Improving the accuracy of expert predictions of the future success of new internet services

One of the challenging issues in new product development is the evaluation of the future success of new product initiatives and the elimination of those with unpromising futures. The difficulty of assessing the future success of a new offering is more highlighted for new services than is for new products because of the unique characteristics of services. Earlier studies suggested that a number of factors can affect the quality of evaluating the future success of new products or services. Building on previous studies, this paper offers the first empirical tests of the roles of two important factors (task structuring and expertise sharing) in improving the quality of expert predictions of the future success of new Internet services. It presents an empirical study based on a longitudinal survey that involves both the initial predictions and the actual results for two Internet services to test related hypotheses. The results showed that these factors can significantly improve the predictive accuracy of experts in evaluating new Internet services. More importantly, the results also indicated a significant substitution effect between the factors, suggesting that either of these factors is sufficient to make major improvements in the accuracy of the evaluations. Because the factors can largely substitute for each other; however, using both can lead to only minor improvements over and above using just one. © 2006 Elsevier B.V. All rights reserved.