Translation Proxy Basics, Part 1


Although the technology has been around for a while, there are many misconceptions about it in the language professionals’ community. This is why we decided to go back to the basics and clarify what translation proxies are – and what they are not, as promised in our earlier article about the 3 ways of website localization.

What is a translation proxy, anyway?

In non-tech terms, it’s a seamless translation layer on the top of the customer’s website displaying content in the visitor’s native language from the translation memory (TM) on the fly. The translation itself can be stored in the cloud, or in a dedicated server of the LSP or the Customer that is connected to the internet.

The content of the translation memory comes from human translation; the translator may work directly in the actual layout of the website, or translate the content with their preffered CAT tool, and upload the translation to the TM. (Of course, the translation memory could also be created from machine translation engine, e.g., for quick presentation purposes.)

Translation Proxy Schematic Illustration by Seth Gottlieb

Translation Proxy – Illustration by Seth Gottlieb

As you can see in this illustration, the source language content sits on the customer’s web server. The translation (but not the whole site) “floats” in the Translation Proxy “cloud”.

When a visitor visits the site from the country of the translation target language, he will see the customer’s website with the translation “floating” on top of it.


Why is this a big deal?
What’s in it for the translator and the content owner?

Remember when getting a website meant that you needed to hire a PHP, Java, HTML programmer and you had to be familiar with HTML to update or add anything? This misery got resolved as user-friendly content management platforms like WordPress, Joomla, Typo3,Drupal and others entered the market almost a decade ago. They removed all the IT frictions, so now anyone can have a website within a day and maintain it without an IT expert’s help.

With the appearance of the translation proxy, something similar is happening in the website translation field. Since the content are not in files anymore, but in content management systems and databases, to extract it and put back the translation involves heavy IT resources. Not anymore. With Translation Proxy Solutions the heavy involvement of IT expertise is not a requirement for website translation.

Let’s review the steps of the traditional process of website translation:

1. Assessing the scale of the project and preparing a quote
To provide a quote, translators need to assess the amount of content, i.e., count the words. But in case of websites, this is not so simple. Websites are built from databases, not static HTML files. It is nearly impossible to explore all content from the browser view of the website. What if it’s an e-commerce site with thousands of dynamic pages? Ok, let’s think. It’s simple: the customer gives the translator direct access to the content management system (CMS). In this case, the translator has to learn to work around the tweaks of the content management system.
Or, the translator may insist that the customer provides the files to translate, someone at the company will do the copy/pasting. Even for a large firm with dedicated IT resources this may involve unacceptable additional time and cost, not mentioning small- and medium-sized companies with limited IT resources. This additional cost could be a show stopper of website localization.

 2. At last! The translator gave a quote and the parties agreed, and the source material gets somehow exported to XLIFF. Off to the translation management software, and you can do the translation. Surely someone at the company will gladly import the translation into the CMS, or copy/pastes it manually. It’s quick, easy and fun, isn’t it?

3. It’s all done and well. Unless one of the international clients contacts the company and lets them know that some translations are out of context, sounds funny, and others simply don’t fit into the layout. Then back to reviewing the whole site, and fixing anything erratic again.

4. Wait, this content is not even valid any more.
And the re-iteration of the whole cycle goes on and on.

From the translator’s and the company’s point of view, this process is a hassle. At global corporations, there are whole departments handling project management of website translation, and technical staff handling content export/import tasks. It takes time, personnel and overhead.

Translation proxy-based tools simplify these steps by eliminating the IT effort from both the customer’s and the translator’s side.

In Part 2, we’ll take a look at the how the translation proxy-based approach can improve the old process.

Dedicated demo for You