Marketing RSS

  Predicting a Product's Success Internationally 

Gelper, Sarah; Stremersch, Stefan
Will a new computer or mobile phone be big in Japan? What about in Argentina, India, Morocco or Russia?

Predicting a new product's growth internationally is a huge topic in marketing. And it's a hotly debated one.

Does a new product's potential growth depend upon a country's wealth, urbanization levels, individualism, education or openness? Which variables matter most?

Professors Sarah Gelper and Stefan Stremersch find a way to analyze a wide range of country characteristics in order to analyze their relevance to predicting the growth rates of six high-technology products in 55 countries. With their technical model, Gelper and Stremersch are able to filter out the white noise to detect the two most important predictors of growth across the board: economic wealth and education.

Economic wealth is, quite simply, the best predictor of product growth in the study. In wealthier nations, new products sell better. The second best predictor is education. Of all the socio-economic, cultural, communication and demographic variables analyzed, wealth and education mattered most to marketers' forecasts.

The co-authors present their results and their means to select their variables -- a model which could be used to study other types of product launches in even more countries, taking a wide array of country characteristics into account.

Global Data, Many Variables
Gelper and Stremersch employ a technical model that accommodates a wide range of variables, even when data is relatively sparse. The following diverse variables were examined in the study:
  • Socioeconomic variables -- economic wealth, income inequality, poverty, education, economic openness, the percentage of women employed and the ratio of economic participation.
  • Cultural variables -- including individualism, uncertainty avoidance, masculinity and power distance.
  • Communication variables -- including media intensity, mobility and incoming tourism.
  • Demographic variables -- including population growth rates, population concentration and the percentage of people living in urban areas.
The model allows the co-authors to include this wide array of data from countries in different stages of development to produce a more global picture, even if some data points are missing.

With the well documented penetration rates of six high-technology products -- the mobile phone, CD player, video camera, personal computer, Internet and ISDN -- the co-authors use their model to home in on the variables that best predict penetration rates.

Homing In on What Matters Most
Stremersch and Gelper find that international differences in new product growth patterns are largely driven by socioeconomic factors. In contrast, cultural variables were not found to be significant in this study.

The co-authors look at the impact of their variables on three key growth metrics in marketing science: 1) market potential, 2) the coefficient of innovation and 3) the coefficient of imitation. The latter two variables, innovation and imitation, interact in marketing models as a certain percentage of a population take to a new high-tech product and then another percentage copy the innovators.

As mentioned, a country's wealth (GDP per capita) had the biggest impact on all three growth metrics, meaning it is the first variable marketing scientists should consult in order to assess the likely success of a product launch.

After wealth appears a more surprising factor: education. The number of university students as a percentage of the total population matters to new product adoption. Education was found to influence both the market potential and the innovation coefficient in this study.

Another socio-economic factor was found to have a negative impact on growth: income inequality. Where income is more unevenly distributed, product penetration rates are lower.

On the flip side, economic openness -- measured as the value of a country's imports and exports divided by its GDP -- was found to positively affect the innovation coefficient, as education does.

Meanwhile, tourism was found to positively affect market potential. The more incoming tourists a country receives, the more contact the country has with new technologies, apparently leading to higher rates of adoption. Finally, mobility -- measured via car ownership rates -- affects the imitation coefficient, due to increased interactions between people.

In sum, six variables were selected as influential. All the others were not. The co-authors explain: "Thus, after controlling for the included variables, [the rest] do not provide any additional information about the diffusion process, in our sample of products and countries." Put bluntly, marketing executives may feel free to ignore them.

Upgrade Your Marketing
The co-authors hope to see their variable-selection model expanded to include other types of products. They also see the potential to analyze even more variables. For example, distribution infrastructure, competition or regulation could be analyzed across countries.

The large number of potentially influential country characteristics combined with the large numbers of products and countries can been daunting to marketing professionals. The co-authors' approach to variable selection, finding the most important drivers of difference in international diffusion, could help focus forecasting on what matters.
This article is based on:  Variable Selection in International Diffusion Models
Publisher:  Elsevier
Year:  2014
Language:  English
Go to source English