Processes surrounding translation and localization are vast, complex and can frankly be confusing to companies buying these services. Many purchasers of translation products feel as challenged by it just like when taking a car to a mechanic. How do you really know what is going on underneath the hood? After all, if you don’t speak the “language”, how can you objectively and quantifiably measure quality and hold translation vendors accountable?
As a result, many clients rely on tactics they feel will get them that measurement of quality, but actually more often than not these get in the way of ensuring the optimal quality. Some of those misconceptions related to translation are:
1. A Linguist Is All That Is Needed
Any writer, even the best professional writer, relies on editors, proofreaders, and others to make their work come alive on the page. Likewise, a professional translation process also involves various resources. Furthermore, translation projects are more often complex involving multiple languages with varying content types, or even file formats. Partnering with an expert agency such as inWhatLanguage (iWL) is a better strategy and solution.
2. More Translators = Better Quality
More translators = better quality. Over time, a dedicated language expert becomes intimately familiar with the writing styles, tone, and messaging of a client’s content. Linguists are not interchangeable. Generally, if the same translator is not used repeatedly and consistently on projects, consistency will decrease, and the translations actually will feature different voices and styles. This impact is further acerbated when using a
team of linguists -- each will have varying experiences, backgrounds and qualifications.
3. Retaining A Second Provider & Using Them Against The Primary Translation Supplier Keeps Quality In Check
Many users of translation services believe it is incumbent to hire one agency (translator) to translate their content, and then have a second provider (reviewer) check the translator’s work for errors. Honestly, this is a recipe for disaster. First, the focus of the reviewer becomes “error detection.” In order to prove the reviewer is doing a good job, the service provider will flag as many “errors” as they can find. The result is very biased and extremely unreliable because
many of the suggested changes are really preferential. The disposition of the reviewer is hoping that if enough mistakes are reported, the translation agency will be dismissed and the second agency will be rewarded with the translation work. Second, the client ends up in a “death spiral of time”, endlessly mediating between the two parties, when ultimately many of the “errors” boil down to one person’s opinion versus another’s. Lastly, the entire process becomes combative instead of collaborative. The primary objective of providing the best translation product for the client is no longer.
4. "Back Translation" = Quality
There is a believe that you can measure quality by doing a reverse or back translation. A project is sent out for translation into a target language. Then, the completed translation is translated back into the original language (by either the same vendor or a secondary company [blind test]). A compare of the two versions to see how similar the produced pieces are is perceived as the ultimate measure of quality.
The assumption is that errors can be spotted by looking at these two versions side by side. This is a failing task because errors can be introduced at any point in the process. If there are mistakes in the back translation, there will indeed be differences between the source and the back-translated version, but the client will not be able to accurately ascertain the source of the errors.
5. Client Employees With Knowledge Of A Second Language (or even bilingual) Provide Reliable Quality Feedback
An often-seen method to measure quality is to ask a favor of a bilingual co-worker or employee by “taking a look” at the translation of content. In reality, this is a perilous endeavor, the feedback is not always relevant or helpful. This person also is unlikely to be an expert in every single area of the content. The only way such employees may provide insightful commentary on the translation quality is if he/she is given specific and focused guidance on what types of things to look for. Unfortunately, most often the question to put to him/her is, “Can you read this and tell me what you think?”
6. The Source Content Has No Impact On Quality
A large percentage of perceived translation errors are directly attributable to source text that may be inadequately written, is unclear or even lacks proper reference and context. It is written for the perspective of the locale and language use of the author. It rarely conforms to a global audience or other foreign locale. When a sentence can be understood in more than one way, the linguist has to make an educated decision (sometimes even a guess) regarding the author’s original intent.
Typically, linguists do not even have the opportunity to clarify the source text to find out what the intention was behind an ambiguous term or meaning. They rely on their research skills and professional work experience in an attempt to figure out the proper meaning. This is obviously not desirable, and may lead to a translation that does not measure up. When the messaging of the source is not clear, the translation output will also not be clear.
What is the takeaway then of all these misconceptions out there? Again, translation quality is not a simple topic. It is a highly subjective discussion.
Let us reference our previous examples of the writer and the mechanic and measuring the quality… Who is the best writer? The answer varies, depending on the writing genre and who the reader is. Who is the best mechanic? This is again difficult for a layperson to judge.
So, let us focus on other performance metrics so we can evaluate our mechanic: How the car runs, how many return visits to the mechanic it requires, how often it breaks down, etc. Just as it is challenging for the average driver/person to understand complicated technical and mechanical car diagnostics; the same holds true of translation. In order to properly understand quality at a deep and proficient level it takes much more than a certain warranted proficiency in both the source language and target dialect.
The best indicator of translation quality, especially for non-language experts, may have nothing to do with typos, misspellings and grammar. From a client’s perspective, a good translation has more to do with quantifiable results related to brand awareness for instance: more customers, more page views, more downloads, more sales, and ultimately more revenue. More simply put Return On Investment (ROI) and Return On Engagement (REO).
The next question is then: How do we get there? Continue reading How To Measure & Ensure Translation Quality - Part 2: Translation Quality
Written By Maurice Van Zutphen, COO
Maurice is a veteran in the translation industry with nearly 25 years helping organizations with their high quality translations. Maurice is an expert in e-learning solutions and very technical translation projects. Having managed and overseen literally thousands of translations at the highest levels, (including highly confidential projects for the White House), there is no one better to look over your translation projects and manage the operations. And in case you are wondering we call Maurice “the flying Dutchman.”