A New Mathematical Learning Curve Model Based on the Empirical Analysis of Japanese Sharing Economy Companies

Abstract

Mathematical learning curve models have been widely used in strategic and production planning. In some emerging industries, the learning curve effect may be delayed or inhibited by various factors. This study proposes a new mathematical learning curve model—the squiggly learning curve model—for these emerging industries. The model incorporates any potential impeding factors in learning. It can present all the learning curve shapes in a single formula, from traditional log-linear learning curves to the S-curve learning curve and squiggly or sawtooth-shaped learning curves. This study adopts a mixed methods approach that combines mathematical modeling, nonparametric regression analysis using smoothed spline methods, questionnaire surveys, interviews, and case studies. The findings confirm that the model suitably represents the shape of sharing economy companies’ learning curve; this learning curve is affected by several factors, including the increase in site-patrol costs to prevent users from violating rules and regulations, site modification costs, and salaries for customer service and sales employees, which occur as the number of sharing economy platform users increases. Thus, the model can be a useful analytical tool for operations and strategic management in the growth phase of emerging industries, such as the sharing economy industry, where learning impediments exist.