Information_Systems

HOW TO BUY INFORMATION SYSTEMS WITHOUT DRIVING YOUR ORGANIZATION CRAZY A citywide information system involves a complex web of sophisticated products and services. It consumes significant resources, has an impact on every, department in a local government, and can change the way a person does things. But buying a system need not be an intimidating task. The fundamental decisions are managerial, not technical, and should be made by civilians. This is not to minimize the role of technical staff but to stress that the process is driven by established goals and expected results, rather than by technical merits. You buy an information system, not a computer. The difference is more than semantic. Knowledge of computer technology is helpful but not essential to decision making, which entails performing fundamental management tasks without succumbing either to technomania or technophobia. Both roads lead to dubious results. The procurement process has four distinct phases, each of equal importance: definition of objectives, process engineering, request for proposals, and vendor selection. (The fifth phase, implementation, deserves separate treatment.) A core project committee should steer the process, with a larger interdepartmental committee to coordinate departments and ensure adherence to overall priorities. But be careful: A camel is a horse created by a committee. It is important to involve staff to the maximum extent but not to let committee-made decisions dilute goals. Place a decision-empowered individual in charge of the project. Define the Objectives The first, too-often-neglected step is definition of operational objectives. Determine what the organization wants to achieve in functional, not technical, terms. A system is a tool used to achieve a goal, not a goal in itself. This is not the time to debate brands of computers, operating systems, speed, or capacity. That is defining inputs. On any project, decisionmakers should define outputs. Three categories of desired outputs are useful in defining objectives: improvement in efficiency of current tasks, improvement in effectiveness of current tasks, and performance of new tasks that were previously impossible or not cost-effective. The overall questions are: What does the local government want to achieve, what tools are available to do it, and will the result be worth the expense' Technology briefings, including vendor demonstrations, are useful at this stage to determine the proper balance between the technologically possible and the economically feasible. Engineer the Process A new system does not solve procedural problems; automating poor procedures just speeds up a poor process. Take a hard look at the organization and the way it does business in all departments. Change may be necessary to absorb new technology and to achieve goals. Now is an ideal time for organizational change. Map current process flows, seek ways to improve them, and set criteria for a system based on the way things should be, not on the way they are. This may mean overcoming the inertia created by departmental fiefdoms, habit, collective bargaining, congenital resistance to change, and outdated ordinances. The challenge is to adapt procedures and staffing patterns to technological change. Without an honest internal process review and adaptations, an information system will provide little more than cosmetic change. The fundamental guidelines for efficient process redesign are to capture data at the source and to create parallel processes with timely access. This approach eliminates redundant entry and ensures that the data are accessible by all departments when and as needed. Issue a Request for Proposals Prepare the document involved in a request for proposals (RFP) carefully; it will last a long time. It should explicitly state or refer to the outcomes of the previous steps and should provide a clear picture of the local government's operational requirements, their purpose, and their context. The RFP should enable vendors to respond with realistic (not "overkill" or "lowball") product proposals and system configurations and should allow tabulation of those responses for fair comparison. The RFP might allow multiple vendors for the various components or might require a prime vendor, who may subcontract as necessary but who accepts responsibility for the entire system performance. The former approach may result in lower component prices, particularly for hardware, but it requires staff to coordinate vendors and can lead to finger pointing when things go wrong. As systems get more complex, the prime-vendor approach offers the advantage of contractual assurances of integration, as well as simplification of management. The requirements statements could take one of two forms: detailed specifications or general objectives statements. The former alternative requires a meticulous checklist for every function, while the latter presents objectives and requires vendors to furnish solutions. With the detailed approach, one gets precisely what is asked for and usually no more. The general-objectives approach involves the vendors in the solution, requires them to share long-term responsibility for it, and encourages innovative solutions. Select a Vendor The relatively small number of vendors that specialize in local governments provide packaged applications, which should be adaptable to special needs. Most software products probably will stand lower on the technology curve than their commercial counterparts. The dizzying pace of technological change makes it impossible to get products that are both tried-and-true and state-of-the-art. Because whatever the locality buys already is obsolete, the decision will reflect a balance between obsolescence and the "bleeding edge," and will be based substantially on evaluation of the vendor's overall credibility. Vendor selection will create a long-term partnership, with complexities that require regular management attention. Examine the vendor's track record in product development, implementation, and particularly support. A vendor that is investing in the future will have several products in various stages of development. While a local government need not dread "beta" products (new product releases), especially if a vendor's development history is good, the contract should be specific about version control and performance specifications. Every vendor will have good and bad references; get the entire client list. One or two bad references should be instructive but not decisive. A high proportion of poor or mediocre client ratings, however, is cause for alarm. Investigate the firm's financial viability, but be careful of imposing requirements that only can be met by large companies; many good vendors are small and regional. The technology procurement process should not be a frightening prospect. With proper management, local governments can obtain the most cost-effective systems for their communities. Costco shops for right outsourcing contract When Don Burdick took over as senior vice president of information systems for Costco Wholesale Corp. in 2000, his department had 60 programmer spots to fill. Employees were struggling to keep up with routine work. Morale was low. "Every day you came in and were fixing stuff that was broken," said Burdick, estimating that less than 30% of the work focused on projects designed to add value to the company. That put IT badly out of sync with a company that has become the nation's largest wholesale club operator with 452 stores. Over the past five years alone, Costco has grown from $35 billion to $52 billion in sales. The fast-track discounter keeps its mantra simple. Give customers a carefully edited selection of high-quality goods at the lowest possible prices. Top-line driven, Costco operates on skinny margins and needs to keep expenses low, said Burdick, who presented his IT story at the recent Gartner Outsourcing Summit. "The fact we walk on a concrete floor and sell goods from high-rack steel is not a gimmick, said Burdick, referring to the practical, no-frills shopping experience. "It was important that IT reflect the business." Today, Burdick still has programmer spots to fill, but the IT outfit is humming along, rather than groaning. Over the past five years, Burdick has developed a "hybrid model" partnership with US Technology Resources LLC, a Aliso Viejo, Calif.-based company that operates a large software company in Kerala, India. Costco employs about 100 people through UST, about 25 of them onshore at Costco headquarters in Issaquah, Wash., and 75 offshore. Execution has been excellent, Burdick said. Of the 20,000 hours logged to date on the company's large joint projects, the difference between US Technology's estimated time and actual time is less than 400 hours. "Best of all' The IT department really has done things they have talked about for at least five years before I got there and would still be talking about today," Burdick said. The upside was not always so obvious. For one, Burdick reminds people that Costco doesn't consider itself in the outsourcing business. "We don't outsource," Burdick said. "If you've shopped with us or been in our warehouses you know our own employees do everything, including cleaning our own restrooms." Contracting with US Technology was pitched as a way to improve the quality of life for IT and make IT more valuable to Costco. "Our decision wasn't to stop hiring and outsource. It was to keep hiring and partner," Burdick said. Faced with his share of skeptics, Burdick took several steps he believes were critical to making the partnership with US Technology a success. He made a list of projects that were not getting done, but he did not require any of his departments to outsource anything. "I said, 'You can use US Technology, or hire more people. Do what you want. Here is the budget,'" Burdick said. Initially, acceptance of US Technology was very uneven. "Some teams figured it out really early and others did not. There are now UST people in every single department," he said. The onshore ratio was initially 1:2 -- one Indian manager on Costco premises for every two employees in India. That changed to 1:3 and eventually Burdick hopes for a ratio of 1:5. Over the years, the pool of Costco expertise in India has been growing, with more and more of the Indian workforce learning about Costco projects firsthand and bringing that knowledge home. "Work that could never have been done in India, all of a sudden is being done in India. Why' Because they know us, they understand our system, they have worked side by side with our people and they have been oriented just like our own employees," Burdick explained. The company has adapted programs used at Costco University for its Indian workforce, so they understand not just the IT but the warehouse club business and its culture. "The team that works for us refers to themselves as the 'Dream Team,' because they really want to work on the Costco account," Burdick said.