According to industry lore, one of the fundamental reasons for building a data warehouse is that it gives users free and unfettered access to corporate data. With a data warehouse, users are no longer held captive by the MIS staff who alone hold the keys to the corporate database. Or so it goes. In the same vein, a data warehouse is supposed to liberate MIS folks from having to toil long, arduous hours building custom reports using COBOL, FOCUS or some other programming language. With a data warehouse, MIS staffs supposedly can devote more attention to value-added activities, such as developing new applications and mission-critical infrastructure..
In short, a data warehouse is supposed to create a utopian "self-service" environment in which users help themselves to data and no longer have to wrestle with the MIS middleman to get custom reports.
Unfortunately, theory often races past reality. In the trenches of most data warehousing operations, it's clear that users don't want unfettered access to data. We've also noticed that MIS professionals are more involved in decision support activities than ever before. It doesn't take long to discover that the notion of self-service decision support is largely a fallacy an outgrowth of wishful thinking on the part of a few overzealous IT managers and power users.
The Impact of the Web
The primary factor contributing to the demise of self-service DSS is the Web. The Web reduces the overhead involved in supporting remote users, centralizes data and application management, and reduces the time required to give new users access to a corporate resource.
In most cases, the first people to access a data warehouse via the Web are those who have never accessed a data warehouse before! These are employees in lower-level or field-based positions or remote geographic locations, as well as customers and suppliers. The Web has effectively democratized information access, enabling companies to open the data warehouse to virtually anyone they want.
The upshot of the Web is that it has tilted end-user requirements from heavy-duty analysis functions to simple reporting. The casual user, who simply wants to review the same data on a regular basis, now represents about 80 percent of a company's decision support populace. Those power users, who once ruled the decision support roost and pushed for self-service capabilities, now represent a mere 20 percent of users.
What's important to know about casual users is that they don't want to get their own data or build their own reports. They would rather have someone else figure out the intricacies of the software tool and spend the time building a report that meets their needs. After all, as business people they get evaluated on their ability to manage or support a business process, not for crafting a pretty report.
So the question becomes, "Who is going to build all the reports that casual users want to consume but don't want to build?" If the answer is "MIS," then it's likely that a company will regress to pre-warehouse days in which MIS and end users squared off as antagonists over the report bottleneck. Enlightened companies, however, create a tiered distribution system that streamlines the delivery of data warehouse information to those who need it.
Tiered Report Distribution System
In this system, MIS sets up and maintains the data warehouse infrastructure and data model(s). MIS then trains power users on the intricacies of the decision support tool(s). MIS also works with power-user representatives from each department or workgroup to co-develop and maintain a semantic layer that is pertinent to a localized set of business requirements.
The power users then are responsible for creating custom reports that meet the needs of their individual groups. Since these power users are business people, they should be able to generate relevant reports much more quickly than MIS. They can use MIS as a help desk if they encounter technical problems.
Making a power user the intermediary between end users and the data makes a lot of sense. The power user is a business person who works directly with end users on a daily basis and understands their requirements extremely well. There is no translation problem between technology and business.
The only problem is keeping a power user embedded in the business and not overwhelmed with technical tasks from an information-hungry workgroup or department. In most cases, companies need to expand headcount by either assigning IT to the field or adding departmental staff so a power user can work half-time as an MIS analyst.
This tiered system also significantly changes the role of MIS. For this tiered system to work, the data warehouse data model must accurately reflect the business environment users want to query and analyze. Since business is fluid, the kinds of questions asked by users change continuously. Instead of taking orders for reports, MIS must now anticipate the questions those reports are designed to answer and modify the data warehouse accordingly. Otherwise, users will stop using the data warehouse and once again pester MIS for custom reports that they and their power users can't build themselves.
This role change requires MIS professionals to adopt new skills. These include interpersonal skills, such as learning how to communicate, listen, negotiate and interview. There is also an imperative to study the business and understand the terminology, processes, and drivers of growth and profitability. Finally, there are also new technical skills, such as dimensional modeling, that involve a paradigm shift in approach and technique for most MIS professionals.
I've seen instances in which MIS didn't "accept" its new role and treated the data warehouse as just another operational system. The MIS staff figured that once they built the data warehouse, they could move on to other endeavors. By abdicating responsibility for the continuous management of the systems, MIS staff let the data warehouse gradually lose its relevance and value to users. Soon thereafter, executives axed the data warehouse like a pile of deadwood, by reducing funding for the data warehouse or terminating the program altogether.
So beware of the notion of "self-service" decision support. Most users want full service, not self service. Ultimately, this means MIS will be more involved in delivering reports, not less involved. And MIS will need to pick up a raft of new skills mostly interpersonal to ensure the success of the data warehouse.
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