With online retail continuing to expand globally and an ever increasing range of industries being served online, many online retailers are looking for ways to overcome barriers of countries and languages in order to expand their presence globally and to increase their revenue stream. This can either be done through aggregators such as Amazon, eBay, Lazada, Alibaba, Rakuten etc. or alternatively retailers can directly or indirectly set up an online presence in the target markets of their choice. Whatever the chosen mechanism, content localization becomes an immediate need and issue. However, unlike traditional localization projects, e-commerce content is typically not economically feasible to be translated by humans because the cost would be prohibitive and the time required too long. Therefore, machine translation is often the only viable option. Some online retailers turn to cloud operators, hoping for them to supply this service, however, sometimes with disappointing results.
This is by no means to say that the cloud operators and free online translation machines do a poor job, but at large the fundamental challenges related to translating e-commerce contents are misunderstood and underestimated. The cloud operators’ aim is to provide a “gist” translation for understanding the content rather than to provide an output with publication quality. However, this is what online retailers would need in order to enable a buying decision. That being the case, a more sophisticated process needs to be employed.
This paper discusses the challenges related to the localization of e-commerce content and offers an approach as well as solutions for employing Machine Translation in content localization for this domain. The approach is based on extensive experience gained by Omniscien Technologies from projects executed for a range of global e-commerce clients.
Read more by downloading our white paper “Challenges and Approaches to Online Retail Content Localization”