Leapfrogging Behavior and the Purchase of Industrial Innovations

Allen M. Weiss and George John, 1989, 89-110

How anticipation of future product improvements affects purchase decisions.

Type of Report
Presentation of theoretical model; empirical test with data from industrial buying.

To understand and explain leapfrog behavior, the practice of deferring purchase of currently available technology in anticipation of subsequent improvements.

Industrial purchasers, particularly those in high-technology areas, will find the model of buying behavior intriguing, academics will want to consider the techniques employed and their possible extensions.

Main Points
In rapidly changing markets, firms often face innovations that represent the latest link in a chain of improvements. Since improvements are expected to continue into the future, some potential buyers choose to bypass the current innovation, since a technology or product acquired now may soon be rendered obsolete by an even more advanced offering. This is leapfrogging behavior--the act of considering the adoption of an innovation, and then deciding to bypass it in favor of waiting to consider an improved future product.

Leapfrogging is thought to account for the sales fall-off for state-of-the-art products when improved versions of the same technology are anticipated by the market. Likewise, leapfrogging is thought to explain why preannouncements of new items forestall purchases of current products. Despite the intuitively apparent importance of leapfrogging behavior, it has not been investigated systematically.

The current report develops a formal model of the firm's decision process that incorporates the possibility of leapfrogging. It proposes that at each point in time, a firm must decide whether to (1) adopt the current innovation, (2) gather more information about it, or (3) stop the search process and wait to reconsider things when the improved technology is launched. The firm's decision is affected by variables which include its perceptions of (1) the benefit gap between its incumbent technology and the current innovation, (2) the cost of adopting the new technology, (3) the size and tuning of the improvements expected in the future, (4) the uncertainties associated with these expectations, and (5) the cost of searching for information. Using an "optimal stopping problem" methodology, the authors solve the model to explain firm behavior.

Fundamentally, the model shows that firms will decide to adopt the current innovation if its extant technological gap (defined as the extent of the advantage of the innovation over the incumbent product) is larger than a certain critical level. If this gap is smaller than another certain critical threshold, the firm will leapfrog past the current innovation. When the gap lies between the two thresholds, the firm will continue to search for more information. While this is intuitively plausible, and is consistent with other recent decision models of innovation adoption, there are some complex dynamic effects involved.

For instance, these thresholds are not constant over time. As time passes, the firm's upper critical gap initially lessens. In other words, as time passes, progressively lower benefit thresholds may induce adoption. Perhaps information becomes less valuable because the amount of uncertainty remaining about the innovation has been reduced. Hence the criterion threshold for adoption becomes less stringent.

However, as the launch date for the improved version draws near, this threshold rises again. In other words, the threshold level of benefits needed to induce adoption of the current innovation becomes larger, since the option of leapfrogging to the future version looks increasingly attractive.

There are other dynamic effects for both the size of improved benefits and the timing of the expected future innovation. The model shows that, when firms expect relatively large improvements over the current innovation, they tend to put off purchase by raising the critical threshold needed to induce adoption. Simple intuition says that firms are more reluctant to commit to the current offering because the option of bypassing it in order to consider the (much better) future product is more attractive. Furthermore, when the current innovation has just been launched, and when expected future products are still far away, firms that expect the future product to be launched relatively soon are more wiring to commit to adopting the current innovation. As time goes on, however, and the expected future product looms closer, these firms are less willing to commit to adoption.

Uncertainty over the launch date of the future product also has some complex dynamic effects. Early in the life cycle of the current innovation, firms that are uncertain about the launch date of the improved product are more willing to commit to adoption of the current innovation, since the value of leapfrogging to the relatively uncertain future has reduced value. However, late in the life cycle, when the future product looms close, uncertainty about its launch date forestalls adoption of the current innovation.

Empirical Validation
The empirical part of the work provides evidence about some of the model's predictions. A survey was done of firms in the market for surface-mount-technology (SMT) equipment, a replacement technique for assembling printed circuit boards. Traditional technology still accounts for about 90 percent of the printed circuit market. Industry observers claim that the adoption of SMT has been stalled because of expectations of improved SMT equipment.

These circumstances permit use of the model developed here. In terms of the model's representation, firms can either adopt the currently available SMT equipment, continue searching for more information about SMT, or bypass the current SMT equipment in favor of waiting to reconsider things when improved versions are launched.

Using key informants, the authors obtained measures of the constructs indicated by the model. Among the prominent independent variables captured were the pace of technological improvement in SMT, the uncertainty about the timing of future improvements in SMT, and the costs of adoption and searching.

Using multinomial logit models to account for the qualitative dependent variables, the data show support for the model's predictions. Firms that expect faster and more sizable improvements in SMT tend to be more willing to pass up the current SMT equipment. However, uncertainty about these improvements had no effect on such behavior. Consistent with predictions, the authors also found that costs of switching to the current SMT equipment tended to make firms more likely to leapfrog. Likewise, firms that were less capable of searching and processing information were more likely to leapfrog.

Only two predictions about adoption behavior could be tested with the data: the data show that firms with higher switching costs are less likely to adopt SMT. However, the ability to search for information has no significant effect.

The most significant implication of this work is the identification of the leapfrogging phenomenon. Once expectations of the future are taken into account, firms' reactions to current innovations are seen to be relatively complex. Rather than simply comparing the costs and benefits of the current innovation against the incumbent product, firms may choose to bypass the current innovation because of the possibility of being saddled with a soon-to-be-antiquated product. Clearly, the timing and volume of innovation purchases are thus affected, and the diffusion curves that are implied are probably quite different as a consequence.

Professor Weiss notes:

"It was the casual observation that technological advance is occurring at an increasingly rapid pace that motivated this research. Both the popular press and trade journals are constantly reporting significant technological improvements in a variety of product markets. Indeed, many industrial buyers who depend on technological capabilities for their differential advantage have become keenly aware of these ongoing changes and must decide when to adopt an improving technology. Oddly enough, the marketing literature has offered little assistance in understanding this purchase environment. Other research that addresses adoption in high technology environments has typically neglected the fact that expectations of technological improvements will naturally arise and can play a significant role in the behavior of potential customers. We hoped to draw attention to this phenomenon through this research. In particular, we wanted to understand, in a precise manner, the nature of leapfrogging behavior. We hope that this will lead to an enhanced understanding of customer behavior in high-tech markets. Naturally, there is much left to do in this area and, since the rate of technological advance is not likely to abate, the relevance of the research should continue into the future.

About the Authors
Allen Weiss is Assistant Professor of Marketing at Stanford University. George John is Associate Professor at the University of Minnesota.

This report is not currently available online. To purchase a print copy of the full report, please contact


  • Corporate: FREE
  • Academic: FREE
  • Subscribers: FREE
  • Public: $18.00



Employees of MSI Member Companies enjoy the benefits of complete online access to content, member conferences and networking with the MSI community.



Qualified academics benefit from a relationship with MSI through access to, conferences and research opportunities.



The public is invited to enjoy partial access to content, a free e-newsletter, selected reports and more.




Learn more about becoming a member of the institute

Read More

Stay Informed

The MSI Mailing List

Subscribe to our email list to stay informed about upcoming events, news, etc.

Social Media