Predicting the Path of Technological Innovation: SAW Versus Moore, Bass, Gompertz, and Kryder
Ashish Sood, Gareth M. James, Gerard J. Tellis, and Ji Zhu*
Marketing Science
Abstract
Competition is intense among rival technologies and success depends on predicting their future trajectory of performance. To resolve this challenge, managers often follow popular heuristics, generalizations, or “laws” like the Moore’s Law. We propose a model, Step And Wait (SAW), for predicting the path of technological innovation and compare its performance against eight models for 25 technologies and 804 technologies-years across six markets. The estimates of the model provide four important results. First, Moore’s Law and Kryder’s law do not generalize across markets; none holds for all technologies even in a single market. Second, SAW produces superior predictions over traditional methods, such as the Bass model or Gompertz law, and can form predictions for a completely new technology, by incorporating information from other categories on time varying covariates. Third, analysis of the model parameters suggests that: i) recent technologies improve at a faster rate than old technologies; ii) as the number of competitors increases, performance improves in smaller steps and longer waits; iii) later entrants and technologies that have a number of prior steps tend to have smaller steps and shorter waits; but iv) technologies with long average wait time continue to have large steps. Fourth, technologies cluster in their performance by market.
The law and the profits
Technology forecasting: A new “step and wait” model claims to outperform industry rules of thumb in predictive power
Mar 9th 2013 |From the print edition
PREDICTING the course of technological progress can be a risky business. Scorn the latest advances and you risk being left behind, as when Sony kept investing in flat-screen versions of cathode-ray televisions in the 1990s while Samsung piled into liquid-crystal displays (LCDs), and eventually replaced Sony as market leader. Embrace new ideas too early, though, and you may be left with egg on your face, as when General Motors spent more than $1 billion developing hydrogen fuel cells a decade ago, only to see them overtaken by lithium-ion batteries as the preferred power source for electric and hybrid vehicles.
To determine when to proceed with a new technology many managers and engineers employ popular heuristics, some of which are seen as “laws”. The best known is Moore’s law, proposed in 1965 by Gordon Moore, a co-founder of Intel. At first it stated that as more transistors are crammed onto the surface of silicon chips, the devices double in performance every year. This law was later revised to two years, and chip performance is now usually reckoned to double every 18 months. Other laws use “S” curves and various other calculations to predict how technologies will evolve.
Many of these laws have become widely accepted and are now applied when drawing conclusions about a broad range of technologies. Some have become self-fulfilling. Chipmakers, for example, use Moore’s law to co-ordinate their research and development (R&D) activity and plan their capital investment. In reality, however, such laws are unreliable because progress is rarely smooth. So Ashish Sood of the Goizueta School of Business at Emory University, Atlanta, and his colleagues have come up with their own law, which is explicitly based on the tendency of technology to progress in stops and starts.
Their “step and wait” (SAW) model, recently published in Marketing Science, notes that advances in performance are often followed by a waiting period before the next step forward. The steps can be big or small, and the waiting periods long or short. The researchers also hypothesise that greater support for innovation means new technologies improve in larger and more frequent steps than old technologies did. This is the result of higher R&D spending, the existence of better tools and the fact that more countries are undertaking research. But as the number of competitors in a new field increases, both the size of the steps and the length of the wait for the next step can change. Read more of this post
