When the stakes are high and uncertainties are great, it pays to build a model of your idea and test it in any and every way that you can. By modeling the a process, system, or program before it is designed and implemented, you can more clearly think through how it will impact overall organizational processes and performance. When the idea "flies on paper" in the modeling stage you can be more confident that it will success in real operation. That is why building models and testing them thoroughly before getting to the final design and implementation phase is an effective way to hold down development costs and minimize risks.

Extensions of a formal model of a problem. Review the earlier section of this handbook on "Models of Problems," because solution models share many features with problem models.

Representations of operation. Like the problem model, solution models often represent processes, information flows, decision points, and relationships -- but this time they are improved to solve the problem. The solution model should show how the new process or system will function within the whole organizational context.

Representations of organizational and customer-oriented effects. These representations are just as critical as the ones that show how the proposed IT system itself will function.

Simulating full system operation. Models of solutions help you describe and simulate how a full system will operate within the context of organizational and human factors.

Thinking big. These types of models help you see the implications of a limited prototype when it is expanded to full-scale operations. Managers are forced to think through technical, organizational, and policy issues in designing these models.

Exploring costs and benefits. You can delve into the costs and benefits of proposed solutions by linking the model to financial data.

Asking "what if" questions. A new or revised system or process will have various organizational and human factor effects. By asking "what if" questions, you can anticipate issues and problems before they are encountered in a real world system implementation.

All depends on the data. Models of solutions are no better than the data and relationships upon which they are built. Your model must be built upon a foundation solid data and analysis if you want it to accurately forecast the impacts of the new system.

Expense. These models can be very complex, expensive and time-consuming to build. Models of solutions may require specialized expertise, which may be unavailable in your agency.

Wolstenholme, E. (1993). Evaluation of Management Information Systems. Chichester; New York: Wiley.