Forecasting is defined by B.R. Martin as "the process involved in systematically attempting to look into the longer-term future of science, technology, the economy and society with the aim of identifying the areas of strategic research and emerging generic technologies likely to yield the greatest economic and social benefits."

Numerous techniques for forecasting technological developments were pioneered in the 1960s in both business and government (particularly military) applications, and the term "foresight studies" is now commonly used. The more important of the techniques are described here.


The growth of a new technological capability typically follows an S-shaped curve that can be divided into three stages. The first is slow initial growth, as the new technology has to prove its superiority over existing technologies. Once this is demonstrated, a period of rapid growth follows. Finally, its growth is limited by technological or socioeconomic factors and levels off toward some upper limit. The commercially successful exploitation of technology often depends upon the astute perception and exploitation of this growth. Thus, forecasters pay significant attention to extrapolating the growth of the S-shaped curve of a technological capability at some relatively early stage of its life. In so doing, they use mathematical functions or models.


Technological evolutions typically progress through successive generations of capabilities (e.g. 286, 386, 486, and Pentium microprocessors) and, as each capability is superseded by its technologically superior successor, overall functional performance continues to rise along an envelope curve generated by successive S-curves. This envelope curve defines a trend against time, which may be extrapolated forward to predict future capabilities.

Richard Foster focuses general management attention upon the importance of identifying S-curves while, for the mathematically sophisticated reader, Meade and Islam provide a critique of the relative merits of some of the numerous technological techniques available.


The Delphi method and its extensions provide the backbone of foresight studies. This method was originally funded by the U.S. Air Force and later developed by Olaf Helmer and coworkers at the Rand Corporation. It derives its name from the Oracle of Delphi, who was the prime source of prophecy in ancient Greece. The method is based upon the premise that the best sources of technological forecasts are the opinions of experts in the given technology. That is, the simplest way of making a forecast is to ask the experts in the field to do it. It is undesirable to base a forecast on a single oracle or expert, however distinguished, so the opinion of a sample or committee of experts is sought. The considered judgment or consensus of a committee of experts provides a viable approach to deriving a technological forecast, but suffers from the disadvantage that it may be biased toward the opinions of its dominant members. The Delphi approach avoids this disadvantage by requiring members to participate anonymously.

The Delphi method is usually conducted by one individual, known as the director. The panel's members are selected based upon expertise and availability, security considerations (e.g., commercial or military) and the avoidance of overall bias. Panel members can usually be selected from peer judgments, literature citations, honors and awards, patents, and professional society status. A typical panel consists of between ten and fifty members.

The approach is iterative, with each iteration called a round. In each round the members are interrogated individually and confidentially (usually by questionnaire) for their views on the likelihood and timing of the occurrences of certain future, technological breakthroughs or other events. Direct interactions among panel members is forbidden; this preserves anonymity between panel members, with controlled anonymous feedback. A unique feature of the Delphi method, as noted by Parenté et al, is that it provides feedback from earlier rounds between successive polls. The results of each round are summarized statistically as median-date and interquartile-range responses and circulated among panel members. In the first round, members often differ widely in their judgments, yielding a wide interquartile range. However, as members anonymously exchange the rationales of their judgments in successive rounds a consensus is reached, usually rather quickly—after about four rounds. The director then consolidates the results of this final round, which constitute the reported forecast. The panel's forecasts are usually presented in the form of the final median dates and interquartile ranges for each of the events considered.

These techniques—the S-shaped logistic curve, envelope curves and trend extrapolation, and the Delphi method—are a primary sampling of the tools used for technological forecasting. Advances in computer technology will continue to provide additional forecasting opportunities for years to come.

SEE ALSO: Futuring ; Longitudinal Scenarios ; Multiple-Criteria Decision Making

Michael J.C. Martin

Revised by Monica C. Turner


Alsan, A., and M.A. Oner. "Comparison of National Foresight Studies by Integrated Foresight Management Model." Futures 36, no. 8 (October 2004): 889–902.

Foster, R. N. Innovation: The Attacker's Advantage. New York: Summit Books, 1986.

Grupp, H., and H.A. Linstone. "National Technology Foresight Activities Around the Globe: Resurrection and New Paradigms." Technological Forecasting & Social Change 60, no. 1 (2 January 1999): 85–94.

Kuwahara, T. "Technology Forecasting Activities in Japan." Technological Forecasting & Social Change 60, no. 1 (2 January 1999): 5–14.

Martin, B.R. "Foresight in Science and Technology." Technology Analysis & Strategic Management 7, no. 2 (June 1995): 139–68.

Meade, N., and T. Islam. "Technological Forecasting: Model Selection, Model Stability, and Combining Models." Management Science 44, no. 8 (August 1998): 1115–130.

Parenté, R.J., T.F. Hiöb, R.A. Silver, C. Jenkins, M.P. Poe, and R.J. Mullins. "The Delphi Method, Impeachment and Terrorism: Accuracies of Short-range Forecasts for Volatile World Events." Technological Forecasting & Social Change 72, no. 4 (May 2005): 401–11.

Porter, A.L., W.B. Ashton, G. Clar, J.F. Coates, K. Cuhls, S.W. Cunningham, et al. "Technology Futures Analysis: Toward Integration of the Field and New Methods." Technological Forecasting & Social Change 71, no. 3 (March 2004): 287–303.

Rowe, G., G. Wright, and A. McColl. "Judgment Change During Delphi-Like Procedures: The Role of Majority Influence, Expertise, and Confidence." Technological Forecasting & Social Change 72, no. 4 (May 2005): 377–99.

Salo, A., T. Gustafsson, and R. Ramanathan. "Multicriteria Methods for Technology Foresight." Journal of Forecasting 22, no. 2/3 (March/April 2003): 235–55.

User Contributions:

Comment about this article, ask questions, or add new information about this topic: