Skip to content
Quick Start for:
Texas Performance Review 
Capital Metro 
Chapter 9
 
Information Systems


Capital Metro's board and managers are fully aware that the authority's core computer systems are antiquated and must be replaced. However, Capital Metro has a long history of mismanaging the implementation of large contracts such as these. New computer system purchases pose a substantial risk to Capital Metro. If the new systems are not planned, installed, and managed properly, Capital Metro could waste millions of dollars and operate under reduced efficiency for the next decade.

Capital Metro collectively lacks the skills needed to develop appropriate specifications for new, multimillion-dollar systems; implement and test them; or ensure that they meet the long-term needs of managers and staff. The authority quite rightly is seeking outside assistance to develop contract specifications and negotiate final contracts. In the end, though, it will be Capital Metro's responsibility to rigorously monitor and enforce vendors' compliance with project timelines and design specifications.

The upcoming purchase of major new information systems offers an excellent opportunity for Capital Metro to rethink its current administrative processes and business procedures and reshape them to take the best advantage of new technologies. The authority should conduct the planning needed to ensure that financial reporting, cost accounting and cost allocation, and route and ridership data can be combined to provide useful performance measures for managers and the board. The new system should be expected to accurately report cost and taxpayer subsidy information on all services and routes provided by Capital Metro.

TPR's proposals are intended to help Capital Metro obtain the information systems it needs for the next decade and put them in place on time and at the right price.

TPR recommends that Capital Metro immediately consolidate the authority's information systems staff, functions, and resources. This also was recommended by two separate consulting studies in 1997, and should be done without delay. This recommendation is even more important now because Capital Metro is in the process of several major information system procurements that need to be managed rigorously and in constant coordination with existing systems. Capital Metro's managers should develop a strategic plan for information technology to guide the agency's needs for the next five to seven years.

TPR recommends that Capital Metro establish a working committee to monitor all information procurement projects budgeted at $1 million or more, to help guarantee that the systems chosen are the best available for the authority's long-term needs.

Capital Metro should continue to hire independent technical firms for all large information systems projects to help assess the authority's needs, write the contractual specifications, manage negotiations, and monitor implementation. Finally, TPR recommends that the authority take advantage of free or low-cost independent advice such as assistance from state and local agencies.

In all, the recommendations in this chapter would save Capital Metro $20,000 in fiscal 1999 and $100,000 over the next five years.

PROPOSAL 44

Promptly consolidate information systems staffing and resources.

Background

Capital Metro's information systems and technical personnel are scattered among a number of departments including Information Systems, Vehicle Maintenance, Finance, Planning, and Customer Service. The fiscal 1998 operating budget for the Information Systems Department alone is $733,464, but this amount does not include related costs incurred by other departments.1 It is difficult to determine Capital Metro's total information systems budget since the costs are divided among several departments. Capital Metro's annual budget does not provide a convenient summary of total costs for computer hardware, software, maintenance, training, consulting contracts, and personnel.

Capital Metro is in the process of replacing some older computer systems. Capital Metro's fiscal 1998 budget estimates its expenditures for new computer equipment and software at $4.2 million in 1998 and $1.7 million in 1999, leveling off to an annual average of about $440,000 in each of the following four years.2 Additional amounts not specified in the budget may be needed for outside contracts to help implement new systems and train technical and operational staff in their use.

Centralization recommended

A July 1997 performance audit by KPMG Peat Marwick LLP recommended that Capital Metro centralize its key information systems under the management of a single group. Such arrangements are nearly always more efficient and more secure for an organization of Capital Metro's size.3

Shortly afterward, in October 1997, a large information systems study by the University of Texas Center for Transportation Research concurred, stating that:
...an integrated information system is now essential if the agency (Capital Metro) is to move forward in the area of improved information for management. There are many sources of information presently flowing into or available to various departments with Capital Metro. This information is typically used...to address a limited specific need and...is not available to enhance and support other activities that may benefit from it. This results in duplication, inconsistency in information format and content, and inefficiency. Integrated information systems across planning and operations can contribute greatly to more effective decision-making at Capital Metro, translating into better services.4
Capital Metro has not followed the recommendations of either of these studies. Rather than consolidating its management of information systems, in fact, Capital Metro has instead further decentralized it.

For instance, in February 1998, a key staff member of the Information Systems Department was transferred to the Operations Division. A number of technical employees have been shifted to Operations, in part to complete the implementation of the $1.3 million Maintenance Management Information System (MMIS), which is intended to track and schedule vehicle maintenance and maintain parts inventories.5 Both the July 1997 audit and the October 1997 study specifically recommended that management of this system be consolidated into the Information Systems Department. At the end of March 1998, the MMIS was about 90 percent operational.6

However, TPR believes that Capital Metro faces a risk that key components of the MMIS will not communicate effectively with Capital Metro's new financial and human resource information systems, which are scheduled to be obtained in the next 18 months. One of the main risks is that a procurement module of the MMIS may not easily integrate with the new financial information system.

Automated Trip System

The Customer Service Department's Automated Trip System provides a case study of the problems that can develop when major information systems are developed as "stand-alone" projects without centralized management and open communications among departments. This system-which involved initial costs of at least $452,838 for hardware, software, and consultants-is significantly behind schedule and its costs may grow as a result. Capital Metro was unable to provide a long-term estimate of future costs for this project but they will be substantial, involving annual costs for hardware and software maintenance and updates, updating of geographical and bus route information, and technical support staff.

In July 1997, the Customer Service Department contracted with Mantech Systems Engineering Corporation, an Arlington, Virginia software company, for a $365,000 "automated trip management system." The system is intended to allow telephone operators in Customer Service to provide fast, accurate information about bus routes, times, transfer points, and trip planning. This is a worthwhile concept, since at present, the telephone operators have to plow through thick, cumbersome schedule books and ring binders to answer the huge volume of phone calls they field each day from Capital Metro riders.7 Capital Metro received five responses to its request for proposals for this system; there are few vendors for transit-specific software such as this. The Mantech contract includes all computer hardware, software, and implementation services.

The project was to be completed by March 31, 1998; at present, however, the final completion date is uncertain due to the vendor's continuing inability to deliver a workable system.

Part of the problem is that Customer Service has not solicited the ongoing advice or expertise of the Information Systems Department. According to the manager of Information Systems, the project has had severe problems in part because the Customer Service Department did not thoroughly validate the project's cost-effectiveness, perform technical tests of the system it was purchasing before the contract was signed, or closely monitor the vendor's work. The contractual specifications for the system were written by an outside consulting firm, Macro Corporation of Horsham, Pennsylvania, at a cost of $87,838 for work beginning June 1996 through the present. While outside technical advice was needed, the Information Systems Department was largely excluded from the process; its only involvement being a two-day review of the final contract with Mantech.8

The Automated Trip System, which was to have been fully operational by March 1998, has been plagued with major problems. On March 16, a team of three Customer Service employees, one Contracts and Procurement employee, and a repre-sentative of Macro Corporation visited Mantech's place of business in Chesapeake, Virginia to conduct a one-week test of the Automated Trip System. The team found so many fundamental problems with the system-63 in all-that it halted its testing after only a day and a half and returned to Austin.9 This trip, including the cost of technical advice from the Macro consultant, cost Capital Metro $12,500.10 On March 20, 1998, the manager of Contracts and Procurement sent a formal letter to Mantech citing the firm's failure to perform a long list of basic contractual requirements.

As of March 1998, Mantech had been paid two progress payments totaling about $100,000.11 According to the manager of Contracts and Procurements, Mantech has been asked to reimburse the authority for the full $12,500 spent on the abortive field test in March, and no further payments will be made to the vendor until this amount is refunded.12

While problems remain, Capital Metro believes that this project will be back on track some time in the late summer 1998. In early June 1998, Capital Metro employees and a Macro consultant revisited the vendor and found the software substantially improved, though more work and testing remains. Capital Metro will have to incur more costs for another visit, termed a "factory acceptance test," before it makes another progress payment to the vendor and begins work on the final phases of the project. Note that, once again, the Information Systems Department was left entirely out of the loop concerning either trip to Mantech.

Sending computer equipment out of town for custom software installation is unusual. In most cases, software is shipped to and loaded at the client's place of business. The Mantech contract has other unusual aspects. According to the new manager of Contracts and Procurement, the project contract has no performance bond, which is standard practice for contracts of this type. A performance bond would help Capital Metro cover its losses should the vendor fail to execute the contract successfully. The contract was executed in July 1997, under a previous procurement manager, and Capital Metro has no documentation explaining why it failed to require a performance bond.

Another oddity is that the software initially will not be delivered to Capital Metro's place of business. Instead, Mantech is purchasing 17 personal computers, two computer servers, and other hardware from an Austin subcontractor; Mantech already has purchased six personal computers and two servers and shipped them from Austin to its headquarters in Virginia. Once Mantech Systems installs and tests the software, Capital Metro's computers will be shipped back from Virginia to Austin. Later, the department will purchase 11 more personal computers in Austin, and the Automated Trip System software for these will be installed directly at Capital Metro.13

Sending computer equipment out of town for custom software installation is unusual. In most cases, software is shipped to and loaded at the client's place of business. It makes more sense to have the Automated Trip System installed at Capital Metro, so that all software and computer networking problems can be field-tested under actual working conditions, instead of the vendor's place of business. Capital Metro's rationale was that it preferred testing to be done at the vendor's site so as not to interrupt business in Capital Metro offices. This might make sense if the Automated Trip System was a stand-alone system. But the software also needs to be linked and tested to computer systems in at least two other departments at the authority. According to the Customer Service Department's plans, the Automated Trip System eventually will be accessible to the public through Capital Metro's Internet site, though a date has not been set for this goal; this website is maintained by the Information Services Department.14 In addition, the system's database will need substantial updating at least twice each year by the Planning Department to reflect bus route changes and changes to Austin-area streets. For this updating, the Automated Trip System must be linked to the Planning Department's "Trapeze" and "ArcView GIS" computer systems.

Because few vendors offer this sort of transit call center system, Capital Metro wisely included a provision in the contract for the vendor to provide an escrow agreement that guarantees access to the program's source code and other files should the vendor be unable or unwilling to do so, which in theory would allow the authority to maintain the system itself.

Nonetheless, TPR believes that if this project had been managed with the oversight and involvement of a consolidated Information Systems Department, instead of as an ad hoc, stand-alone project, many of the problems stalling it could have been avoided. Other departments, moreover, should have been involved throughout the project, particularly Planning, which by default will have substantial ongoing responsibilities for updating the system's database.

Strategic planning for information systems

Any large organization needs an information technology strategic plan to manage the performance of its information systems and to anticipate its informational needs. But such plans are not effective when produced in a vacuum. A workable information systems plan must be harmonized with an organization's overall strategic plan.

One of the keys to successfully managing expensive information system projects is to make sure that the projects support the organization's mission and that the projects' strategies and goals are clearly understood by all stakeholders. To do this, information system managers must listen to their customers and translate customer expectations into practical projects that are planned and implemented in open cooperation with their customers.15

As discussed in Chapter 2, however, Capital Metro also lacks an overall strategic plan. This has made it difficult for Information Systems and other departments to plan and coordinate future information technology projects. Instead, computer projects end up being developed by separate departments, without the involvement of other departments and without sufficient budgetary and quality control.

To its credit, in mid-1997 the Information Systems Department prepared an introductory strategic plan for Capital Metro's information systems.16 The department manager believed that in the absence of any apparent interest by authority managers, a stand-alone plan for the department would be better than no plan, and could be used as a starting point for coordinating technology projects among departments. However, Capital Metro as a whole has not acted on Information Systems' strategic plan, in effect shelving it with little discussion. The board considered hiring an outside consultant to help prepare an information system strategic plan, but this has been placed on hold pending consideration of Capital Metro's general strategic direction.

TPR believes that elements in the Information Systems plan should be revisited, particularly when Capital Metro begins preparing its overall strategic plan. The plan includes a vision statement, mission statement, goals and objectives, major technology projects to support future operations, and clearly stated strategic five-year assumptions. The Information Systems strategic plan also provides detailed lists of performance measures and best practices to hold the Information Systems Department responsible for results to its customers.

But important items are lacking from this plan, most notably requirements for guidance or approvals by the board, general manager, or other senior managers. The Information Systems strategic plan also makes a number of assumptions, including the following:
  • Capital Metro will develop a light rail project, with construction beginning in five years.
  • Capital Metro will directly provide bus service to the University of Texas and discontinue contracting for this service during fiscal 1999.
  • Capital Metro will open a new bus service and maintenance facility in north Austin in 1999.
  • New farebox systems will be purchased that will serve as the authority's primary source of ridership information.
  • Capital Metro will provide communication and automation services to contractors who provide transportation services on behalf of the authority.17
  • It would be a mistake for the Information Systems Department to proceed with the development of future information technology projects on assumptions that are contingent upon uncertain future decisions by the general manager, the board, and the electorate. Clearly, however, the department realized the limitations of this effort, as it also requested funding for an outside consultant to help complete a more detailed strategic plan.

    Still, this example serves to illustrate the point that planning for the optimal use of information systems is the responsibility of the entire management team, with input from the board. If the task is shifted by default to a single department, too much money is tied up in the wrong systems; computer systems do not work efficiently with each other; and the needs of customers are not met.

    Computer training costs

    Since 1995, Capital Metro has had an ongoing contract with Productivity Point International for computer software training. The contract has been amended and renewed during the last five years for an average of about $51,000 per year.18 In 1995, Capital Metro was moving from a mainframe to personal computer environment, and Capital Metro's professional and administrative employees needed basic training on personal computer software. By all accounts, the Information Services Department was stretched thin during this transitional period, and Capital Metro wisely decided to outsource this initial round of training. More recently, Capital Metro stated that it has negotiated lower training rates through an interlocal agreement with the Texas Department of Information Resources.

    However, the transition to the personal computer environment is complete, and much of Capital Metro's computer training needs now are for routine training in applications such as word processing, spreadsheets, and general personal computer operations. Routine software training often can be conducted less expensively in house. In addition, administrative employees will need training on the authority's new financial and human resource information systems, but the experience of various state agencies studied by TPR suggests that large-scale outside training provided directly by software vendors can be extremely expensive.

    Capital Metro should explore ways to mitigate computer training costs, while at the same time providing prompt training for its staff. A useful alternative is a "train-the-trainers" program; a small group of Capital Metro employees could receive extensive training from the software vendors and then become responsible for developing and delivering computer training programs for the rest of the authority's administrative staff. Such training would be cheaper and can be more effective. In-house computer training can be custom-tailored to meet the specific needs of each department, and the training can be shorter and more to the point. Some of the authority's more experienced computer users also could be recruited to help guide and train computer users in their departments.

    Recommendations

    A. The board and general manager should immediately consolidate all information systems staff, functions, and resources.

    This action was recommended by two separate consulting studies in 1997 and is even more important now. Capital Metro is embarking on several major computer system purchases that should be managed rigorously and in constant coordination with its existing information systems.

    As a first step, computer operations of the Customer Service, Planning, and Vehicle Maintenance Departments should be consolidated into the Information Systems Department. The consolidated department should be involved in all procurements for information systems.

    A priority for the new Information Systems Department should be to ensure that Capital Metro's new computer systems for vehicle maintenance, procurement, finance, human resources, and customer service are closely integrated and operate smoothly with each other.

    B. Capital Metro should include performance bonds in its information technology contracts and ensure the continued usefulness of specialized software systems by using software code escrow agreements.

    Some of Capital Metro's largest computer software investments are with specialized transit software vendors. Capital Metro should take steps to ensure that it can continue operating this software in the future if the software vendors go out of business or otherwise stop supporting the software. Capital Metro has such provisions in its contract for the Automated Trip Planning system, and software code escrow agreements should be used for all major purchases of information systems.

    As a standard business practice, Capital Metro should establish three-way agreements in which software vendors agree to place their computer software source code with a designated escrow agent. If a vendor ever fails to support its software, Capital Metro then could obtain access to the software code and modify it as needed to keep it running in the future.19

    C. Capital Metro should develop an information technology strategic plan to guide the authority's computer usage and acquisition for the next five to seven years, in connection with an overall strategic plan.

    One of the first priorities of the newly consolidated Information Systems Department should be to develop a detailed information technology strategic plan to serve as a blueprint for all information systems, in place or planned. The plan must support the authority's overall strategic long-term plan recommended in Chapter 2, and should be clearly written in language that all managers can understand.

    The information technology strategic plan should consider issues such as the integration of purchasing functions with the new financial system and the electronic dissemination of budget and performance management information for easy access by managers and staff.

    The plan also should address future needs for competitive salaries for technology employees; detailed descriptions of the skills needed for operating new information systems; and recruitment plans for hiring and retaining qualified information technology specialists.

    D. Capital Metro should use in-house personnel to provide routine software training for administrative workers.

    Now that Capital Metro has shifted from a mainframe to a personal computer environment, much routine training could be conducted less expensively in-house using a "train-the-trainers" approach in which a few employees receive extensive outside computer training and then become responsible for training the rest of the staff. Some of the authority's more experienced computer users also could be recruited to help guide and train computer users in their departments. Capital Metro should contract for computer training sparingly and selectively, preferably for highly advanced or specialized computer applications. The authority should reduce training expenditures by 40 percent beginning in fiscal 1999.

    Whenever Capital Metro hires new information systems staff, one of the job skills sought should be the ability to train users

    Fiscal impact

    These recommendations would save Capital Metro money by eliminating duplication and gaps in computer systems management and by ensuring that additional technology is purchased only when it fits with an overall plan and when proper contract oversight is in place. These savings could be substantial but cannot be estimated.

    Minor savings would result from reducing Capital Metro's reliance on outside contracts for computer training. A 40 percent reduction in outside contracting for this training would save the authority about $20,000 each year.

    Fiscal Year Savings
    1999 $20,000
    2000 20,000
    2001 20,000
    2002 20,000
    2003 20,000


    PROPOSAL 45

    Strengthen planning and oversight of major new information systems.

    Background

    Capital Metro's financial information systems are more than 11 years old. The authority plans to replace them due to "Year 2000" problems. The widely publicized "Year 2000" problem is caused by older software improperly reading dates beyond the year 1999, which in turn causes severe software and computer system malfunctions. Capital Metro must act quickly to replace its aging systems; yet the board and senior managers also must remain aware that they are making expensive and potentially disastrous decisions the authority will have to live with for the next decade.

    Capital Metro collectively lacks the technical and negotiating skills to buy or contract for the best financial and human resource information systems. Computer-related procurements have been a perennial stumbling block for the authority; for example, Capital Metro took more than two years to buy and install its automated vehicle maintenance system.20

    Oversight of information systems purchases

    In the early 1990s, Capital Metro had an Information Technology Planning Committee consisting of 11 managers and senior staff members. Its goals were to obtain authoritywide input on information technology planning and ensure that all managers were involved in the management of computer and software purchases and projects. According to the manager of the Information Systems Department, this planning committee worked quite well until it was disbanded by the then-general manager.21 Since this committee disbanded, Capital Metro managers have communicated only infrequently concerning information system plans and purchases.

    In addition, until fairly recently, Capital Metro lacked a sufficient number of information systems employees. The authority lagged in keeping its information technology up to date until a new information systems manager, hired in 1996, was allowed to begin hiring staff with deeper technical skills. But the Information Systems Department has had to play "catch up" to meet the daily demands of its users, and has not been able to monitor some new information system projects.

    Recently, Capital Metro began using outside consultants to act as quality assurance experts for the authority's acquisitions of major new information systems. In planning its purchase of its new financial and human resource systems, Capital Metro has sought the assistance of the Texas Department of Information Resources (DIR) to facilitate vendor negotiations. Capital Metro, under DIR's advice, also is hiring an independent technical quality assurance firm to assist it in writing contract specifications, negotiating with vendors, evaluating technical proposals, integrating new computer systems with older ones, and providing ongoing testing and quality assurance during the implementation of new systems.22

    Capital Metro believes the use of an independent technical advisor to monitor all aspects of its purchase of new systems, from creating specifications to ensuring that vendors deliver quality products on time, is well worthwhile, and TPR concurs. At the same time, Capital Metro also must closely monitor both the technical and fiscal aspects of these contracts throughout their implementation.

    Specify expectations in detail

    The most critical issue involved in information systems procurements is the need for a thorough understanding between Capital Metro and any vendor concerning the authority's expectations for its new systems. Well-written information systems contracts include very detailed expectations and specifications. Contract deliverables, quality checks, and delivery timelines must be thoroughly specified in the written contract. All oral conversations with vendors should be documented in writing and kept in well-organized files.

    Given Capital Metro's history of weak contract administration, it is essential that basic contract controls be established for new information system purchases. After vendors are selected, Capital Metro should document and prioritize its key expectations for the implementation process and the final capabilities of the systems it is buying. After this information is summarized, it should be presented at a final meeting with the winning contractors to verify that the expectations of Capital Metro and its contractors mesh. Only after this process should the final contracts be signed.

    Example: purchasing system software

    Contracting and purchasing have been a major source of problems for Capital Metro throughout the 1990s. The main problems have been human-poor management, a lack of monitoring and oversight, and a lack of appropriate policies and enforcement of those policies. Another potential problem, however, has resulted from Capital Metro's automated purchasing systems. Under Capital Metro's previous general manager, the authority attempted to improve its purchasing systems by adding a procurement software module to the new vehicle maintenance system and using it as the authority's general purchasing system.

    After an almost two-year implementation process, Capital Metro began using its Maintenance Management Information System (MMIS) on January 28, 1998. The MMIS cost Capital Metro nearly $1.4 million, as well as an unknown amount of additional labor and miscellaneous costs. The MMIS is intended to provide Capital Metro's vehicle mechanics and managers with a more effective means of tracking vehicle maintenance schedules and an automated process for ordering parts, assigning tasks, and reviewing vehicle maintenance histories.23 The MMIS procurement module, however, was designed specifically for maintenance operations and not necessarily for the wide-ranging needs of a major organization such as Capital Metro.

    The MMIS is an "open system" based on the widely-used Oracle data management system and the Windows NT operating environment. Still, very few vendors specialize in this type of asset maintenance management software. According to Capital Metro, there are only four major U.S. vendors of this type of software. On the other hand, general procurement software is commonplace and is marketed by dozens of major national and international vendors. Capital Metro needs a flexible, standardized procurement package that links directly and easily to its financial system, sharing common data to the full extent possible. Procurement modules are very common in financial information systems.

    According to interviews with Capital Metro staff, it may be technically possible to link the MMIS procurement module to Capital Metro's forthcoming financial information system, but it will not be an easy, seamless process. The MMIS procurement module and Capital Metro's planned financial systems both will be based on Oracle software. The vendor of MMIS and at least one major vendor of financial information systems recently informed Capital Metro that their systems can be integrated. However, the costs and practicality of integrating MMIS with other financial systems cannot be determined without further technical study by Capital Metro or an outside consultant.24 Moreover, being industry-specific, MMIS exposes Capital Metro to the risks of uncertain future vendor support.

    Capital Metro added the procurement module to the MMIS as a stop-gap measure to improve its well-publicized contracting and procurement problems. Capital Metro claims that the procurement module added no additional costs to the MMIS vendor contract, although considerable staff time and effort were involved getting the new procurement system to operate properly. As part of its ongoing project to completely replace its financial information systems, the authority is considering the purchase of another procurement software package to better meet its long-term needs; if it does so, it will discontinue its use of the MMIS module. This is probably the best solution. Nevertheless, the whole episode illustrates the sort of needless duplication that can result from Capital Metro's ad hoc approach to the development of important computer systems.

    Plans for a "data warehouse"

    Capital Metro has been considering purchasing the hardware and software needed to create a "data warehouse," which would consolidate the organization's electronic data in a uniform format for easy access by multiple users for a wide variety of purposes. A data warehouse integrates information from multiple departments for faster and more informed decision-making, while eliminating the need for special computer runs by an information systems department. When properly developed, a data warehouse can improve customer service and increase employee productivity through fast and comprehensive access to vital information.

    The initial estimate for a Capital Metro data warehouse was $120,933. This price included a dedicated database computer server and upgrades to the agencyís database license. TPR was specifically asked to review and comment on Capital Metro's plans for a data warehouse. Capital Metro wanted to explore the concept because of the authority's present, decentralized computing environment. According to Capital Metro, the authority operates in two hardware environments and three networking environments, and houses five critical applications on an older VAX mini-computer and eight more on several computer servers. The Information Systems Department justifies the need for a data warehouse because its current computer environment makes it extremely difficult to share, integrate, and maintain information. 25

    TPR found, however, that the initial cost estimate for Capital Metro's data warehouse is too low, as it does not include the full spectrum of costs needed to maintain the system. Capital Metro's estimated cost for the data warehouse project does not include estimates for the substantial costs of implementation, data analysis and conversion, reporting tools, and user support and training. In addition, many other questions need to be addressed by Capital Metro before proceeding with the project, including whether the hardware can be easily expanded as the data warehouse itself grows; whether Capital Metro has the expertise to establish the data warehouse on its own; and whether the board is prepared to provide more funding for ongoing support of the data warehouse.

    The board has placed the data warehouse project on indefinite hold. This is prudent, as the installation of a data warehouse prior to the creation of a strategic plan for information systems carries some risk that the data warehouse may not meet the authority's future needs.

    Recommendations

    A. The board and general manager should establish a working committee to monitor all information procurement projects budgeted at $1 million or more.

    Capital Metro is in the process of purchasing multimillion dollar systems that will be expected to function for at least the next eight to ten years. A joint management-board working committee would help ensure that the best systems are procured to meet the authority's long-term needs. The committee's overall goal should be to use new information systems to better manage Capital Metro's operations and, where possible, reduce administrative staffing through increased productivity.

    B. Capital Metro should continue and expand its new practice of hiring independent technical firms for large information systems projects to help assess the authority's needs, write contractual specifications, manage negotiations, guide project implementation, and provide quality checks during the implementation of contracts.

    Given the time and money Capital Metro will spend on new information systems, it should make use of independent third parties to manage contract specifications, negotiations, and project implementation. Independent firms also can help Capital Metro by establishing realistic cost estimates for the type of information systems typically used by organizations of Capital Metro's size and type.

    An outside technical firm should be used to determine whether a new procurement module should be purchased as an integrated part of the authority's new financial information system. The outside firm should be assigned to prepare an analysis that compares the procurement module of Capital Metro's vehicle maintenance system with the use of general purchasing software that integrates with the new financial information system.

    The use of free or low-cost independent assistance also is recommended, such as that available from state and local agencies. Capital Metro also should seek to develop ongoing contacts with other state and national transit authorities for computing expertise.

    C. Capital Metro should continue studying the long-term benefits of establishing a data warehouse so that key management information can be accessed timely and accurately and effectively shared among departments.

    Planning for the eventual procurement of a data warehouse system can proceed so long as it is incorporated into Capital Metro's overall strategic planning process. Before buying a data warehouse system, however, Capital Metro should address these basic questions:
  • Is the selected hardware expandable, so that capacity can be upgraded later without loss of value?
  • Will all planned products associated with the data warehouse support the database software?
  • Does the Information Systems Department have the expertise to implement a data warehouse without outside help?
  • Is the board prepared to provide funding for additional resources for the implementation and ongoing support of a data warehouse?
  • Fiscal impact

    Capital Metro's board already has a 1998 capital budget of $4 million for the acquisition of major new information systems, including those for financial and personnel applications. These recommendations would entail no additional costs. These recommendations should save Capital Metro money by making sure it buys only the technology it needs, with the flexibility to meet future needs. Savings from better information systems and cost avoidance cannot be estimated.


    Texas Comptroller of Public Accounts Window on State Government
    Contact Us
    Privacy and Security Policy