Verisign has analysed the type of web content associated with Arabic script second level domain names in .com and .net. The data sample was 27 000. The sample excludes those where there is too little text to determine the language of web content (21 000), and where the type of website cannot be determined (2 000), leaving 4,104 Arabic script .com and .net domains. The low usage rates exemplify the inherent complexity of combining right-to-left Arabic script second level domains, with left-to-right domain extensions (.com and .net).
Although English is the most popular language associated with .com and .net Arabic script IDNs, it is striking that nearly 90% of the sites are typical of low involvement web content – meaning that the domain name holder has invested little or no energy in content creation (pay per click, parking pages, starter pages, and redirects). Similarly, 70% of “Other” language websites (none of which are strongly associated with Arabic script), are used for low involvement content.
A very different picture emerges where the language of web content is Persian and Arabic. These have “high involvement” content (blog, online business, e-commerce), indicating that the domain name holder has invested energy and resources into content creation.
Nevertheless, the majority of Arabic script domain names in .com and .net which are associated with Persian and Arabic language content are used for online business purposes. This fits with general usage patterns:
- Individuals are tending to migrate away from domain names to social networks, but a minority create blog sites to publish their writing (some organisations use blog sites of course).
- Businesses are more likely to register and use domain names to create an online presence
Note that the data sample is small, and therefore generalisations may be of limited value. However, in this small sample, Arabic script IDNs are more likely to be used for “high involvement” sites where the language is strongly associated with the IDN script (eg Persian or Arabic). Where the language of web content is not strongly associated with the IDN script (English, Other), the content is “low involvement”.
Strong correlation between local language content, country of hosting and script of IDN
We have seen that local language content is more likely to be hosted in country.
In the .com and .net sample of Arabic IDNs, 1132 Persian language websites were categorised as “online business”. Of these, 82% were hosted in Iran. Only 35% of Persian language blog sites were hosted in Iran. The reason for this difference is not clear. It may be that business websites tend to be more sophisticated, favouring a personal relationship between the business and web developer in the business’ own language or locality. In contrast many blog sites, if not exclusively for personal use, can now be set up in multiple languages, using popular templates (eg WordPress) without the involvement of a web developer.
The hosting country for Arabic language / Arabic script IDNs in the sample of .com and .net names were overwhelmingly out of region (98%). This goes against the correlation found in the ISOC study, but the data sample is small (215 domain names) so it is difficult to draw general conclusions from this.
Originally published in World Report on Internationalised Domain Names, 2014