Document Type

Closed Project

Publication Date

Summer 2002


Toryos Pandejpong

Course Title

Research Methods in ETM

Course Number

EMGT 565/665


Diffusion of innovations -- Mathematical models, Technological innovations -- Management, Engineering management, Decision making


Historically, technology transfer within the software industry has been a difficult, slow, and poorly understood process. Redwine and Riddle found that a mean time of 17 years was required for software engineering technologies to pass from the initial concept stage until to usage by 70% of the industry. In a survey of software process innovation (SPI) champions, Goldenson and Herbsleb discovered that lack of guidance, mentoring, and assistance for change agents was a major factor in retarding progress:

Our data suggest a number of factors that can make process innovation difficult to achieve. Aspects of organizational culture are among those most likely to inhibit such change. When our respondents say that they have seen excessive turf guarding and organizational politics, they also report less success in addressing the findings and recommendations that were raised in their appraisals ... We need to learn more about how to make change happen, not just what needs to be improved.

The social barriers facing SPI are especially formidable when infusing software innovations into large organizations. Technology infusion refers to the micro-level organizational process of adopting a new process technology. The macro-level industry counterpart is technology transfer. In large organizations recipients may be widely scattered, inhibiting interpersonal contact between SPI change agents and software engineers.

Internal SPI agencies face formidable project planning difficulties. They lack the tools needed to accurately estimate required resources and campaign duration. It is not well understood how to forecast the number of internal change agents who would be needed to infuse a software process innovation. Humphrey states that full-time staffing levels for internal change agencies in software organizations should run to about l-3% of overall staffing levels, but this is only a rough estimate.

Change agencies also the ability need to assess the impact and effectiveness of communication strategies in support of infusion campaigns. Formative evaluation techniques can help to refine the message, but the ability to experiment is limited by the modest resources of most change agencies. As staff :functions their budgets tend to be limited in comparison to line :functions, so it is important for them to leverage their resources as efficiently as possible.

In short, technology infusion involves a good deal of guesswork, which contributes to the lengthy periods of time reported by Redwine and Riddle. As part of an effort to address the problem this paper poses the following longterm management questions:

  1. How can we improve estimates of the cost, effort and duration of technology infusion projects?
  2. How can we identify points of maximum leverage for ehange agency resources?
  3. How can we evaluate the effectiveness of change agency communication strategies?
  4. How can we identify points of resistance in the organizational culture and test the effectiveness of risk migration strategies?


This project is only available to students, staff, and faculty of Portland State University

Persistent Identifier