BLOGS
How to Avoid Costly Localization Issues in Structured Content
May 4, 2023
"My team spends too much time and effort fixing issues in localized content before publishing. What can I do?" This is a common question from companies that handle a large quantity of content in a structured content system.
Your team creates the original English content, you send it off for translation and localization, and then the problems begin. Even small formatting issues can quickly mount up to become hours of manual work when you are dealing with multiple languages and dozens of content files.
With the proper knowledge and systems in place, you can avoid the top 3 issues when localizing structured content.
In our recent Rubric Roundtable, our structured content experts sat down and discussed some useful strategies for avoiding issues.
Here's how you can streamline your localization pipeline…
Why waste time and effort fixing localized structured content?
Structured content can be a highly effective tool for streamlining your content delivery process, saving your teams many hours of unnecessary duplicated work. But this framework is only effective if you use it in the right way.
With most structured content systems, you set the rules for how content is stored and managed. This is useful because it allows you to cater your system to your specific needs. However, this flexibility can introduce inefficiencies that you might not notice… until you start localizing your content.
An example: Poor use of structured content "snippets"
Imagine a manufacturer uses a structured content system to create product catalogs. They set up their system to store the product information in "snippets," which is the name for content fragments in structured content.
The team has stored the product variant number in the same snippet as the product description.
This solution may work fine in English, where every unique product variant is associated with a corresponding product description. The problem with this approach might arise when the company localizes their product catalog for different international markets.
Let's say the company serves France and Canada, both markets that speak the French language. The product descriptions in the two markets are the same, as the team has identified with their localization provider that the regional French language differences are minor enough that they won't create two separate versions. However, the product numbers need to be different for both markets.
The simple choice of storing product numbers and descriptions in the same snippet now causes a problem. It creates unnecessary extra work after translation as team members may need to manually enter the product number into each snippet.
To avoid such problems, you need to design your structured content models in a way that makes localization easy.
The 3 key sources of localization issues with structured content
What causes localization issues when you are using structured content?
According to Rubric's Global Content Business Analyst, Rebecca Metcalf, there are 3 common sources of localization issues that come up again and again.
1. Import issues when integrating localized content
Content teams often first notice localization issues when they try to import localizations back into their systems for publication.
Rebecca says:
"The team receives content back from the localization vendor, but it won't import into their system again. Often, this is because of the way the content has been handled. Structured content files are often quite complex in how they're coded. The files can get broken easily, depending on how they have been put through different systems or tools."
An example of this is with HTML files. The translation process might create syntax errors that need to be fixed before publishing.
2. Unexpected formatting that breaks structured content.
Another common area where fixes are required is when the English source itself has formatting that breaks during the translation process.
Rebecca says:
"An example would be putting quotation marks in ALT text fields. These are often changed during translation, which then gets interpreted by the system as an "end of string" code that breaks the process."
3. English-focused content structure
Finally, problems can occur when your structured content is too reliant on the rules of English.
Rebecca explains:
*"If variables or snippets have been used in such a way that they're built around English sentence structure, these aren't the same in other languages. Units of measurement and numbers also aren't formatted in the same way. These generally cause quality issues."
A simple example of differing grammar structure is the fact that English sentence structure is usually Subject-Verb-Object. By contrast, Japanese uses the format Subject-Object-Verb. If you were to use structured content to automatically replace the "Object" for different product names, this grammatical difference could end up causing issues.
This one handy rule can save you time and effort
How can you reduce the issues that occur when localizing structured content?
The key is to view your localization activities as a process. When there are problems with the translated content, always go back to the beginning of that process.
Rebecca says:
*"For many issues, the answer is to go back to the English and fix it there so that the problem doesn't occur in the localized versions. If the problem is in building content, they should be worked out by improving the localization process.
This simple rule runs counter to how many people view localization. When they find problems with the translated content, they try to fix the translations. But this approach can lead to having to fix the same issues again and again.
How to work with your localization provider to improve your localization pipeline
Working closely with your localization provider is key to ensure your process is as efficient as possible.
Dominic Spurling, Rubric's IT Director, says:
"When we have a new client working with structured content, we generally find that we discover a lot of the niggles in the first couple of handoffs. We introduce configuration into our pipelines to handle anomalies or unusual aspects of the content. We also encourage the client to improve the quality and consistency of the English."
With structured content, our emphasis is always on repeatability.
By having a repeatable process that can easily handle bulk content, your localization process will become more stable with every run. This can contribute to huge savings of time, budget, and effort in the long run.