Internationalize your app and multiply your user reach easily. Sign up now

OneSky Logo Blog

Insights on international expansion

A Step By Step Guide to Pre-translation

pre-localization-guideAccording to Daniel Gouadec, author of Translation as a Profession, the translation process is a series of translating activities involving three phases: pre-translation, translation, and post translation.

Daniel suggests that most translation problems and potential risks could be resolved by getting as much information as possible prior to a translation project. Therefore, investing more resources in the pre-translation phase is a more efficient way of managing risk than having to deal with different pitfalls during or after the whole translation process. Let’s take a look at how to do just that.

What is pre-translation?

The “Pre-translation Phase” includes everything that takes place up to the moment the translator receives all the materials and resources for translation.

In order to minimize risks in a translation project, translators and their clients should do a pre-translation analysis to clearly discuss terminological issues (e.g. glossaries, previous translations, audience profiles, tone of voice…etc) and technical issues (e.g. CAT tools to be used, previous translation memory, contextual resources…etc).

That’s a lot of stuff to discuss. How does a project owner get started and actually do all of that? Well, this blog post will guide you through all the major steps in the pre-translation process.

Step 1: Separate Text (Strings) from Code

Oftentimes, developers make localization complicated because they didn’t manage their apps’ UI strings properly in the first place.

Move all text into a string file

To make localization straightforward, always separate string resources from your code and markup to create a single language-independent codebase. Hard coding any strings will make it difficult to update and localize. Instead place all the strings into a resource file, like a .strings or .xml file, so your strings can be extracted, translated and integrated back into your app without any changes to compiled code.

Provide sufficient context for declared strings

When declaring strings in your resource file, make sure to describe the context clearly in which the string is used. For example, what is this string for? When and where is it showed to the user? Is it a button or a text box?…etc. This contextual information is critical to translators to produce better quality translation.

Mark message parts that should not be translated

Over-localization is a common source of problems. To avoid that, ensure that only strings that need to be translated are provided to localizers. And if certain parts of a string shouldn’t be translated to other languages (e.g. a piece of code, a placeholder for a value, a special symbol…etc), mark text that should remain as-is, so that translators don’t touch it.

Step 2: Make Sure to Support Plural Forms

The next very important step, but often overlooked, is making sure to support plural forms.

Project owners and developers, who haven’t been involved in localization projects in the past may not be aware of these very important linguistic nuances. Different languages have different rules for the formation of plurals. For example, English has two forms: a singular form for “one” (e.g. 1 book), and a plural form for “everything else, including zero” (e.g. n books).

This may sound simple in English, but other languages make finer distinctions. For instance, Russian has different plural forms for numbers ending in 1 (except 11), numbers ending in 2-4 (except 12-14) and other numbers.

While deciding which case to follow for a given language and quantity can be complex, linguistic plurals are a key part of creating high-quality multilingual products that are relevant to local cultures.

Step 3: Pseudo Translation

Pseudo-localization could be the most cost effective way of finding localizability bugs in usability and functionality because it gives you a dummy translation without the cost of a real localization project.

What is pseudo-localization?

It is considered to be a part of the internationalization testing process performed during the product development phase. Instead of professionally translating the content of the software right away, the textual elements of a product are replaced with a dummy translation. So your team can get a feel for what the localized product will look like.

What are the benefits of pseudo-localization?

It helps to find out how prepared your product is for localization. This testing method can reveal hard-coded strings that should be translatable, find strings that shouldn’t be translated, and identify design/layout issues that will affect the localized product in the future. Generally speaking, the earlier you can identify localization problems, the faster and less costly it will be to address and reverse them.

Different ways of pseudo-localization

Pseudo-localization could be done via script or a tool. Below are some common techniques for different localization objectives:

  • Replacing characters with Xs
  • Pseudoese (similar-looking diacritical characters)
  • Lorem Ipsum
  • String Identifiers
  • Simple prefixes/Suffixes
  • Machine translation

For the detailed benefits of each method, check out this blog post.

Step 4: Define Glossary Terms

In every project, there are usually a few key terms that define the product (e.g. “Like” in Facebook) that we need to make sure they are translated accurately and consistently in every use case. This is where a glossary – sometimes called a terminology database or termbase – comes in as it is specially designed for this purpose.

A good glossary eliminates ambiguity and serves as a guide to translators on how to manage key terminology. After generating a list of key terms (e.g. acronyms, names, titles or subject specific terminology) and providing a clear definition and proper context, project owners will work closely with translators to review and approve the final translations.

When translators are working in a CAT tool, these key terms will be highlighted together with the pre-approved translations which enhance consistency and accuracy each time a key term occurs. In cases where there is a high volume of repetition, it can save both time and money as well.

You may learn more about how to set up a translation glossary via this blog post.

Step 5: Use Translation Memory

Translation Memory is a translation tool used to monitor and assist with the translation process by storing all the content that have been translated previously. Similar to a glossary, Translation Memory maintains consistent translation throughout the project.

So what’s the difference between Translation Memory and Glossary? A Glossary is a guide for translators on how to translate specific key terms, whereas a Translation Memory automatically translates repetitive content, eliminating the need to re-translate the same text in current or previous projects.

During the pre-translation analysis, a Translation Memory is useful to calculate the estimated word count and costs for new or pre-translated segments. As a result, updating Translation Memories is a very important step after each translation project. They will not only significantly reducie translation costs and turnaround time but also improve overall translation quality in future projects.

Step 6: Create a Localization Style Guide

Many people mistakenly think that localization means to translate their content word-for-word. They didn’t know that in order to create a successful localized product, they should adapt their messages to the local culture and tone of voice. In other words, the brand sentiment shouldn’t be ignored during the localization process.

Therefore, project owners are recommended to prepare a localization style guide before starting a translation project. It works as a framework for translators to understand the company’s voice, personality, and brand goals. A consistent translation voice can be achieved if all the translators involved in the project follow these guidelines as closely as possible.

Typical elements in the style guide include:

  • Punctuation (spacing, quotation marks)
  • Branding elements (unique to the country or language)
  • Formatting (bolding, fonts, trademarks)
  • Localization or Adaptation (formal vs. informal tone, how to deal with currencies, addresses, phone numbers)

Learn more about localization style guide via this blog post.

Step 7: Prepare Screenshots as Contextual Information

In a poor organized localization project, translators have to work with anonymous strings in a translation tool without looking at the actual product, which increases the chance of contextual errors and excessive project management time.

As the project owner, if you don’t want your translators to do guesswork, you need to provide them with a sufficient amount of reference materials, such as screenshots. And if possible, it’s recommended to let translators use the actual product during the project.

Screenshots allow translators to see their works in context, so they can adjust the text length or translation meaning in the best possible way. The most helpful screenshots should be able to indicate where the associated strings are located on the screen and all necessary elements that may influence the translation process.

If you want to learn more about other ways to make translators’ lives easier, check out this blog post.


It’s true that setting up your pre-translation process the right way might take a little bit of time, but in the end it’s worth it and will produce the type of results that turn into successful localizated products and happy international customers for your company! If you want to learn more about how this can be applied in your own business, I invite you to get a Free Pre-translation Assessment below. We’re here to help!

Image source: Flickr

Written by Vincent Chan

Vincent Chan is the Head of Product at OneSky, focused on making the best localization product possible.