The definition of prototype first needs to be put into the context of a domain-of-discourse, which for this article is PD. This is necessary because, undoubtedly, there are as many definitions for 'prototype' as there are areas of human endeavor.

PD is a business process that covers the definition, design, implementation, validation, production and service of a product that is to be sold into a market at a profit. Prototype in this context is defined as a goal-driven, recursive model (or tool) which represents some (to all) aspects of the product, and is used to learn about the product's functionality, performance, interaction with its environment, its economics, legalities, intellectual property (IP) value and its usefulness.

A goal-driven prototype is very necessary in the PD business because there are almost always definite Resource, Funding and Time (R) limits by which a project is constrained. The prototype must have goals against which you gauge the prototype's usefulness and success. The goals need to be measurable (Go / No-Go, statistical, etc.) or assessable (by agreement) decisions. If they can not be, then the prototype process can fall into a state of indecision; and in a PD environment this could pose a problem for the business case.

A prototype's usefulness is finished when the goals have been achieved. The usefulness must be relevant to the prototype stakeholders - those people who are involved in the creation, development, assessment, production, support, regulation and use of the product. The responsibility of each stakeholder is to provide discussion and feedback during interaction with the prototype.

The goals must be absolutely measurable so they can be gauged by what the product's evolvement and the stakeholders require, not by what the vague egos of the engineering manager, or the marketing guru or the CEO demand.