Doing More With Less or Why Change if the Applications Work |-------------------------------------| | Paper presented at CAUSE92 | | December 1-4, 1992, Dallas, Texas | |-------------------------------------| DOING MORE WITH LESS: WHY CHANGE IF THE APPLICATIONS WORK Morris A. Hicks Associate Director, Computer Services University of Hartford West Hartford, Connecticut ABSTRACT While the impact of the economic downturn and unfavorable demographics has hit institutions of higher education, the effect has been compounded on the University of Hartford because it is an independent university with a modest endowment. In this environment, many operations of the University were reviewed with the goal of "Doing More With Less," including the highly successful portfolio of the University's administrative systems. Since information technology has rapidly advanced since these applications were designed, it was concluded that there was a "window of opportunity" to reduce costs by moving to a new set of applications based on newer technology. In addition, administrative applications systems would be strategically positioned for future. Essential to this project are the goals of and support by senior level management and the high level of experience, expertise, and commitment of end user and CS staffs. _____________________________________________________________ Doing More With Less or Why Change if the Applications Work PROJECT COMMENCEMENT MEETING All the project team members of University were invited to the official "kick-off" meeting of the project to change the entire administrative applications. The implementation team members number over 50 and come from a wide range of University departments. At this meeting the University president and senior administrators congratulated all of the new team members and emphasize the significance of this project to University. This conversion project was termed as being absolutely vital to the University and as playing an important role in working smarter and faster. Why the University decided to change from administrative information systems that were working well and how the project started and is progressing is the subject of this paper. UNIVERSITY OF HARTFORD - "TUITION DRIVEN" To understand the forces behind the project to change administrative systems that working very well, the current economic circumstances of the University of Hartford must be related to the poor economic and demographic conditions of the recent past. Located on a 300 acre campus in the greater Hartford area, the University is an independent, comprehensive university. With a modest endowment portfolio and other few sources of significant funds, the University's economic dependency has been aptly described as "tuition driven." The impact of fewer students also decreases of revenues from other sources, such as residential housing. Most of the more than 3,800 full-time undergraduate students and 4,000 others come from the northeastern part of the United States, which has been economically depressed for a number of years. Furthermore, the University competes to some extent with the less expensive state institutions in its region of the country. The situation is further compounded by the general decline in the pool of appropriate high school graduates in the University's historic recruiting area. Understandably, there is increased competition for students, which has prompted additional resources to be committed to recruitment and to enhancing academic programs. As the economic situation has squeezed available resources, the internal competition for funding of valuable programs forced a rigorous review of doing more with fewer resources. DOING MORE WITH LESS - AN INSTITUTIONAL URGENCY When the impact of the declining number of students was felt by the University, an exhaustive review of operations was initiated that resulted in a series of difficult budgetary reductions. In the fall of 1991, budgetary reductions were made in many areas including personnel costs. While programs such as early retirements and 10 month staff positions reduced personnel costs, some layoffs also occurred. It was hoped that these actions would be sufficient. However, as the forecast for the next academic year became clear in the spring of 1992, further actions were required. Once again the budget cuts included personnel reductions and elimination of some projects designed to enhance the University's attractiveness to prospective students. DOING MORE WITH LESS - CS DEPARTMENT The Computer Services department (CS) was not spared the impact of the two general budgetary reductions and lost over 20% of its personnel. The strategic alternatives for administrative systems had been under continual review by CS within the framework of reducing costs while maintaining the current level of support for administrative information systems. A strategic plan outline had been created and, where possible, the reductions were made in groups that would need fewer personnel in the projected administrative system environment of the future. With the instructions to do more with less, CS began a new comprehensive review. Due to its on-going planning activity, the principal considerations and primary alternatives[l] for the future were well understood. The conclusions from the review can be summarized into three points, which are discussed in more detail in following paragraphs. 1. The current administrative information systems are very successful and integral to the administrative operation of the University. 2. But, these systems were designed a decade ago with the information technologies that were available at that time and there were major concerns about the future of the current applications. 3. Hence, the dramatic changes in information technology and the University's information technology infra-structure provide an opportunity to reduce costs while positioning the administrative systems for the future. [1] Vendor products and services are discussed only in the context of the University's environment, experience, and plans rather than any endorsement or criticism of any specific vendor product or service. ++ 1. Current Application Systems - Successful and Essential Over the past decade, the University of Hartford has invested in a comprehensive portfolio of integrated administrative information systems. Although the application software had been acquired from a major software vendor, Systems and Computer Technology (SCT), the software had been substantially modified and enhanced to satisfy the specific needs of the University. The applications systems are the alumni and donor development system (ADDS), the student accounts receivable (ARIS, which is integrated with ISIS), the human resource and payroll system (HRIS), the financial information system (IFIS), and the student information system (ISIS). The ISIS system is by far the largest of all the systems and incorporates most of the University's modifications and enhancements. The present portfolio of administrative information systems have been integrated into the daily operations of the University. A few quantifiable measures are displayed below to indicate the size of the administrative systems as well the support needed to maintain the current systems by the present staff of seven CS application professionals. Current Administrative Application Systems Item Number Source Code Modules 3,040 On-line Screens 477 Data Base Records 15.4 million Another important factor in the usage growth has been the expansion of the information technology infrastructure that now includes a campus- wide data communications network. The number of users with privilege to access the on-line production administrative systems has more than doubled during the past few years to over 500. ++ 2. Current Application Systems - Old Design; Future Path? The administrative applications were originally designed by SCT with the pervasive information technology of that time. While the University's applications had been modified, new enhancements made, and new modules added, the underlying technology is basically from the same era. The current status can be characterized as systems designed with older information technology using a proprietary platform. It is the cost of this proprietary platform that attracted the most scrutiny in the new review to reduce costs. Current Administrative Systems + Applications + Modified and enhanced vendor application software (SCT Symmetry series) written in COBOL. + Environment + IBM 4381 with MVS/XA and CICS, ClNCOM's RDBMS Supra 1.3.5 (non-SQL) The on-going strategic review process in CS had identified the long-term information technology objectives that would protect and enhance the University's investment in their administrative systems. It was also realized that the process would require intermediate steps that would allow the University to eventually achieve the longer term objectives in a judicious plan that also recognized the other competing demands that must be satisfied in the short term. Administrative Systems - Long Term Objectives + Achieve flexibility in selecting future computer platforms. + Improve ease of use and training of application end users. + Operate with a relational database management system that has SQL functionality. + Increase end user access to information, including English language query capability. + Position systems to exploit emerging and future information technologies (e.g., client/server, automated work flow, imaging systems). + Acquire and employ more productive application development and maintenance tools. In light of the long term objectives, the important aspects of the current situation were reviewed while considering the following constraints. Administrative Systems - Planning Constraints + CS staff is to remain few in number. + Any new system must provide the essential functionality that is currently provided. + Existing applications are to remain fully operational during all changes or transitions to new products. + Alternative computer platforms must be robust, mature and from a leading vendor. As the principal alternatives to the current environment were assessed in the context of the above objects and constraints, three general strategic directions were identified. Administrative System - Future Direction + Continue with the current applications and compatible platforms, + Move to other platforms with the same applications, and . + Move to different applications on a different platform. + Continue with the current applications and compatible platforms---As the usage demands continued to grow, staying with the first alternative meant that the current platform would need to go through an expensive upgrade and after the upgrade the same issues about the future would remain. The real opportunity would be to move to different, less expensive platforms with the same or different applications. + Move to different platforms with the same applications---One of the prime concerns with the existing application software was that the vendor (SCT) had developed a very successful new product line (BANNER) that had over shadowed the older product line (Symmetry). This created the predicament in which clients with the old product line decided to drop the enhancement and maintenance program because of an expectation of reduced support from the vendor and the vendor having to reducing support because of fewer clients subscribing to the enhancement and maintenance program. For the University, this trend meant that increasing resources (staff and computer systems) would be needed to achieve the same level of application system support in the future. In addition, the eventual phasing out of this old product line by the vendor was anticipated. Staying with the current applications but on a different, less expensive platform would not address these issues. Another information technology concern was the lack of SQL capabilities with the current RDBMS (Supra version 1.3.5). Since the logical upgrade from Supra 1.3.5 is Supra 2.X (which has SQL), the upgrade effort and requirements have been monitored and under consideration for some time. Based on our situation, our analysis determined that moving from Supra 1.3.5 to Supra 2.X would be much more than just a version change but a major undertaking for the University. It was concluded that just going to Supra 2.X would not yield sufficient benefits to justify the effort and that achieving SQL capability should be a component of a larger project rather than an isolated effort. + Move to a different platform with different applications---The alternative of moving to other platforms with different applications raised the issue of outsourcing application development. Considering the extent and use of the current applications as well as the above constraints that included keeping the professional staff small in number, it was concluded that any future administrative systems would be appropriate vendor software with the necessary modifications and enhancements. The use of vendor application software is, in effect, "outsourcing" a large part of the development effort, while continuing the enhancement and maintenance program with the vendor is an "outsourcing" of some of the on-going systems analyst and programming work. Another consideration for vendor application software was that it must be successful in the higher educational market place and that, preferably, one vendor could supply all applications. The latter consideration would provide a consistency across systems for users and allow for more efficient support of systems with a small CS staff. The leading candidate was the new, successful application series (BANNER) from the current software vendor, SCT. The obvious hardware alternatives for the University were UNIX systems and/or DEC VAX platforms. DEC was a major consideration because of DEC's strategic direction toward openness, the University's experience with DEC platforms, and DEC's program of the free licensing of its software products, including systems software, to higher educational institutions. Since the University had a DEC VAX system for academic use since the late 1970s, it had taken advantage of this program as it upgraded its academic computer system, installed new computer equipment for a library system, and installed its campus data network. Thus, the University has a high level of experience and expertise with DEC equipment and software with a very small staff. The DEC platform is robust and mature and DEC is a leading computer vendor. Given the University's circumstances, the DEC platform emerged as the prime candidate for a move to a different platform for administrative applications. ++ 3. Opportunity for Doing More With Less While Achieving Strategic Objectives As indicated above, the implications for the major information technology alternatives for University were well understood. With the directive of "Doing More With Less," starting on the path to the future became an immediate opportunity. Plans were developed to move to the new systems both quickly and prudently while maintaining the old systems. Administrative Systems - Future Production Status + Applications + Modified vendor application software (SCT BANNER series) + Increased ease of use and user access to information-report writers, English language query product + Environment + DEC 6620 with VMS; Oracle RDBMS 6.0; Oracle SQL*Forms & SQL*Menu The feasibility of the project was carefully examined as to achieving the objectives within a reasonable time frame, since the immediate reason for the project was not achieving information technology objectives but reducing costs as soon as possible. Accessing the Feasibility of and Creating a Project Plan - Major Factors + Senior staff support. - End of changes to current applications except for regulatory or University requirements. - Commitment of user departments and the time of user team members. + Expertise of the applications users and CS development staff. - Accelerated training and consulting. - Use of temporary, contract personnel. + Same vendor for both application software series. - Logical progression of core functional design from the old to the new applications. - Fast "maturing" of functionality of the new applications versus the mature functionality in the University's current production systems. - Consultants with knowledge of old and new applications. + "Freezing" of systems on the old computer platform. - No changes on the operating system, equipment, etc. on the IBM platform. - Accomplish transition to the new systems before an upgrade of the old platform is needed. + Replacement of equipment to allow access to the new system - Network infrastructure in place except for near-by buildings. The feasibility of the project depended upon a "window of opportunity" in which the transition to the new systems could be completed before an upgrade to the old computer platform would be needed. The systems analysts and key users reviewed the new and the old systems and created an outline plan and production dates for each of the major application products. Other institutions were contacted and visited to discuss their experience with the BANNER products and their conversion projects. For each of the major application products, a general outline of the categories of plan activities was followed by key users and CS systems analysts. Conversion Plan - Task Categories + Training/Consulting - Functional and Technical - Accelerated for those experienced in current applications - Audience: Implementation team members + Data conversion - Data element analysis and conversion programs + Validation tables and rules creation for new applications + Modifications to baseline - Changes to data base, forms, and programs - New forms and programs + Interfaces - During transition (between applications in the old and new systems) - After transition (between the applications in the new system) + Departmental Training - Creation of training for the production system - Training to occur just before the move to production + User Access to Information - Selection, installation, and training - report writers, text processors, and other query tools - Selection, installation, and training - English language query products + Internal CS Management, Control, and Procedures - Change management, scheduling software, backup/recovery, security, etc. - Productivity tools and products for systems analysts and programmers The result of this analysis was a plan with the target production dates for the various products that took into consideration the work needed to be done in the Task Categories and the user departments estimate of the best time of year for them to change over to the new system. Since the English language query facility would not be useful until sufficient data was converted to one of the new application products, selection of this product was scheduled to begin in the first quarter of 1993. Conversion Project - Dates of Major Events DEC 6610 system June 1992 (Completed) Oracle products, Baseline applications July 1992 (Completed) BANNER Finance December 1993 BANNER Human Resources January 1994 BANNER Alumni/Development May 1994 BANNER Student, Financial Aid August 1994 Necessarily, the review was based on assumptions about converting to a "moving target," since the very successful BANNER series was maturing with a series of updates from SCT. The actual release of the BANNER products that would be used for production would be decided after the project began and would be based on a more detailed implementation plan and the vendor's release schedule. Considering the time frame and cost for the implementation of the project, the financial analysis resulted in an estimated savings of over $400,000 per year. Naturally, it was based on a number of factors subject to change, but there was confidence that this figure will be attained. REVIEWS AND APPROVALS With a project this size, it not unexpected that there was an extensive review and approval route. Revisions and improvements were made at each step along the way. Starting with the internal reviews with CS personnel and key users, the process was to involve the entire user community at some point in the process. As an independent check, external consultants went over the plans and made recommendations. Senior staff reviewed and finally approved the project for presentation to the Board of Regents, which made the final approval. Major Reviews and Approvals + Internal CS reviews with key users and contacts with other BANNER clients + General presentations and review by the user community + Senior Staff reviews + External consultant review + Senior Staff approvals + Approval by the Board of Regents IMPLEMENTATION The approval of the project was given by the Board of Regents in May 1992 and the official start of the project began with the new fiscal year on July 1. The training of the Computer Services staff began at the end of May when the data base administrator (DBA) started DBA Oracle training sessions in order to be ready for the installation of Oracle and BANNER products in early July. Computer Services personnel initially trained in DEC VMS and Oracle RDBMS and SQL, using computer based training (CBT) or self-paced instruction, and Oracle developmental products. The project organization was designed to be user-driven with a predominance of users on all of the project teams. There are "link pins" (common members) between teams to share information and project news. BANNER Project Management Organization Project Advisor Board (7 members) Communication Team (3 members) Implementation teams (Average 19 members): --Project Team --BANNER Products Alumni Development Team Alumni/Development Finance & Human Resource Team Finance, Human Resource Student 8 Financial Aid Team Student, Financial Aid The Project Advisory Board is composed of senior administrators and functions as the University's overseer for the project and resolves cross BANNER product issues as necessary. The Communication Team is to inform the University community of the progress of the project. The implementation project teams were organized to include key user, application end users from various departments, and CS system analysts. Since the software products are composed of a number of modules designed for particular departmental functions, a variety of representative end users were included in each team to ensure an institution-wide involvement in the implementation. The teams are each led by an experienced user and were grouped by related BANNER products. It was expected that sub-teams will be formed to handle specific products as appropriate. Training is a very important part of the project. The earliest training was to prepare the information systems professionals for the new technology. The next block of "training" was to both train and provide consulting to the members of the project teams as they began the project tasks, which includes the creation of a detailed project plan (e.g., data conversion, modifications, new programs, and training,). Soon after the installation of the baseline application systems on the University's new DEC system, the training by the application vendor consultants began. Training was divided into a series of sessions by application and type (which is either functional training for users or technical training for information system professionals). Over 25 weeks of on-site training/consulting was scheduled. The production training of the departmental end users will occur just before the move to production and will be based on the actual production system with the University's validation tables, rules, changes, enhancements, etc. EVALUATION OF THE CURRENT STATUS As of this writing, there have not been any surprises that would imperil its success or hasten its completion. However, there have been some notable events and experience worth noting. On-Site Functional Training/Consulting Sessions The extent of on-site functional training was expanded beyond what was originally planned. Scheduling of the larger implementation teams was a major problem, even though most team members attended training only for those application modules that directly concern their departments. Changes in the implementation dates also prompted an accelerated schedule for some systems. However, the faster-paced on-site training with correlations to the old applications is considered the better alternative for the team members than the lengthier standard training in the baseline product. Most of the trainers/consultants from SCT were knowledgeable in both the old Symmetry and the new BANNER systems. Drawing upon this experience is a major benefit to the project. Unexpected Loss of Personnel Key University personnel were lost for a variety of reasons including unexpected early retirements and new jobs at other institutions. The result has been a delay in some areas, which is expected to be made up over the life of the project. However, loss of personnel who have experience with the old systems is a definite problem. There has been another temporary loss of a key vendor consultant due to a major medical problem. This misfortune is not expected to significantly delay progress of the project. Changes in Project Organization Two changes occurred for different reasons. The first was a general perception from the implementation team members that a new committee should be formed to address the common data issues. This new Common Data committee has the goal of investigating issues surrounding the common data and validation tables in the new BANNER products and making recommendations to the Project Advisory Board. As with all other implementation teams, the leader is a knowledgeable user and the members come from other teams to ensure dynamic interaction and sharing of information among the teams. One of the hardest issues is determining a procedure of updating common data in a controlled, rational approach that is acceptable to all involved departments. The other organizational change was the splitting of the Finance and Human Resources team into two teams with a new leader for the Human Resources team. This change was prompted by the loss of a senior person in the financial affairs area and the additional workload that the leader of the old combined team had to assume. BANNER Project Management Organization (Revised) Project Advisor Board Communication Team Common Data Team Alumni Development Team (BANNER Alumni/Development) Finance Team (BANNER Finance) Human Resource Team (BANNER Human Resource) Student & Financial Aid Team (BANNER Student, Financial Aid) Implementation Dates - Changes The most significant change in the implementation dates was for Finance, which is now scheduled five months earlier to coincide with the beginning of the new fiscal year. This earlier implementation date will result in not having to convert certain detail financial records of the current fiscal year to the new system. Plan for Production Status by BANNER Product (Revised) Finance July 1993 Alumni/Development February 1994 Human Resources March 1994 Student 8 Financial Aid August 1994 Unexpected Demands for Recruitment. etc. Recruitment and retention activities and other support demands have been much higher than expected and has, consequently, drawn both user and CS resources away from the project. The impact of these activities has been mainly felt in the Student and Financial Aid project. Production - BANNER Version As discussed above, SCT has constantly been enhancing and expanding the application products. At some point, a BANNER release had to be picked that would be the basis for the production systems at the University. SCT has announced that the first quarter 1993 release will be a "very clean" release that will resolve all problems as of October 1992. This release will also be the basis of the next major technical upgrade of BANNER (named BANNER2). It is now planned that this is the release that will be the basis of the new production systems. CBT Training for CS Personnel In general, the approach was successful because of its flexibility to each person's schedule. However for those assigned to applications that demanded critical, additional support, the progress was slower than planned. As is well known, off-site training usually provides better conditions to concentrate on the learning the material. SUMMARY - WHY CHANGE? The question of "Why Change" can be briefly stated. The circumstances of the University and the alternates available to it created a "window of opportunity" to reduce costs while positioning the administrative applications for the future. While the existing systems could have been continued for a number of years, it would have been increasingly expensive and would have left the University with older systems on a proprietary platform. Furthermore, these older systems would continue to face an uncertain future. After considering the alternatives and issues, the prudent decision was made to change applications that work before the window of opportunity closed.