Software Localization: 5 Crucial Challenges Resolved!

Key Solutions to the Biggest Software Localization Challenges

1. Localization as a Second Thought: More Costly and Time-Consuming Software Localization

Many software localization errors or costly workarounds occur before the actual localization has even started. You may not realize this, but developers already hold some power over how successful and economical the software localization process will be.

The Solution: Think Ahead and Internationalize! 

Before you get started with the localization of your software product, internationalization should come first. When done right, internationalization makes localization easier, less time-consuming, and more cost-effective, since the content can more easily be adapted to different languages.Internationalization involves the preparation of your software for the localization process and removing all elements that refer to a single language application making it functional for multiple languages. Here is what that means in more detail.First of all, internationalization will simplify localization for the User Interface (UI):

• A skillful adaption to text expansion and contraction is vital. Did you know that some languages can take up to 35% more space than others? If not dealt with on a professional level, text expansion can lead to errors on both the technical and visual front.
• If you are translating into RTL Languages such as Arabic or Hebrew, you should add the markup that supports bidirectional content in your DTD. Of course, some languages, such as Asian languages, are written vertically. In this case, or in the case of any other typography with no Latin elements, you can use CSS support.
• You may be using a lot of colors and images, and these will also need to be adapted. To make this easier, do not hardcode your images and use editable formats instead.

Your strings and data format should also be developed in a manner that makes localization easier. This includes:

• No hardcoding! We may have mentioned It before, but we can’t overstate it. It will be best if you enable Unicode. Note that the most popular encoding is UTF-8.
• Similarly, your coding should be enabled to support local formats such as numbers, time, dates, and so on. For time and dates, the ISO 8601 standard is commonly applied, while numbers such as currencies are commonly standardized through the ISO 4217.

Let’s look at a striking example of internationalization done well. Developers can prevent possible time-consuming fixes related to, for example, text expansion or contraction by performing the so-called “mock translation†or “pseudo localizationâ€. This approach estimates text expansion and contraction ratios by replacing specific text segments with random letters.Netflix, for example, heavily promotes pseudo localization as one of its pre-localization strategies.1 The company stated that “as we added language support for our members residing in 190 different countries, scaling globalization at Netflix has never been more important. Along the way we’ve built out countless solutions to help us achieve globalization at scale.†Have a look at an example of their pseudo localization for their mobile app!

2. Upholding Cultural and Linguistic Standards

When it comes to the linguistic side of things, localization is about more than just language translation. You have to make sure you choose the appropriate regional version of the language, and the cultural references should be effective and respectful.If you are in a situation where the language, as well as the culture, are not completely respected, you may just alienate your target audience rather than connect with them. Here is how you nail it!

The Solution: Rely on Native Experts and Linguistic QA

It is a challenge getting both the language and the cultural references completely right. Have a look at this scene from Netflix hit show Stranger Things, which is available in over 20 languages including English, Castilian Spanish, Latin American Spanish, Italian, German, Japanese, Brazilian Portuguese, and French. The official Stranger Things video explains “When language isn’t a barrier, great stories have the power to travel the world.â€The best way to ensure these kinds of powerful, accurate, and culturally savvy translations, is to connect with experts. These experts should:

• Be native in the target language and reside there. Note how the Stranger Things team makes a distinction between Castilian Spanish, Latin American Spanish, and Brazilian Portuguese. That is because even if the same language is spoken in different countries, each country and region will have its own dialect and unique linguistic and cultural traits.  
• Have translation experience and industry experience. This way, the translator can make intuitive choices.
• Follow a proven translation methodology that follows the TEP (Translation, Editing, and Proofreading) Process, so that any inaccuracies and cultural faux pas are weeded out.

To further confirm the quality of the translation, linguistic testing should be performed. Here, the spelling, grammar, cultural validity, and so on, are all double-checked for quality and compatibility with the culture the content is aimed at. A fine example of top-shelf translations in the game and software localization industry is PUBG Corporation. The South Korean online game developer has a worldwide player base of hundreds of millions 2 and with games such as New State available in 17 languages.3

Part of what makes PUBG’s localization strategy effective is the considerations they have given to cultural sensitives as well as languages. For example, when their PlayerUnknown’s Battlegrounds was considered too violent for the Chinese market, PUGB commissioned a fully localized version of the game and renamed it “Game for Peaceâ€.4 And while guns are still used in the game, the level of gore prevalent in the original game is no longer present.

3. So Many Things Can Go Wrong on the Technical Front

When not handled professionally, Software Localization can lead to a myriad of bugs and errors undermining the system. And one of these is caused by the above-mentioned text expansion. What if the developers did not apply internationalization effectively or were not very successful in their pseudo localization efforts? Have a look at this example of a “Play Now†button.Text expansion such as this can wreak havoc with the software’s operating system. Visually, text expansion causes quite some issues as well. We will elaborate more on some other visuals in the next part, but when it comes to text expansion, text can appear truncated, flow outside of the boxes, or even overlap.  Beyond text expansion, other possible issues such as compromised code during the translation stage can contribute to bugs and errors.

The Solutions: Localization Engineers, Tools, and Functional QA

If the internationalization stage hasn’t been effective in predicting text expansion issues, there are still ways to resolve them. This is where localization engineers and their sophisticated translation and localization tools have a vital part to play.Translation CAT Tools are not only there to help make the translation process run more smoothly for the linguists. Localization engineers also leverage the tool’s Quality Assurance (QA) capabilities to deliver an error-free product. For text expansion, the following techniques can be used.

• The maximum number of characters can be specified, and the tools will notify you whether the limit is met. This can be performed on the segment level. 
• Alternatively, the engineer can apply a certain value to the entire file rather than just the segments and can apply a certain percentage under the QA parameter settings that cannot be exceeded.
• Some CAT Tools also have a translation preview option or a Software Translation Simulator, so translators can view their work in the correct framework and make adjustments accordingly.

But there was another concern we raised. How can localization engineers ensure that the code will not be compromised during the translation process? Enter regular expressions!Regular Expressions are implemented into the CAT Tool by the localization engineer. It identifies the content that should not be translated, or the code, and locks it so it cannot be edited. This non-translatable content will be displayed as a tag. Have a look at this screenshot from one of our projects. Locking down the code (purple) means that linguists cannot accidentally edit these strings, and a wide range of possible errors are prevented in the process.Finally, any software localization project must include functional testing. What does this entail? The software performance, code, text expansion, and more are tested and a smoothly running UI is confirmed.