoft Office's drop-down menu approach, icons and navigation, each product differs in features and functionality. As BI products are used, practitioners become familiar with them and over time become knowledgeable about their features and functionality. The greater the usage, the higher the comfort level with the appearance and navigation of the BI software product. Obviously, this assumes that the underlying data structure that the BI software product is accessing is stable and supports their desired performance and information needs. A BI user's investment in time and application of a BI software product grows as they use it.
Once a BI user reaches a certain comfort level, they become very loyal to the BI software product and it becomes very difficult, if not impossible, to entertain using or switching to a different BI software product with similar capabilities. For example, if you use Microsoft Office Word for your word processing software, image the challenges you would face if you had to switch to another word processing program such as UNIX vi, Corel WordPerfect, Lotus Word Pro or Adobe FrameMaker, or vise versa. This is the dilemma that IT professionals are confronted with when they attempt to standardize on one BI product suite or reduce the number of BI products with similar capabilities within their organization.
For organizations new to BI, the opportunity exists to develop common goals and objectives that align IT and BI users' perspectives before purchasing more than one BI software product. Alignment of IT and BI users' perspectives can also be achieved for those organizations that have several BI products but with different capabilities such as Business Object for querying, Cognos' PowerPlay for OLAP/analysis, Hyperion's products for business performance management, SAS's products for data mining and MicroStrategy's Report Services for enterprise reporting. Although a medley of BI products and different vendors are listed in the preceding example, each product serves a specific purpose. Standardizing on a BI product suite has its advantages from an IT perspective for sharing of meta data, security and common infrastructure. However, a BI product suite where all of the components are best in class does not currently exist. Accordingly, standardizing on one BI product for each capability is a compromise and the next best alternative.
Organizations that use BI software with overlapping capabilities need to standardize on one BI product to achieve goals from both the IT and BI user perspective. The reality is that inefficiencies are created when more than one BI product with the same capabilities are used within an organization.
Here's an example to illustrate my point. One of our clients uses three different BI products that provide query and reporting capabilities. The IT group is supporting all three products. However, the redundancy in effort would be reduced by 66 percent if they were to standardize on one BI product for query and reporting. While each functional group, such as finance or marketing, has different BI query and reporting products, redundancy exists because both have common data sets that they are accessing. In addition, the ability to efficiently share queries or reports is eliminated. For example, an analyst in finance who develops a report that assesses the financial impact of a marketing campaign using Business Objects cannot share that report with an analyst in the marketing department if marketing has Oracle Discoverer as their only BI query and reporting product. The inefficiencies associated with multiple BI products with the same capabilities can be exponential for BI users.
In cases where multiple BI products with similar capabilities exist within an organization, a formal software selection process should be conducted with the goal of standardizing on one of the existing BI products. Representatives from every BI user community and functional group must actively participate in the process in order to ensure proper representation. Other factors that influence the standardization process is the number of users of each of the BI products, the frequency with which each group uses it, and executive support, just to name a few.
So, how many BI products does your organization need? Ideally, just one - the best-in-class BI product for each of the capabilities that are needed.
Register or login for access to this item and much more
All Information Management content is archived after seven days.
Community members receive:
- All recent and archived articles
- Conference offers and updates
- A full menu of enewsletter options
- Web seminars, white papers, ebooks
Already have an account? Log In
Don't have an account? Register for Free Unlimited Access