To reach global markets efficiently, you need a CMS that’s capable of integrating 3rd-party translation services as well as serving different variants of the same content in various regions. Moreover, localization is an effort of multiple teams sitting continents and time zones apart. That’s why it’s vital for the CMS to have support for granular role permissions and seamless collaboration across time zones.
Kontent.ai – the localization-ready CMS
There are several approaches to content localization and you can integrate services based on that. Do you want the best-possible human-made localization? Or do you need machine translation done fast and cheap?Usually, your organization’s approach depends on the content volumes it needs to localize, the nature of the content, and the amount of time and resources it’s willing to invest in the localization.Kontent.ai enables you to embrace any approach to localization, be it your local translators manually typing every word of every translation, completely automatic machine translations, or anything in-between. Every approach requires certain project settings, strategic planning, and coordination. Kontent.ai supports it all.
Strategic planning for localization
Content localization must be considered during source content planning and creation. Localization priorities and capabilities of localization teams need to be considered when you set up the localization process.
The structure of content meant for localization needs to support different local scenarios, such as differences in campaign requirements.
The source content must be ready for use with localized assets that might be completely different from those in the source.
Different localization teams have different capabilities. Some need only a little overwatch and can decide and act independently, while others need to be directed in more detail by the organization’s localization managers.
Your localization process also needs to take into account that localization speed isn’t uniform across locales and teams. This is especially important if you need to release certain content in different locales simultaneously.
Not all locales are of equal value to your organization. Your processes need to support that. For instance, locales with high revenue rates require meticulous human localization, while the low-revenue ones can do with mere translation.
Below are some aspects of localization you need to consider when designing your localization process.
High-volume localization
To achieve efficient localization of high content volumes, it’s best to outsource translation and localization services to a 3rd-party company specialized in localization or to local vendors. These other companies or vendors don’t have nor need to have access to your CMS. With a headless CMS like Kontent.ai, the best way is to employ APIs to exchange content between the CMS and the 3rd-party systems.
Machine vs. human translation
The decision whether to go with full machine translation, full human translation, or a combination of both depends largely on the content in question.Legal and regulatory content always needs to be checked for mistranslations, but this also goes for highly visible marketing content. A mere translation is not enough to make the content appealing to the target audience, and localization by a local team is recommended.Machine translation is called translation for a reason – computers often don’t know the context of the target audience and can’t localize the text’s subtleties.For sensitive content requiring human translation, the approach called computer-aided translation is recommended. People make the translation, but software handles certain aspects of the task, such as providing translation memory, grammar, spelling, terminology checks, and so on. Software providing these services can be integrated with Kontent.ai if it supports communication via API.
Roles and permissions
There usually are many people across many teams working on content translation. That means your CMS needs to provide options to control access to content.Access control, in the context of localization, means that, for example, the Japanese team doesn’t need to and shouldn’t be able to modify German content. And vice versa. You should also set your project up so that local teams can’t modify the source content.In Kontent.ai, there are several ways to achieve this control:
Set up languages and roles for each locale so that the team for the Brazil locale can’t edit Bengali content item variants.
Put localized content into separate collections based on the locale and set up local roles so that teams with roles for the French locale can work only with the collection for French content.
You can, of course, combine these approaches.
Tools for localization in Kontent.ai
Every content item can have one variant per each language you set up in your project.
Manual translation
Manual translation and localization can be done directly in Kontent.ai – your translator opens the source and target content item variants in two windows and creates the localized variant. However, this process is slow and error-prone because there’s no option for computer-aided translation, such as translation memory.
Computer-aided translation
There are various systems for helping translators create consistent localized content. These systems have their own interface that can’t be integrated into Kontent.ai per se. However, if they can communicate with external tools via APIs, you can integrate them with Kontent.ai for content exchange:
The 3rd-party service gets source content from Kontent.ai via API.
Your translators do their job in the 3rd-party tool’s interface.
When they’re done, the 3rd-party tool uses Kontent.ai’s API to push the translated content back into the right content item variant in Kontent.ai.
The advantage of this approach is that translators have access to all the features of computer-aided translation, such as translation memory or machine-based pre-translate, while the localization itself isn’t machine-based, so it can be of the highest quality.
Machine translation
If you only require simple translation that’s fast and cheap, machine translation may be the right way for you. You can integrate a service that provides machine translation via APIs. It then works the way that Kontent.ai pushes the source content into the service via API, the service machine-translates the content and pushes it back to an appropriate content item variant in Kontent.ai.The advantage of this approach is that it’s fast and cheap. The downside is that any content, such as idioms, that requires localization to be intelligible for the target audience, gets only translated, and the intended meaning likely gets lost.
What's next?
Content-first approach
With a new content management system (CMS), there are usually changes coming. They’ll indeed come with Kontent.ai as well. But they’re good changes.Kontent.ai enables everybody to work with content efficiently by separating content creation and content presentation. With this division, you can provide smooth customer experience by effortlessly publishing on every channel your customers are using.
