4 Common Software Localization Challenges And Their Solutions
Earlier this year, we had the opportunity to hold a very informative webinar about software localization with experienced localization managers of Stripe, Uber, Evernote, JDA Software, and InEight, during which they shared some of the biggest challenges they are experiencing and even some of the solutions they are applying.
In this article, we will further review those insights.
Coping with Agile software localization release cycles
The waterfall model, in which a certain task needs to be finished in order to start the next one, is a thing of the past. This affects not only how the software is developed, but also the software localization team and processes.
Problems
- You have to get faster. Everything is faster now, and companies are noticing that the localization process needs to work with quicker organization, planning, updates, and also turnarounds.
- You need people on your side. You might need more buy-in to get the solutions you need.
Solutions
- Agile software localization tools. You’ll need a translation management systems that is web-based and has an API to connect with. More connectivity between your TMS, your CAT tool, and your software, the better.
- Buy-in. The fastest software loc teams have significant buy-in from management. The easiest way to get there is to make sure your team has a firm grasp of the technical details.
- Collaboration. You’re going to have to work together. The best TMS solutions help with in-project communications, but you’ll need more than that with your technical cohorts. A dedicated Slack channel is a good idea.
- Automation. There are lots of opportunities to automate if you choose the right TMS. They require some investment and best practice, but once it is running, you’ll save alot of time.
Lack of context and arduous functional testing
Unlike regular document translation projects, in which linguists can see full sentences and paragraphs with lots of context clues, software localization strings often have no context whatsoever and can be very enigmatic.
Problems
- Translators to spend extra time guessing, asking questions, and waiting for answers.
- Beefing up functional testing for localization becomes a priority. This testing considers reviewing the complete software, which can be very time consuming, repetitive, error-prone, and therefore expensive.
Solutions
- Live Preview. Wordbee offers Live Preview for software localization that goes beyond the limited scope of other systems. It’s also perfectly compatible with your world. It’s perfectly legit live preview that connects directly with your system via API.
- Pseudo-translation (see text expansion below).
Inconsistent terminology in the software interface
When starting localization projects, companies tend to just jump right in, not paying enough attention to important details that may have big consequences in the future. One of these “details” is terminology.
Problems
- Translation inconsistencies. Login or Log in? Send, Submit, or Next?
- Confusion everywhere.
Solutions
- Standardization from day one. Nobody wants to do this, ever. But you’ll be glad you did.
- Terminology databases. There can be many ways to translate a single term, so it’s very important to create glossaries and terminology databases that include accurate and approved keywords and phrases.
Text expansion wreaking havoc with localization
The words in different languages aren’t always the same length, and how you deal with this challenge might determine whether you can keep up with your agile software localization schedule or not.
Problem
- Messed up interface. The difference in length and density is a real problem, and you don’t prepare for it, you can end up with crowded texts or overlapped strings that will affect your original design and layout.
Solutions
- Use pseudo translation. This process lets you simulate how translation will look when you expand the text by a certain percentage, so you can spot potential problems beforehand.
- Design your front-end application to accomodate text expansion from a technical point of view.
Building the right software localization translator team
Technical aspects are one thing, but we can’t forget that the human factor is the most important thing.
In a software localization project there are many actors involved, including engineering teams, marketing teams, product teams, and language teams. All of them are very important, but only the language team is the most directly connected with the final message that your company is going to convey.
Despite this, some firms don’t worry too much about hiring the right linguists with the proper cultural background for the job, nor do they maintain a regular staff of translators. This affects the overall quality of the translations, making the final message seem uneven and unprofessional.
What can be done, then?
Create language teams with local and specialized translators that actually speak the target languages to keep quality standards high. And just creating teams isn’t enough: You need to maintain them over time, building good work relationships, and giving them the right tools and trainings to support their work.
Is this interesting?
Subscribe to get interesting localization podcasts, discussion panels, and articles every month.