Keeping abreast of technology is difficult. It might seem that you’ve no sooner unpacked your new laptop before its manufacturers introduce an upgraded or entirely new product. When the next generation of computer, digital camera, mobile phone or even vacuum cleaner is launched on the market, it will often co-exist with its predecessor for some time. Companies then must continue to devote time and resources to improving the old generation and reducing production costs while pushing ahead with developing the new. In the competition for firm resources, how companies balance the need to focus learning resources on two generations of products is a crucial problem.
In this Working Paper, Professor of Technology Management Lieven Demeester, and Adjunct Professor Mei Qi, at INSEAD argue that firms can use learning resources such as experienced manufacturing personnel, technical experts, process engineers and test laboratory time to influence the rate at which operation costs fall. By studying the decision problem mathematically, they identify some important factors that firms should consider when timing the switch from one generation to the next.
Thus far, studies of firm learning have tended to focus on the learning curve, documenting and quantifying the relationship between cumulative production experience and production costs. While there is a link between the two, in-depth studies have revealed that cost reduction does not automatically occur with production experience. Deliberate learning activities matter so the authors identify the need to model the managerial decision problem of allocating learning resources. By looking at the example of a large semiconductor manufacturer, the authors model the effect of several factors on cost reduction.
They derive and discuss several propositions regarding management decisions regarding product generations. First among these is that firms should abandon the older product generation quicker if either the market potential, or substitution rate for the two generations is increased, or if the potential to reduce production costs for the newer product is increased.
These propositions are intuitive: firms should devote all their learning resources to the newer product generation earlier when it promises to increase revenue or have substantial market appeal. Second, firms should postpone abandoning the older generation if the firm’s learning rate potential is increased. If firms abandon the older generation too early, they risk not exploiting its cost reduction opportunities.
Cross learning, whereby the lessons learned from one generation can be transferred to the other, must also be considered. If a firm’s level of cross learning increases, it should also switch more quickly to the new generation, as learning will continue for the older model even when it is deprived of resources. Conversely, as returns on learning resources diminish, firms should postpone abandoning the older product generation, spreading its resources for a longer period of time.
Having identified the factors that affect cost reduction, the authors provide managers with a useful tool for considering the timing of reallocating learning resources. They also recommend that the decision-maker – be they an engineering manager or a manufacturing director – stay informed about production costs and market forecasts. In this way, this paper identifies a useful starting point for managers building decision support systems ideally matched to their firm’s specific situation.
In sum, and in line with other studies, the authors recommend that the optimal use of learning resources involves spreading them firm-wide, and keeping them flexible and mobile. In this way, firms can take advantage of cost reduction opportunities and be ideally positioned to harness any changes in the strategic importance of learning.