We released 4 versions of T-Index to allow for personalized and much more precise results.

• T-index version is designed to guide companies in choosing the best languages and the best markets to target, regardless of the product to be sold or the country to which to export.

• The country-specific T-index version takes into account the country in which the company operates and from which the products will be exported.

• The product-specific T-index version takes into account the country in which the company operates and from which the products will be exported and/or the particular product that will be marketed.

• The service-specific T-Index version takes into account the service that will be marketed.

T-Index

The T-Index value is calculated by taking into account the purchase power of the internet population. In order to estimate the annual per capita expenditure of each Internet user, we used the HFCE, (household final consumption expenditure) indicator from the World Bank expressed in current US dollars, the income share, the total population and the internet penetration rate. You can find the details of our calculation method below.

Each country was classified according to the language most commonly used by the local population to browse and make purchases on the Web. The T-Index study only includes languages used for browsing the Web. A number of languages are currently very scarcely represented online, with some not being used at all. In many countries, the language spoken on a daily basis by the majority of the population cannot be found online. This is often due to a low Internet penetration rate, which prevents local people from accessing the Web to create content in their native language. 

For countries for which there was no official breakdown of the percentage of the population that speaks each language, or where there were many dialects, we made an educated guess: we normalized the data or carried out qualitative research, always keeping in mind the economic potential of the use of the languages online. Most of the languages with international variants have been grouped together. For example, UK and US English have both been classified as English. The T-Index study includes all the countries listed by the WorldBank, which is our main data source, except the ones for which data is not available. 

Calculation

We downloaded the most updated data about HFCE, Income share, internet penetration, total population and GDP (which is used only when the HFCE data is missing). We then analyzed each country’s Internet penetration rate and income distribution to determine the proportion of the HFCE theoretically spent by Internet users. Finally, in order to obtain the “HFCE per capita of Internet users”, i.e. an estimation of their annual expenditure, we made the following calculation:

(Country’s total population x Country’s HFCE per capita x Percentage of HFCE theoretically spent by Internet users)

_______________________________________________________

Number of Internet users

For countries where income distribution data was not available, we calculated the average income distribution for all countries and used this estimate.

In order to determine the proportion of HFCE theoretically spent by Internet users, we assumed that the Internet users in each country belong to the richest segment of the country’s population. We made this assumption bearing in mind that a certain income level is necessary in most countries to purchase an Internet subscription and take part in e-commerce activities.

The product-specific T-Index

The product-specific T-Index is available for the countries covered by the COMTRADE database. Much in the same way, the value of the product-specific T-Index is obtained by taking into account the product import-export customs data and the wealth of the population segment with internet access in the buyer country. The difference between the two indices therefore lies in the fact that the first takes into account the total flow between the two countries, while the second only uses the flow of the particular product that we want to export.

Calculation

To calculate the country-specific T-Index we need three ingredients, namely trade flows between countries for the specific product, internet penetration rate, and income distribution. 

The percentage of spending by internet users is calculated by assuming that these users belong to the richest segment of the population, thus combining the data on internet penetration with the income distribution data.

If we are exporting a product P from country A, the product-specific T-Index of country B is calculated as: 

(percentage of spending by internet users) * (total exports from A to B of product P)

_______________________________________________________

Ʃ (percentage of spending by internet users) * (total exports from A to X of product P)

For countries where no income distribution data was available, we used an estimate by averaging the distributions of other countries.

The service-specific T-Index

The T-Index for services is calculated by multiplying the volume (in 2010 dollars) of services imported by the various nations by the usual corrective factor, given by the percentage of expenditure by internet users. The service import data is obtained from the Balance of Payments (BoP) provided by the International Monetary Fund. We can therefore see that, in this case, the information for the country of origin is not used. This is due to the fact that, unlike products, services do not involve transport costs and can even be supplied remotely in many cases. The T-Index for services is available for the nations listed by the BoP and covers 12 different classes of services.

Calculation

When exporting a service S, the T-Index for services of nation B is calculated as follows:

(imports by B of service S) x (percentage of expenditure by internet users of B)

Once again in this case, for countries where no income distribution data was available, we used an estimate by calculating an average of the distribution of other countries.

Study carried out by CNR-ISC and Sapienza University, in partnership with Translated.

The internet penetration in each country for the country-specific T-Index was obtained from the International Telecommunication Union (ITU) report titled ‘Proportion of individuals using the Internet’, while the income distribution by quintiles was obtained from statistical data from the World Bank’s Development Research Group. With regard to the data on trade flows between countries, we used a database built by our research group starting from the data collected by customs authorities and pre-processed by COMTRADE. In particular, the cleaning and reconstruction of the flows was carried out using special machine learning methods.