A Decision-Support Program Can Unearth a Trove of Savings
By Todd Bashor

One of the many things you do as an IT manager is gather data that enables your organization to make better business decisions. In other words, you provide decision support. To do this successfully you need both quality data and a sound analysis method. Unfortunately, many programs disregard the necessary processes and software tools, an omission that impedes progress and frustrates internal customers. But early, careful attention to your toolset can improve the quality and quantity of data—and produce better results. High-quality output will, in turn, justify your expenditures and help unearth a buried treasure of cost and time savings.

"Decision support" can mean a number of things. It includes business intelligence, metrics, analytics, dashboards, data mining and enterprise reporting. My definition encompasses all of these terms: A decision-support solution enables stakeholders to get information at the level of detail they need and in the format they want. It advances reporting to a higher level by allowing end users to access, query, analyze and format data in a way that is meaningful for them. All reporting tools do this by definition, although data presentation, level of detail and support for business-scenario analysis might not suit every user.

Organizations with mature systems support both operational-data consumers, who depend on predefined report formats, as well as analytical data consumers, who depend on ad hoc query and formatting. Tools that support mature systems satisfy different types of users; can scale up for large amounts of data; offer interactive Web-based interfaces for querying and customizing the data presentation; are able to publish and distribute reports via the Internet, Web site or e-mail; and, using semantic rules, enforce apples-to-apples data comparison. (Semantic rule enforcement—making sure reports compare equivalent data—is a basic element of a mature system).

A successful program allows stakeholders to get data efficiently, do their jobs more effectively and make to more informed business decisions. However, to get past “Go” you will inevitably have to justify spending time and money. Your priorities will determine how much is spent and where.

Five Points to Help Justify the Program

Improve data integrity. Use problems as leverage. If decision makers depend on data that, despite your efforts, is unreliable, explain that buy-in from management, participation from end users, and a disciplined approach to processes will increase both the amount and reliability of data. For example, after we reported a trend of suboptimal audit results to senior management, our data center managers made data quality a priority.

Support underserved consumers. Focusing on decision support will allow you to improve service for your data customers. Examine your user community and determine how well it uses data collected by asset management processes. You may be surprised to learn who is underserved. When we found out that our chief financial officer needed monthly spending reports for laptop and desktop systems, our group provided better quality and more relevant data than was available from our accounting systems. It was a real coup for us and helped our credibility immensely.

A unified “view” facilitates quicker response and faster decision making. How many times have you been caught off guard by a request for data? It is nearly impossible to anticipate every need, but a solid, well organized system is flexible enough to handle most requests. Streamlining data access and enforcing apples-to-apples comparisons also gives decision-makers more confidence in the data. Now we fulfill most requests in minutes, not hours, giving our business analysts and managers time for deeper analysis and exploration of alternate approaches.

Robust, accessible reporting and analysis tools cost less and are more efficient. Many organizations lack self-service reporting. After waiting days, or even weeks, the report may fall short of expectations, or worse, miss the mark entirely because the decision had already been made. Our decision-support systems make it easy for IT support staff to deliver more relevant reports more quickly; they also allow end users to do so themselves if they prefer. This, in turn, helps our data experts and analysts add value to the business by giving them time for complex, thought-intensive tasks that computers cannot do .

A decision-support system pays for itself. Cost reduction and avoidance opportunities are frequently buried in impenetrable islands of data. Once the data is connected to related information in a meaningful way, the relevant numbers tend to jump out, revealing a treasure trove of savings. For example, our analysts had a hunch that a gap in the IT asset-receiving process cost over $200,000 a year. After the procurement and IT asset data were linked it became clear that more money would be saved by redeploying assets instead of repurchasing them. They key was to receive IT software and hardware and items properly and link the process to the IT asset-tracking system. Lessons learned: Validate hunches, examine processes, and equate data with dollars.

We used the ongoing savings illustrated in these examples to justify the expense of a decision-support program. Other data consumers (customer service, accounting, and marketing) may have different concerns than those addressed by your asset management decision-support program, so tailor your justification to your audience.

An all-or-nothing approach is rarely successful. Even during the planning stage, cut scope in favor of steady progress. Start by identifying and categorizing your organization's data consumers. Consider the information consumed (costs, utilization, efficiency, etc.), type of analysis or report manipulation performed (data in one database versus data from multiple sources), how the data is accessed (self-service or request from IT), and which formats (graphs, charts, tables, spreadsheets, etc.) your customers will want. Distinguish between casual users who simply view “canned” reports and power users who produce reports and analyses from scratch; this will help you evaluate tool options and determine a phased program schedule.

Have a thorough review process. Some individuals may not know about your plans, the existence of the data, or how to use it. They may even have given up looking for it. Interview representative consumers, document and prioritize their requirements, then perform a gap analysis against your existing reporting tools. What do you have? What’s missing? You should be able to formulate a high-level list of requirements and create a road map to help increase your capabilities.

Know what you need. It’s almost pointless to begin otherwise. Explore vendors and options. Consult reviews by industry analysts such as ECP, Forrester, Gartner, or Meta Group. Although the business-intelligence market is maturing, it’s still fragmented. In the short term you may not find a vendor able to meet all your requirements.

Plan in hand, push for progress. After you’ve put your business plan together and secured funding, start a pilot program. The firsthand benefits of a state-of-the-art tool will create excitement and hopefully, lead to more support for the initiative. Use the pilot program to gather feedback, develop an implementation plan, and refine your program’s road map. Once you feel comfortable, make your move to put the tool into production. Even if it doesn’t meet all requirements immediately, your goal is to make it an indispensable source of information. Your efforts to plan, research, and promote the success of the pilot program are sure to help you achieve your goal.

Todd Bashor manages Intuit's asset management program. Meet him at SAM Summit ’05 in Los Angeles on May 16 to learn more about this topic and Intuit’s program when he presents Creating Effective Reporting & Business Analytics.