Catch up with the first post in this series, where we discuss with localization experts from Netflix and Skyscanner.
As we build and develop Frontitude, our team meets with leading product and UX teams worldwide, discussing their UX writing workflows, challenges, tools, and more.
In this blog series, we focus on companies with highly complex localized products and interview their team’s localization experts and translators, as they are the ones who initiate and own the localization process.
So, grab your favorite coffee. We're waiting for you.
All set? Please meet...
What do you consider the most significant challenges in product localization these days?
Natalya: Building good communication between all stakeholders and decision-makers and creating an effective process. Developers, UX designers, content writers, and everyone involved in the product creation, need to put localization in the loop. They need to understand what they are creating and why in order to make the right decisions and provide the best quality. In some cases, it's better to create a different flow for a specific market instead of localizing it or perhaps create a single product that fits all your focus markets.
In any case, you need to dedicate lots of time and effort to integrating the localization process into the product development lifecycle. Localization is not a delivery service that can be provided after a product is ready. It requires teamwork that involves all your product creators; it touches and depends on all aspects: code, design, product copy, and content. Even when you choose a name for your product or company, discussing it with your localization specialists is better – it can save your reputation and save your money.
"Localization is not a delivery service that can be provided after a product is ready. It requires teamwork that involves all your product creators..."
Ketty: When talking about software products, the biggest challenges encountered are related to design. The design often doesn’t accommodate for languages much longer than the source and contains concatenated strings or un-extracted strings, resulting in partially translated final products and a tight time for market readiness. The final outcome is often a rushed localization process that leaves little or no time for translation, internationalization testing, and in-context review.
Finally, a lack of proper localization practice knowledge often leads to wrong decisions which tend to make the overall process cumbersome, much longer than anticipated, and, frankly, painful for all parties involved. Lack of instructions for translators, errors in the source, and last-minute changes can result in issues in the final localized product and/or with the localization process itself.
"Lack of instructions for translators, errors in the source, and last-minute changes can result in issues in the final localized product and/or with the localization process itself."
What information (i.e design context, content style guide, etc.) do you think translators need in order to provide high-quality translations?
Natalya: It should be the same information your original content writers get: briefs, guides, visuals, and any other necessary context (for instance, not only where on the screen the element is located but also what happens when the user clicks something or what their state of mind or intention at this point, etc.). Also, one of the most important things you can share with your linguists is a product glossary with definitions and context. Sometimes, it's critical to explain why a specific term was chosen and how your content writer came up with it. If your linguists understand the product logic, they can find the right message that will save devs and designers time.
Ketty: Translators should understand the final use of their piece of work, how the translations are placed in the wider scheme of things, and their final usage and audience.
The language might significantly change depending on the answer to all of the above questions: from the tone of voice, to the level of formality in language, to the way of addressing. This is why language assets, such as robust term bases and style guides, are essential to the delivery of a good translation that uses a language that resonates with, is appropriate for the target audience, and is in line with the tone of voice and intentions of the source writer.
In addition, getting full context is essential while translating: previous translations, both for the source and the target, contextual explanations and visual aids such as screenshots, or a design of the source text in context to gauge the appropriate length of the final translation. It would also be ideal to create a communication channel for translators to interact with the source writers.
"language assets, such as robust term bases and style guides, are essential to the delivery of a good translation that uses a language that resonates with, is appropriate for the target audience, and is in line with the tone of voice and intentions of the source writer."
Have you explored new tools or features to improve localization quality/process recently? If so, what tools/features did you use, and how did it go?
Natalya: Well, it's not new, but we started using Figma not so long ago, and we are working on implementing the design stage localization and pseudo localization. It helps a lot and saves time and effort dramatically. We are also looking for tools to automate our language QA process for linguists.
Ketty: We’re constantly on the lookout for new features, tools and integrations with our current tools with the objective of streamlining and making our localization processes more robust.
We have started leveraging the Figma plugin for image localization specifically, ensuring that the most prominent images in our UI, that were previously left in English, are now translated through a simple click of a button.
Also, using the pseudo-translation function in Figma, it’s possible to foresee translation text expansion and adjust the designs accordingly ahead of translation. By seeing the translations in context, the final layout can be tweaked and correctly adapted.
Additionally, we’re employing our translation tools in favor of a robust linguistic quality analysis that can help us identify potential quality concerns as early as possible and tackle them accordingly.
Our team also worked on the implementation of a tool that helps expedite vendor testing sessions through automatic UI screen grab capture. It releases pressure and lightens the workload of internal teams. Also, it allows more frequent testing and overall quicker bug resolution, as well as an increase in the percentage of content functionally and linguistically tested ahead of product release.
What do you wish to improve or find missing in your localization processes?
Natalya: We wish for a quick and easy language QA process. Before each release, we ask our linguists to review all product flows and screens, trigger all content, and ensure everything is logical, and enables users to complete their tasks quickly and smoothly. We want to ensure that each language version looks and sounds as good as the original and that our users would feel that the product was not localized but created in their language. QA is an essential step in our process – it's technically complicated, involves a lot of manual work, and requires lots of time from our linguists. So there is plenty of room for improvement. Also, it would be great to have a good glossary management tool that enables smooth collaboration.
Ketty: I believe localization teams are in constant need to improve and adapt their practices to the fast-growing, fast-changing localization environment. Overall, the main objective tends to be making processes more streamlined, efficient and faster.
In our case, despite the sophisticated localization processes we have in place, I’d say there is still room to improve in a few eras: streamline the feedback loop to keep translations and linguistic assets always up-to-date; work on delivering faster translation output by further employing machine translation capabilities; build a robust integration between our TMS system and our internal managing tool for source and target strings and making the data gathering easier, centralized and more efficient.
Thank you both, Natalya and Ketty, for taking the time to speak with our team about your localization process and for sharing with the entire community your insights and knowledge.
Did you find this content helpful? We're about to publish the next blog on this series -- stay tuned so you won't miss it!
If you want to participate in one of our upcoming UX Writing Over Coffee posts, or suggest any questions for next posts, feel free to drop us a line.