Jaap van der Meer interviews Rory Cowan and Mark Lancaster
Surprised by the recent shake-up in the translation industry? The SDL acquisition of TRADOS and the Lionbridge acquisition of Bowne Global Solutions. No, it was about time something big happened. Wasn't it? The previous round of mergers and acquisitions did not change very much, except for the cosmetics of creating some fatter players. The same old pre-occupation with translation being an artistic craftsmanship dominated the heart and soul of every language service provider, large and small. That paradigm is now broken down very fast. The industry leaders have spoken.
Firmly back into technology
"You need a certain mass, before the mathematics start to work," says Rory Cowan, CEO and President of Lionbridge, now by far the largest language services company in the world with revenues in the range of 370 million dollars. He thinks back of his many acquisitions of relative small and bigger translation companies like Harvard Translations, Japanese Language Services, International Communications and ILE; and he thinks ahead of the bigger fish still to swallow: Bowne Global Solutions. "We would not have acquired Bowne Global Solutions if we had not had the Logoport technology. Logoport helps us to make this business scalable and manageable."
After having jumped in and out of technology in the past, Rory Cowan is now definitely into technology. "Our previous endeavors with Liontrack workflow and Foreigndesk translation memory stem from a time when every customer and every supplier had conflicting ideas, tools and processes. The technology was not ready. We shied away from further technology investments when we saw the venture capitalists backing up the new GMS sector with huge sums of money. At that time it was best for Lionbridge to concentrate on professional services and work with industry-standard technology. Now the picture has changed. The GMS companies are history. Two hundred and fifty million dollars have been sunk in the sector. TRADOS is the latest victim. We had already given up paying the hefty license fees for each new TRADOS workstation or additional volume of translation work. And many will follow us. Last year it was time for Lionbridge to firmly get back into technology when we acquired Logoport."
Salesforce.com vs. Siebel Systems
Logoport was a tiny small investment for Lionbridge and yet it dominated my conversation with Rory, making the hundred and sixty million dollar acquisition of Bowne Global Solutions almost irrelevant. What makes Logoport so different from for instance TRADOS? "It is Salesforce.com versus Siebel Systems. Logoport does not need to be installed on every translator's and every reviewer's desktop. Translation memory files do not need to travel from the customer to the project manager to the translator and back. Every member of the project team can access the files that she or he is working on through the Internet. All linguistic data are always up to date and centrally stored.
"The Logoport principles help us to cut on translation management overhead. At the same time we improve our value proposition to customers. The scalable and easy to manage Logoport infrastructure convert translation memories and glossaries into linguistic assets that can be leveraged for multiple localization applications in our customer's environment. And most importantly this infrastructure comes without the hefty cost of licenses and complex implementation and training. Logoport is installed on one central server. This means that everyone is always working with the same and the latest version of the software. What a relief after all these years of expensive licenses and never-ending upgrades on tens and hundreds of workstations."
Vision of sharing
The strategic value of Logoport may be less in the short-term savings on licenses and implementation costs and more so in the new perspective of sharing of linguistic data. Who said that localization is proprietary and that the choice of one translation or another for a menu option on your computer screen would change your mind and buy the other system? Just imagine, if we could throw everything in one big bucket and organize our linguistic resources - translation memories and terminology - by industry domains rather than by the limited company confines. Wouldn't everybody benefit? The end-user not in the last place: less confusion in an ever more complex world of digital convergence.
"Definitely, the vision is there. It is a matter of time. Time is changing everything. Remember the days we were concerned about providing credit card details through a web portal? Similarly our concerns about sharing content and translation memories will be fading away. The benefits are becoming too great. When customers ask, we'll be able to quickly assemble unified linguistic databases for industry domains. Once that happens, we'll also be able to train and customize machine translation systems using vast linguistic resources. What our customers really want is just-in-time translation instead of just-in-case translation. Logoport puts us on track to meet these requirements.
"As early as end of 2006 we see the benefits of combining Logoport with the machine translation capabilities in our newly acquired Bowne Global Solutions business coming to fruition. But first things first. Logoport needs refinement and workflow needs to be built around it. And then we must proliferate deployment throughout our global network of more than four thousand staff and a multifold of translators and reviewers."
The ‘craftsmen' strike back
Challenges? Oh yes, there are plenty of them. No matter the amount of progress in our industry in the past ten years, the clinical and process-oriented approach to translation does not always get the support of a hard core of language professionals. Just when you think you have got translation defined as process and nothing but process, the craftsmen strike back and claim it to be a work of intelligent design requiring perhaps not a divine but certainly more human creativity than we had in mind for them. And they are right in their own right. If the profession is downgraded to just cleaning up fuzzy matches from machines, then the resistance is all but natural. Rory Cowan is a long way away from retirement. He has got the mass, but to make the math work, he must overcome a few hurdles.
The Lionbridgers have adopted his vision and understand the principles of outsourcing and off-shoring. Lionbridge is nimble in the western hemisphere and heavy in India where more than half of the 2,200 employees are located. The newly acquired Bowne company works differently. Language management and linguistic quality assurance are managed largely internally in the in-country offices. Integration of the two global operations will inevitably come with fundamental changes, and most likely the Lionbridge way of thinking will prevail.
In the meantime the industry that Lionbridge is arising above, but that it is still part of, will formulate its response. If the industry leaders take their position so clearly and state that technology can help them to grow and prosper, then why would not every small and medium-sized translation company do the same. After all, technology is available to everyone, and even Logoport is not so unique that it can not be developed by someone else. For many customers the small and dedicated providers of language services offer more attraction, especially if translation workflow and linguistic asset management become standard features.
Changing business models
But the biggest hurdle for Rory Cowan is not the internal battle with the Bowne integration issues, nor the external battle with the hundreds of small and medium-sized language service providers, who step in his footprints and make technology the key to their strategies. After all, there is plenty of room for on the one hand a few global players and on the other hand a multitude of regional players. No, the biggest hurdle will be to persuade his customers to trust linguistic asset management fully to Lionbridge.
The growing awareness of the value of linguistic resources makes it less evident for customers to throw all translation memories in one bucket that is owned and controlled by one vendor. Customers who are exploring the opportunities of sharing linguistic data will seek different business models based on mutual reciprocity. In other words, benefits must be shared. The only way to achieve this is by keeping control of language management infrastructure or possibly trusting it to a third party, but not necessarily a language service provider. The biggest challenge for Rory Cowan will be to anticipate the changing business models when sharing and trading of linguistic data become a common thing to do for publishers and translators. Would you not rather be the Siebel Systems of the translation industry when that happens?
Shockwaves through the industry
The two industry leaders have spoken in harmony: technology and process determine where we are going. And yet, the viewpoints of Mark Lancaster, this other daring entrepreneur of the language industry, could not be more different from those of Rory Cowan. Having been questioned about his technology strategy for years Mark Lancaster continues to invest in translation technology to underpin his service operations. His acquisition of TRADOS sent shockwaves through the industry.
The TRADOS company celebrated its twenty years anniversary last year. TRADOS became world famous among buyers and producers of translation around the world. TRADOS is a household name in the localization industry often referred to as a sign of success and prestige by the thousands and thousands of translators and translation agencies making their way into the modern age of L10n and G11n. Despite it being an envious brand, you wonder why would Mark Lancaster, CEO and President of SDL, be interested in buying TRADOS? After all he already owns an impressive range of translation software.
"You know that my roots are in software. I learned about localization when I was doing international marketing for Lotus Development and Ashton-Tate. I look at this industry a bit differently. My interest is not to compete on a micro-level, producing translations at one or two cents less. We like to innovate on an enterprise-level. Instead of translating a lot of content, we like to manage our customers' global assets. This requires investments in enterprise translation software. We at SDL have been doing this for many years. The TRADOS acquisition should be seen as an accelerator in our strategy. Together we have invested some ninety million dollars in translation software development. This beats anyone else in the industry."
Aggregator of language technology
Where Rory Cowan reveals himself as the aggregator of language services, Mark Lancaster can be seen as the aggregator of language technology. The language technology sector is even more fragmented than the language services market. This sub-sector of the GILT industry has led a marginal existence ever since it was incepted by the first machine translation experiment in January 1954. If it was not for the funding from communities and the academic world, language technology would have made very little progress. The general business relevance has been limited to spell checkers and voice recognition and generation.
Translation technology as such - machine translation and translation memory - has never raised much awareness outside the relatively small translation industry. The rise of TRADOS in the last ten years gave hope that language technology would deserve its own space in the global software market. The emergence of the GMS (globalization management software) companies, who focused on the presumed needs for global web content management, added to this optimism. But now this hope seems to be idle.
Systran, Idiom, Across and Language Weaver are the only translation technology companies with the outlook of a more or less independent existence. The translation technology sector does not attract many customers wanting to pay for licenses. Rumors were that TRADOS was not able to make the transition from desktop to enterprise sales. In the late nineties the company invested millions in the development of an innovative hosted service model for the global translation industry. This ambition collapsed even before the Internet bubble burst. The Uniscape acquisition was TRADOS' own attempt to aggregate the GMS sector, but realizing the benefits was not easy. Perhaps it was not all that glittering from the inside.
Growing sales in enterprise licenses
Being king in a sector that is small and weak.... It makes you wonder again why SDL would go down this route. Rory Cowan may be perfectly right saying that the GMS sector is history and that no customer likes to invest in this heavy infrastructure. Or is there something that we overlook?
"Let's first set things right here about TRADOS," says Mark Lancaster. "It would fit me well to say that they failed and that SDL did everything right and that that is the reason we end up owning TRADOS. But that is simply not true. TRADOS has shown a 50% year-on-year growth consistently. They did sign on many enterprise customers and the number of orders of the new TeamWorks translation workflow system in the last eight months proves that there is a market for translation software.
"But again, both TRADOS and SDL needed more volume of business to be able to afford the ongoing development and support of global customers. The combined companies will be better and more efficient in research and development, support and integration and in training users. That is the simple truth behind our acquisition of TRADOS."
Global information management solutions
With some eighty people in development SDL is clearly the leader in the language technology sector, a fact that will be appreciated by investors who value technology companies usually twice as high as service companies. But what about SDL's customers? Will they be able to recognize the value that Mark is building for them?
"No matter where the customer is coming from, he will benefit from our strong technology advantage. Sometimes they will buy the software licenses to do their own thing. Sometimes they will ask us to build a solution and operate it for them. And quite often they will simply buy the translated words from us. In all these scenarios the technology benefits both us and the customer. Look, every translation agency has some kind of a portal these days. But none of them have the wealth of automation as we do. Tools like SDLAuthorAssistant and SDLPhraseFinder may be small products, but they make a huge difference to our customers. We can build end-to-end global information management solutions that help our customers to win time, translate more and save costs."
Standard pricing
After the acquisition of TRADOS, SDL can offer a choice of workflow, translation memory and terminology management systems. In addition the company offers an authoring system and machine translation for a limited number of language pairs. The machine translation is probably the weak spot in the tools portfolio, because it is very costly to develop new language pairs. Where SDL is pushing its KbTS (Knowledge Based Translation System) to the market as the all-encompassing solution - combining MT, TM, terminology and workflow - it will most likely start shopping for external MT systems. Other than that SDL has got what it takes. The temptation is there for SDL to keep it for themselves and create an unbeatable competitive edge. Would that perhaps be a hidden strategy?
"Oh no, we would much rather sell the enterprise licenses than the translated words. Our marketing strategy is focused on selling enterprise licenses and then possibly be part of the delivery of support and language services. We will soon come out with a standard price for TeamWorks translation workflow for language service providers. So, rather than creating barriers, we will make it easier for every translation company to work with our technology. Besides, if we don't offer the tools, there will be other software companies who will."
Thinking out of the box
Glancing over the case descriptions of some of SDL's customers, we must admit that this company has a talent for thinking out of the box. When Best Western hotels knocked on their door with a more than 20 million words translation requirement, SDL immediately realized that the conventional translation process would be too time-consuming and too costly.
They built a highly automated and dedicated solution for Best Western using a mix of MT, TM and post-editing. A similar solution was developed for CNH, the global manufacturer of agricultural machines, to translate technical support information at minimal cost and very fast. These types of solutions bring the cost per word down to a few cents perhaps. If you are a translation company, you would call this ‘cannibalizing your business'. But SDL is not a translation company. Or are they?
This may be the battle for Mark Lancaster, even more so after the TRADOS acquisition, to find the right balance between services and technology. The challenge for the SDL sales organization is to understand what the customer really wants. The temptation will be to have every customer standardize on SDL technology and buy SDL services for all language needs. But most customers will prefer the freedom of choice. An overzealous SDL sales team may benefit the competition more than SDL.
New inspiration
New inspiration is what the localization industry needed and this is what the news of the SDL-TRADOS marriage has given. Technology is the name of the game. All translation companies will be striving to make themselves equal to SDL in their offerings of efficient processes. They may simply buy the workflow systems from SDL or buy it from other sources. ‘Other sources' may have their preference: who likes to buy from a competitor. Idiom has already announced their LSP Advantage Program giving translation companies access to their World Server technology at no cost.
The merger of the world's leading translation memory and workflow systems will inspire new competitors. The loyalty to TRADOS exists primarily in the translators and agency community. Will Mark be able to maintain and build this loyalty? Chances are that he will be losing some of this over time, especially when the TRADOS brand is slowly disappearing. Surely he is aware of that but he bets that he will gain more from the enterprise customers than he may lose from the traditional customers.
