Institutional Barriers
Generally it is recognized by those of us who have been involved in the GIS "profession" for a long period of time that management acceptance is essential for a successful GIS implementation. The Book supports that notion and in fact, provides a theoretical foundation for the successful implementation of GIS from an institutional level. In the preface of the Book the authors suggest that the capabilities of GIS have exceeded the needs of most end users. Furthermore, the discrepancy between users' needs and GIS capabilities is largely due to institutional limitations including standard operating procedures (SOPs). Generally, SOPs within large organizations tend to propagate the status quo rather than promote change or innovation.
      So what are the requirements for implementing a successful GIS? The greatest impediment to introducing any new technology or enhanced information system capabilities such as GIS has been the difficulty in gaining widespread acceptance and effective utilization of the new system. It is primarily an organizational rather than a technological barrier that needs to be overcome.
      I once worked for an organization that attempted to implement a GIS to manage its timber inventory. At first the system was not very successful because (at least in my opinion) the management tried to make the new software operate the way they were accustomed to doing things. After several iterations of hardware, software, and personnel changes they began to develop new procedures based on GIS capabilities. Ultimately the system paid for itself and it is now used on a regular basis. Their GIS is used to keep track of timber stands and detailed information about the trees such as average diameter at breast height (DBH), crown density, species mix, silvicultural practices, and other vital data that are used to make cost-effective management decisions.

Organizational factors
There are many organizational factors that influence the adoption and eventual success or failure of a GIS. The authors of the Book state that some of those factors may include:
1. Clearly defined goals - this includes the overall mission of the organization, but more specifically, what are the functional requirements for the new system.
2. Sufficient resource allocation - money, time, training, and personnel all need to be allocated in adequate quantities to support not only the installation, but also the ongoing maintenance and support of the system.
3. Upper management support - sometimes this happens because of a "champion," or someone with enough influence and decision-making authority to see that the implementation is a success. If there is no "champion," at least the support of top management needs to be made known to all concerned parties.
4. Implementation schedules - this relates to the planning aspect. The implementation should be well thought out, preferably with detailed schedules and milestones for the use of the system and training for users.
5. Competent technical support - this may include external support from the software vendor, but internal technical competence is also an essential factor.
6. Adequate communication channels - The goals, status, schedule, users' needs, training requirements, etc. need to be adequately communicated with all concerned parties.
7. Feedback capabilities - there should be a formal mechanism in place for suggestions and changes to the system.
8. Client responsiveness - if the system is not capable of responding to the needs of the end users, it clearly will not be successful.

Implementation types
Many different types of GIS implementations exist making it awkward to develop a cookbook approach to assist those involved in doing the work. Often times a particular GIS implementation is so site-specific that the same steps would not be appropriate for other organizations. I believe that for this reason primarily, there is a lack of good written material in the form of guidelines or how-to documents regarding the steps required to implement a successful GIS.
      There are many possible ways to categorize the potential types of different GIS implementations that may exist. For simplicity sake, let us assume that a GIS can represent either: 1) a large, transaction-oriented system such as that used by a large public utility or municipal public works department; 2) a medium-sized, structured system that supports data sharing and broad access to support organizations like a city or county government agency, oil company, or data conversion shop; or 3) a small, flexible system that may be supported by only one or two staff members within a broader organizational context such as an engineering consulting firm, a small design group, or a small city or county with a limited budget.
      There will obviously be differences in the procedures used to establish a GIS depending on the type of implementation. For example, a large, category 1 type of system requires extensive planning with a long lead time and a solid implementation plan. A medium, category 2 implementation may include some detailed planning, but might forego some of the more extensive design considerations by using a pilot study to determine the detailed requirements. The small, category 3 types of systems can usually be changed or enhanced without a lot of lost time or wasted effort and thus the need for a lot of pre-implementation planning is greatly reduced.

Introducing change
For most types of implementations, a needs analysis should be conducted to determine the functional system requirements, types and quantity of data required, recommendations for hardware, software, staffing, and procedures for updating and maintaining the database, with schedules, and perhaps even a cost-benefit analysis. The needs analysis may be a formal process conducted by outside consultants, or it may be an internal review completed by staff members who are knowledgeable about GIS hardware, software, data, and staffing issues.
      A detailed database design is also a crucial step in most types of implementations. The database design should build upon the needs analysis to further define the system requirements in terms of amount, accuracy, and specific attributes of the data that will be stored and analyzed using the new system. Ideally, a pilot project should be conducted using the initial database design and a small but representative study area to uncover any additional quirks or concerns that could affect the usefulness of the system. The database design should also attempt to anticipate future requirements, potential analyses, and the specific types of data required to satisfy those needs.
      Finally, an implementation plan should be developed that includes goals and schedules for procuring and installing hardware and software, for providing user training, that outlines staffing requirements and responsibilities, and that provides a starting point for developing new procedures for using the GIS. As long as upper management embraces the plan and provides the support needed to maintain staffing, equipment, and other necessary resources, and assuming that the plan is a good one, a successful GIS is certain to evolve.

About the Author:
Damon Judd is manager of GIS and computer applications for RUST Environment and Infrastructure in Broomfield, Colo. He can be reached at 303-469-6660.

Back