Sharing Data Across Companies – is it the Holy Grail?
Some of you may have been reading up on the idea of sharing the translated data that you have within your organization, so that other companies can leverage from your data and vice versa. What you probably need to look at before this is the bigger picture of your translation costs and see how you can do things in your business that will bring about other benefits such as productivity and time-to-market improvements:
- Centralize the process you have for storing your own translation memories – bring all your translation memories that are scattered across your business and that are used across different translation agencies and bring them into one. We have seen companies get anywhere between 60% and 90% reuse of their own translated content through doing this. At Philips, for example, operational costs for localization/translation have decreased by 70%, customer satisfaction has increased and time-to-market has been cut dramatically, all through the implementation of centralized leveraging of linguistic assets
- At the same time as saving on centralizing, you will save in productivity by automating many of the manual tasks involved in localization. There are translation management systems that do exactly this for many global businesses such as Dell, FICO, Philips, Renault and many more
- Look to improving the authoring processes and ensure consistency of style, which will reduce costs in the localization process later on. At the same time you can have authors see what has been translated before and you can ensure they don’t make unnecessary changes that will push up costs later on
- Ensure the terminology used in both authoring and translation is consistent to your brand and in every other language and that the people doing the translation for you are translating those terms consistently every time
- Look to automated translation – companies reduce the costs of translation by 25-50% by using this as part of a human process for translation. Again if you want to use a statistical process, you will benefit from sharing data if you have no translated content of your own. You will gain less if you have high volumes of your own, personalized content that has been translated
- Go wider than your own department – look at the processes used in marketing, sales, training, support, technical documentation, engineering and bring them all together into one unified strategy that reuses the same content
Once you have done all of this across your organization, you will have done the maximum to leverage the best and most relevant data – your own.

How to put yourself out of work by another automatic system which replaces people with machines!!!
Posted by: Eileen Brophy | 12/15/2009 at 12:18 AM
I have used Trados for years and continue to do so, but would NEVER consider sharing my ALL my TM with others on a general basis. After all, why have translators, when you have all terminology and phrases translated and online to be freely accessed? Why did I pay for the programme? To put myself out of work? That was not the exact reason at all, I wanted to be faster and more accurate, nothing more, simple professionalism.
Posted by: Eileen Brophy | 12/15/2009 at 12:22 AM
It's absolutely not in the interest of translators.
That means having your TM compiled by some software company so they can improve their automatic translation software.
Besides, whatever you call it, a monopoly remains a monopoly and there are lwas against trusts.
Posted by: DAVID MAROTE | 12/15/2009 at 08:02 AM
Without proper review, the re-used materials will be obsoleted. Having those translation memories from different context and organization compiled into one can be dangerous if you are ignorant..
Posted by: ade indarta | 12/15/2009 at 10:22 AM
Thanks all of you for your comments on this and I have seen some other blogs written in response. So it is clearly an interesting discussion point on what the quality can be, what is the impact on translators etc. As per my blog article, I think there is a lot that companies can do to improve internal efficiencies and leverage their own data before looking at sharing their data. It can also depend on the amount of translated data a company has as to how much value this will add.
All good discussion and time will tell how much organizations decide to adopt this and how it impacts the industry. As a point of clarification by the way, SDL is not planning on sharing its customers' data - I wished to merely open the discussion on the topic. SDL considers the translation memories of its customers to be their own intellectual property and therefore it is up to them as to how they wish to use it.
Posted by: Sophie Hurst | 12/16/2009 at 12:09 PM
Centralize the process you have for storing your own translation memories – bring all your translation memories that are scattered across your business and that are used across different translation agencies and bring them into one....Thats is the best thing to do !
Posted by: cheap computers canada | 01/09/2010 at 08:49 PM
This is a great way of getting better leverage and therefore cut costs and reduce time-to-market. However, it is not only a benefit for the business sending the work, but also for the translators, who get to benefit from previously translated content, sometimes key in understanding new translations or maintaining the style when you are working on several types of files from the same client. Of course, proper review of the leveraged content during each translation cycle and maintenance of the TM is key to ensure the quality is not affected and consistency in terms of style and terminology is maintained.
Posted by: Ana Arias | 02/10/2010 at 01:21 PM