In view of The Language Technology Industry Summit 2017, we have spoken with Massimo Ghislandi, Executive Vice President, Translation Productivity at SDL about some of the trends and challenges facing the translation and localisation industry:
Philippe Wacker, Secretary General, LT-Innovate: SDL has reorganised in the past years and focussed more centrally on translation. Where do the principal challenges come from in today's business and technology world for a large LSP?
Massimo Ghislandi: There has been a marked change in the last few years, as highlighted in some of our industry research. First of all, volumes have grown. This has not come as a surprise; the digital content explosion has been happening all around us for some time. Secondly, the type of content, and more importantly how that content is sent for translation has changed. Today, content is authored and sent for translation in a more agile manner. More and more content is created in components / chunks (also known as micro translations) and sent for translation. We have seen this trend ourselves and so too has the industry; our research tells us that there has been a 70% increase in chunking in the last 5 years. The result for large LSPs is that they are receiving smaller chunks of content but at a much faster pace. This in turns forces LSPs to be much more efficient when it comes to the translation process.
Past ‘Translation Industry’ common wisdom indicated that in a project 50% of the overall project is translation, the rest is project management. Increasing the number of projects, or files within a project, but not necessarily the overall word count can impact that ratio, and LSPs might find themselves in a situation where project management goes beyond 50%, to 60-70 or even 80%. This is due to the overheads caused by the myriads of small files which need to be prepared, sent for translation, tracked and reviewed. As a result maximizing process efficiency becomes essential; to ensure margins are retained, quality not impacted and translator satisfaction is also maintained.
Philippe Wacker: In addition to your legacy technology offering, what new kinds of translation technology will you be marketing to individual/small translator players in the coming years? Is this a growth business or does it simply remain stable? Could you expect to acquire technology by buying up a competing start-up?
Massimo Ghislandi: Up to now, sharing of translation assets (Translation files, memories and terminology) was predominantly only feasible for larger corporations and LSPs. We have been working on ways to make sharing more accessible to a wider audience.
We have a “GroupShare Cloud” option which is ideal for smaller set-up for 5-10 concurrent users, paid on a yearly fee, with no installation or hardware required. This is ideal to streamline project management tasks and minimize upfront investment.
We have also recently launched online services to enable sharing of translation assets. Our beta cloud Terminology sharing is a good example. This is now live and people can collaborate and share terminology via the cloud. This will enable even smaller groups to gain the project and process management efficiency that a centralized stored set of assets can offer, and at an affordable price point.
We at SDL are committed to delivering innovative ideas and solutions. Offering new services and solution to our existing customer base is one of our key focus areas, but we are also looking to introduce a wider audience to translation technology through initiatives such as SDL’s online translation editor. This has been designed for more casual translators, who might be looking for a tool to help them with ad-hoc translations, but do not require the functionality that we see in professional software.
Lastly, on the subject of acquisitions, we have in the past acquired a number of companies to enhance and extend our offering and we will continue to do so, where appropriate. That said, we are well positioned internally to deliver innovative ideas like, AdaptiveMT, our self-learning MT that was developed internally and delivered as part of the launch of SDL Trados Studio 2017.
Philippe Wacker: Which piece of language technology are you waiting for to be designed and built(!) ?
Massimo Ghislandi: Artificial Intelligence and Neural networks are two areas that could impact translation technology over the next few years. We do not believe machines will replace translators in the foreseeable future, but some of the jobs might change. How translators interact with technology might indeed change over the upcoming years.
Philippe Wacker: Which (new?) verticals do you think will draw the most business for translation in the coming 5 years?
Massimo Ghislandi: It is difficult to tell. Companies and governments ultimately translate for two reasons: if they have to or if they want to.
Regulations have always been a driver for translation. So pharma, legal and financial sectors all remain key growth areas, as more companies expand globally and need to provide content in the local language.
Entering new markets and the ability to connect with the local audience is also a must for many companies. As a result of this we are seeing the emergence of new language pairs in traditional markets, like high tech and manufacturing, where companies are trying to grow in new geographies and new language combinations emerge e.g. in Asia, where English tended to be the main pivotal language, but today we are seeing more Asian-Asian language combination being translated.
Philippe Wacker: The role of human translators appears to be changing: some insist on a central place for human translation at least in commercial translation in future, others see translation merging into post-editing. What is for you the most effective configuration of humans and technology in the translation pipeline?
Massimo Ghislandi: This is difficult to predict and there are some variables that complicate the picture. First of all it is still difficult to predict how Machine Translation (MT) will evolve, meaning the role of the translator will continue to be essential. In some cases the way of working could change, especially where MT could offer more intelligent support, such as becoming a smart phrase dictionary that will help the translator. However there are also language combinations where machine translation still provides poor results, meaning changes in ways of working will be minimal or non-existent in the near term.
As I see it there are two strands: one where translators are working with technology to accelerate the more monotonous part of translation, such as repetitive simple sentences or looking up potential translation in reference materials. The other sees translators continuing to rely on translation memory for more complex language combination or for more complex content.
Philippe Wacker: What needs to change in the industry, or in the broader business world, (regulations, standards, training, accounting, methods, cost structures, etc.) for you to achieve your most significant objectives as a company?
Massimo Ghislandi: Ideally the more translation becomes a must, the easier it becomes for us as a company. Even today translation remains an afterthought in many cases, and something not central in the content strategy of many organizations.
Although perhaps not always the most positive message – but if regulations forced translation of content into more languages this would certainly help. Our most recent research study found that a quarter of respondents still do not have an allocated budget for translation – this shows how there is still a lot of work to be done in making sure that translation is central to content and globalization strategies.
Philippe Wacker: What kind of support would you like to see from the European Union in this field?
Massimo Ghislandi: Continuing to spread the message about the importance of translation is key. So too is the continued support for EU companies who are looking to go international / global. Both of which are good for trade, whilst simultaneously driving translation demand.
Regulations to ensure important content is translated in as many languages as possible would also help - both us and those people within organizations looking to change attitudes towards translation.
