Localization Testing

After an application has been localized, it must be tested before market release. While some may worry that testing increases time-to-market, it should be noted that the cost of correcting a problem increases dramatically over time.

There is a slight but significant difference between localization and linguistic testing. Here are simple definitions:

• Localization testing focuses on the correct functionality, appearance and completeness of the localized product.

• Linguistic testing takes care of ensuring the correct language rules are being used and focuses on correct in-context linguistic usage.

Testing has often been considered only for software that is localized. But, in fact, all localized content should be tested to make sure it is correct. Whether the localized content runs a version of software for Asian audiences, or whether it appears on the side of a box containing the company’s product or in an online ad, it represents the company and should be considered as important as the original content.

Localization testing focuses primarily on user interface but it also reaches farther—in fact, the localization process can introduce severe functionality problems to the software. Those problems can be caused either by over-translation of some system variables that are invisible to the target user and must not be translated, or by modified functionality, which sometimes must be implemented to the product to meet local market expectations. Letter wizards and spell checkers could be the typical examples.

Localization testing requires both source and target language versions of the product installed on the environment that a typical user would use. Therefore attention must be paid to the correct version of the operating system, language, regional settings and more. The builds used for this testing must also match in terms of functionality – localization starts at an early stage of product development where all features are not yet implemented, and mismatched localized and English builds cannot provide the expected functionality testing consistency.

Testing teams usually follow the steps that lead them through the software, paying attention to the tiniest details and differences. They look for truncations, misalignments, untranslated strings…even a missing period is an issue. Plus they verify the functionality, as mentioned. All defects must be well documented, tracked in a database and of course verified once the new build with fixed defect is created. Manual localization testing can be – to a great extent – replaced by automated localization testing that can discover truncations and other user interface defects but has limited possibilities when performing functionality testing. Automation also requires someone to prepare the scripts first; therefore automation is often efficient only for products localized to a higher number of languages.



Localization Blunder
In a Belgrade hotel elevator: To more the cabin, push button for wishing floor. If the cabin should enter more persons, each one should press a number of wishing floor. Driving is then going alphabetically by national order.