Offshoring considerations

By Willem Stoeller,
PMP and VP for Globalization at WeLocalize

willem.stoeller at Welocalize.com

There is an ongoing debate about the benefits of offshoring services to lower-cost countries. Myself, I believe that offshoring of localization can produce cost benefits, but more importantly, it can provide scalability of localization services.

However, before you start sending your localization to a vendor in India, Russia, or China, it is important to recognize that this will have little impact on one of the largest cost components of your project: translation.

Virtually all of the localization industry uses freelance or in-house professionally trained translators. And typically, a localization vendor will only find local translators for the languages of the vendor’s own region. For all other languages, vendors look for professional translators in their respective countries. Indeed, one of the axioms of professional translation is that you always translate into your mother tongue.

Thus, do not expect lower translation costs from your India vendor for any languages other than the Indic ones. The same goes for a China vendor: do not expect lower translation costs for any languages other than Chinese (all dialects), and maybe Korean.

Project Management – There is quite a shortage of project management talent in India and China. The profession is growing rapidly in those countries, but mainly in IT and software development.
Localization engineering – Staff with only a basic understanding of the target languages can do this well. Native speakers are not essential.
DTP and multimedia – Staff with a basic under standing of the target languages can do this well. Native speakers are not essential.
Internationalization and functional testing – Staff with a basic understanding of the target languages can do this well. Native speakers are not essential.
Localization testing – This requires native speakers for each target language.

From this list, we see that potential cost benefits can be had for localization engineering, DTP, and internationalization and functional testing. I use the term “potential” here because such benefits can be offset from other costs that are inherent in such offshore scenarios: there is a frequent need to audit these services, plus there is still a need for a just-in-time localization solution. This refers to a quick turnaround of last-minute changes and defects. For example, a US company might need last minutes changes turned around quicker than it can be done by a vendor in a time zone 14 hours away.

Cost benefits to offshoring might seem less apparent, but with the right model, the use of offshore resources in a reduced-cost localization effort can still be accomplished. Consider the dynamics among the following types of vendors:

Global localization vendors – These are typically US or Western European vendors with substantial production offices in China or India or both.
Regional vendors – These are companies in China or India that focus only on their respective regional languages.
Local multi-language vendors (MLVs): These are localization companies in China or India that offer localization into languages other than their own.

Several large Western localization providers, such as SDL, Moravia, Lionbridge, Welocalize, etc., have obtained sizable production facilities in India or China. These offshore facilities can be used for the bulk of localization engineering, DTP, multimedia engineering, or functional and internationalization testing. However, small local teams are still needed to audit (sample) the in-country work or to provide quick, last-minute turnaround for US and European clients.

When you combine the costs for these needed local teams with the overhead of distributed project management, you do realize an increase in total cost for offshore localization engineering and other offshore services, but there is still potential for savings, depending on the type and size of the project and the number of hours required for offshore services.

“When you combine the costs for these needed local teams with the overhead of distributed project management, you do realize an increase in total cost...”

Localization testing requires the use of native speakers for the linguistic aspects of localization testing. It is possible to use a distributed team approach for localization testing, where the production facilities in China or India handle the technical aspects of localization testing, and remote native speakers handle the linguistic aspects. For example, capturing screenshots can be done in China or India, but the actual screenshot reviews are done by the remote native speakers.

These are also known as regional single language vendors (SLVs). In India, SLVs are a good source for localization into the Indic languages and Urdu. SLVs in China are a good source for localization into Chinese for PRC, Taiwan, Hong Kong and Singapore.

A drawback of using regional vendors directly is that you, as a buyer, need to interface with a number of different regional vendors; this implies different contacts, different work processes, and different technologies. This is typically only feasible for large localization buyers with their own in-house staff of localization specialists and project managers.

India and, to a lesser degree, China have a number of MLVs that offer localization into many non-local languages. These companies typically do not have the experience of the larger Western MLVs, however. And there are no substantial cost benefits to using these regional MLVs as long as they use in-country, native-speaking, professionally trained translators.

For your larger projects, you can obtain cost savings on localization line items, services such as localization engineering, testing, and DTP, if you use one of the larger Western MLVs that have substantial production facilities in lower-cost areas such as Russia, China, or India. But for smaller localization projects, the overhead of distributed project management outweighs the potential gains.

There is no possibility to offshore professional translation with the goal of reducing translation costs. This can only be done in country (or possibly in North America, where there is a regular influx of immigrants).

Regional SLVs in China and India are a feasible solution for mature buyers with seasoned in-house localization staff. But I hesitate to recommend the MLVs in those regions.

To get the best quality translation, Welocalize works only with professional translators in country. One exception to this rule is Spanish, which is spoken in Spain and most of South and Central America. It is possible to get lower cost translations in the Americas by native speaking, professional translators, as long as pure European (Castilian) Spanish is not a requirement.

A frequent contributor to CSN Magazine, Willem Stoeller is a former professor of localization principles. Currently a VP at Welocalize, Willem still teaches localization topics at industry conferences. CSN recommends Willem’s previous articles regarding the “path of localization.” See the September 2005, January 2006, and May 2006 issues of CSN Magazine.