Balancing Time, Quality & Cost of Translation

Last Updated January 14, 2015

I recently put this simple chart together to explain in a graphical way, my own understanding of the relationship between cost, throughput time, and quality across the translation continuum. This can help you when thinking about your translation priorities.

A few colleagues asked me to blog about it. So, here we go…

Translation Priorities: Expectation vs. Reality

This chart shows that different approaches yield different results in terms of price point (cost per word), quality (fluency and adequacy of the translated text), and speed (words translated per day), moving along what I call the Translation Continuum.

Clearly, you shouldn’t focus too much on the numbers; they’re mostly an order of magnitude. Prices and volume per day tend to vary greatly over time and across geographies. Working as a US-based freelance translator (English into French or English into Spanish) over 15 years ago, I was able to charge $0.30/word or $50/hour when translating “challenging” source content. In recent years I have had customers telling me they can get $0.02/word with some unknown vendor, to which my response has been, “if they’re good, please send me their contact info, so I can hire them – but I won’t try to compete with them on price only”. In my experience, on average across multiple languages, and on aggregate, taking into account TM leverage, it’s possible to achieve a price per word of approx $0.08 per word – especially when leveraging large translation memories or taking advantage of machine translation capabilities.

The Power of Crowdsourcing

When adding crowdsourcing to an already broad range of options, it becomes possible to offer a tailored solution that meets almost any localization need or constraint, by combining approaches as needed or selecting a single method, when relevant or appropriate. Highly visible content of technical or marketing nature can be localized using professional translators, a slow and expensive process that typically provides the highest quality. On the other end of the spectrum, machine translation can be used to process large knowledge bases, hundreds of support articles, or constantly growing user-generated content on social media: All this represents billions of word that would otherwise never get translated due to budget and time limitations. Technology allows us to harness the power of the crowd, even of a specific or specialized crowd, to translate large volume of content in a quick and somewhat economical manner. Post-editing ensures that questionable MT output is reviewed and improved, increasing the usability of the resulting translation.

Achieving Low Price, High Quality Translation

At the end of the day, when speed of execution is the main factor, MT has to be part of the solution; it’s almost a no-brainer. If “perfection” is required, a professional translator or editor will need to work on your content. A combination of MT plus post-editing and crowdsourcing can help achieve lower price points without sacrificing quality or throughput time. Yet the most effective way to lower localization costs is quite simple: Tell your content writers to use clear, concise, and grammatically correct language.

Related Posts

Summa Linguae uses cookies to allow us to better understand how the site is used. By continuing to use this site, you consent to this policy.

Learn More