After a project has been selected and the planning has begun, there are two important steps that must be taken by a system's analyst before design can begin. The analyst(s) must elicit requirements from the stakeholders and then structure those requirements in a way that can be used for the design specifications and for further follow-up and analysis in future iterations. Requirements gathering is widely considered to be one of the most important steps in ensuring project success and catching problems early on before they become too costly or time consuming (I will not loom on this topic, as it is widely discussed elsewhere, such as (Requirements Elicitation) (Isserman), and (Masters)).
With a process that is so important to companies and the success of projects, any improvement or gain that can be made in the way analysts elicit requirements (and the resulting increase is quality or number of specifications for the designers) is highly desired. There are a very wide range of ways that researchers have gone about uncovering problems in the analysis process and methodologies to improve them. In this paper, I am going to focus upon the issue of unconscious knowledge, and how we can improve our requirements elicitation given the existence of unconscious knowledge and processes that people have but cannot explicitly identify themselves. The question is this: If people have a range of knowledge and responses that are important for the requirements we need, but they don't necessarily know about them nor can describe them, how can we improve our own methods to recognize and identify these requirements? I will attempt to address that question in this paper, based upon research that currently exists and is available.