A Necessary Evil

In the couple of short years in which localization has gone from a handful of cottages in Ireland to a global and in a number of cases a publicly listed industry, its translation aspects have not been the center of attention. “Strategic” aspects such as making money, mergers and acquisitions, employee recruitment, project management and the Internet revolution all commanded much greater airspace at industry gatherings. This omission, though perhaps understandable in an industry metaphorically trying on its first adult suit of clothes, is nevertheless surprising given the fact that translation is the largest single budget item in localization projects.

What is more, where translation was actually addressed, it was regarded as definitely problematic—a source of rationalization efforts by large client corporations and a bugbear for service providers. As Claudio Pinkus, the former CEO of global service provider Bowne Global Solutions, put it at the LISA Budapest Forum in December 1999, “translation means adding one person for every 2,000 words per day, and this is not the way to add value”.

The logical business model for such a low-value, non-core activity was generally felt by both clients and service providers alike to be a mixture of automation and outsourcing. In practice “daisy-chain” structures involving multiple intermediaries evolved to handle the supply side of the market. Thus a client would contract with a large international localization agency to localize an application into numerous languages. In turn, the agency would contract with a lead subcontractor for each language pair, who in turn subcontracted further down the chain. It was (and still is) not infrequent for the ultimate translator to be seven or more links removed from the software manufacturer, and it is not uncommon in the lower echelons for price to be the only driving factor behind supplier selection. At a macro level, the effect of this on a profession as heavily fragmented and individualistic as translation itself started to look like a self-fulfilling prophecy, with much of the traditional low end being sucked into the downward price spiral.

The introduction of translation re-use technologies, though driven as much by the need to ensure the repeatability and reproducibility of original content as by cost-related considerations, also had a similar effect. By—naturally enough—emphasizing the primacy of existing text and inserting a further layer of technology as an intermediary, translation memory tended to remove translators and the creative process of translation from the center of events. This further reinforced the perceived drop in value added. Other language technologies, such as machine translation, have not presented the same problem in such an acute form, since the human input needed—except in highly controlled environments that have generally themselves been the recipients of man years of consulting effort—and hence the value added is much more immediately apparent.

To sum up, therefore, traditional localization companies generally adopted a highly paradoxical attitude to translation. On the one hand, they regarded it as a commodity product, with reproducibility and repeatability as key criteria. In other words, translation was a substitution-driven activity in which individual translators or corporate translation providers were—in theory at least—effortlessly interchangeable. On the other, it was also regarded as a high-maintenance, non-scaleable and irksome activity, to be outsourced if possible. In both cases, translation was a necessary evil rather than a differentiating factor, and having the words “translator” on your business card—if indeed you had one—was not exactly regarded as the stepping stone to a brilliant corporate career. (This despite the further paradox that many localization companies actually made a large proportion of their money from translation).

Things Fall Apart

Much of the blame for this development can, however, be laid firmly at the door of the translation community itself. Accustomed over decades to almost mediaeval working practices and a self-image that sought to set translators apart from such “unworthy” matters as business and customer orientation, the translation profession in the late 1980s and early 1990s was a classic case of an industry largely blind, or reacting wrongly to, the changes happening around it. The paradigm of the translator as artist (i.e. only minimally a scientist and definitely not a business person) was extremely well rooted—and in fact actively encouraged by some academic institutions and professional associations, which spent more time in internecine warfare than in equipping their (admittedly often extremely recalcitrant) members for the new world order.

Against this backdrop, the impact of localization (and of automation in general) was to split the industry into a low end and a high end, as well as causing a big shake-out among traditional proponents of the latter. Unwilling to adopt fast-moving information and communication technologies and facing growing time pressures, among other things, many specialist translators simply first burned out and then dropped out of the business altogether, or turned to areas of work that were less affected, at least temporarily, by the pace of the new developments.

Those translators who avoided the localization trap in the 1990s but who nevertheless enthusiastically embraced computing, the Internet and language technologies started redefining the role and the image of the translator. The new high-end specialists are multilingual content providers, offering business solutions rather than “mere” translations. In fact, many of them also started adding consulting services to their offering, leveraging their know-how and adding value to both their own and their clients’ processes.

All this will sound remarkably familiar to observers of the current round of repositioning on the localization scene. One version of the “high end” is represented by the technically sophisticated global service providers, which offer global reach and economies of scale and are increasingly taking on the role of facilitators. To be able to compete with these often cash-rich companies on something approximating equal ground, and to escape the trap of working for “pennies for words”, many small and medium-sized localization vendors are repositioning themselves as specialty providers. There are many different versions of such specialties, including geographical, technological, and/or domain expertise, as well as unique combinations of services and target environment (e.g. Web site) know-how.

Other factors influencing this redefinition are the fact that as we now automatically get “chips with everything”, the boundaries between the localization and straight translation models are inevitably becoming blurred, and the traditional product release model is breaking down in the wake of the Internet. While overall translation volumes (and hence senior management attention) are continuing to increase, the multiple target audiences and the move towards mass customization are leading to a greater emphasis on content adaptation and personalization rather than “commoditized” translation. Equally, product and content liability (and hence translation liability) is becoming more of an issue for clients, and hence for service providers. And, last but not least, as the Internet tears down entry thresholds and markets become more competitive, quality is emerging as a differentiating factor for both clients and service providers.

Process and Content

What, though, is this quality that everyone is now talking about? It has long been the case in the translation industry that, while there is a general consensus that quality is something we all need, defining quality—and translation quality in particular—is a far more difficult task.

The deleterious effects on translation quality of the commodity model have certainly been identified as an issue. Thus Claude Pesquet, the former Digital Equipment senior executive and LISA Board member demonstrated in Budapest that the French version of Visual Basic 6.0 contained a large number of “anomalies” and called for a change of heart: according to him, “being ashamed of specializing in translation is wrong”. Nevertheless, the dominant mood among clients and service providers alike still seems to be fatalistic. To quote Claudio Pinkus in Budapest again: “the problem is that translations are so often wrong, and there is always a risk no matter how much effort you put into getting them right”.

Of course, it is clear that in a localization context “quality” will refer not just to translation but to the quality of the finished product as a whole (and the renewed focus of the past few years on software testing is no accident in this respect). It is also clear that, in the real world, cost and time to market, resource availability, and expertise and experience can all combine to relativize quality. Nevertheless, the more the industry moves to a specialist model, the more issues of both process quality and content (or output) quality will continue to rise on the radar screen.

The first area is the one in which the most progress has been made. Standards such as ISO 9000 ff., the LISA QA Model and various proprietary enterprise quality systems are designed to ensure that the workflows and processes involved in localization are optimized at the level of organizational units and individual procedures. Adoption of such quality standards in recent years has done much to professionalize and streamline the localization industry, and hence indirectly to facilitate the global expansion of the IT industry. The implemented process quality requirements behind the simship of a major software product, for example, are something of which localizers can rightly be proud. What is more, on the vendor side, such levels of process sophistication have helped ensure corporate survival in a heavily competitive environment and a market dominated by growing volumes and ever-shorter deadlines.

However, much less progress has been made in the area of output quality, even if the incidence of downright howlers has now declined with the spread of basic PC literacy and the increasing use of style guides and standard terminologies. The latter are, incidentally, common ways of trying to build in a certain degree of quality from the start, and other standards such as the LISA QA Model have also addressed this issue in passing. Nevertheless, the more dominant focus of work up to the present has been on various ex post methodologies of analyzing and evaluating output quality. This is perhaps understandable given the pressing need to establish and agree on “objective” criteria for translation quality before doing anything else, and the difficulty of the issues involved. However, such approaches assume—or at least can facilitate the assumption—that the (translation) content received for QA is likely to be flawed in some way, and that the QA process is the primary stage for identifying and rectifying errors.

The Quality Challenge

Nevertheless, designing in quality from source is precisely the challenge now facing the high-end sector of the translation/localization industry. In today’s fast-moving business world, the old triangular model of “quality/price/deadline – pick any two” that service providers used to show to clients has now been replaced by “consistent quality, value-added and time-to-market”, with all three factors an absolute must. This means that the only chance to achieve translation quality is to get it right first time, not to build it in at a later point.

However, this has significant implications for both translation/localization companies and individual translators. In particular, a number of not insubstantial challenges need to be addressed.

• Productivity
  At Budapest, Claudio Pinkus also said that translation means “adding one person for every 2,000 words per day, and this is not the way to create value.” Within the logic of his equation, he is quite right. Despite the move to a value added (and hence not exclusively price driven) model, in today’s fast-moving environment, translators will need to accomplish often much more, and on a regular basis. This requires a change in the “quota-based” mentality sometimes visible in the translation profession (itself often a byproduct of or reinforced by a commodity-based model). Obviously, language and other technologies that give translators the tools they need to do their job have a key role to play here, as do the greater speed and assurance that come with sector specialization. However, translator training and translator profiles also need to be rethought, in some areas substantially (see below).
  
• Translator training
  To use a somewhat old-fashioned military metaphor, new translators, like any other recruits, need to acquire all the skills they need for survival during their basic training. While many universities have made substantial progress over the past few years to modernize their courses, the gap between the curricula and real life is still too large for everyone’s comfort in too many cases. In addition, many universities in turn deplore the basic native language skills (starting with but not confined to grammar) exhibited by school leavers, and in at least one case a remedial program has been set up (the title of which was carefully disguised to avoid falling foul of university regulations on the purpose of tertiary education institutions).

  In addition, much greater sectoral and technological expertise will be required in future. Developing corporate knowledge bases and other forms of IT support is one way of shortening the learning curve here and leveraging existing human and written resources, if still a labor-intensive one. In a knowledge-driven economy, it’s not a matter of having all the knowledge yourself: it’s about knowing where to find it and how to integrate the people who have it. The new breed of translation/content adaptation providers, who have invested heavily in recent years in domain expertise, training and technology, are well positioned to fill this gap. Nevertheless, ongoing training and knowledge collation and dissemination will continue to represent a not inconsiderable effort for smaller specialist shops in particular.

  What is more, the need to transfer knowledge within organizations, its often highly uneven distribution (often solved in the consultant firms by the “principal/senior consultant/junior consultant” syndrome), and the motivational and human resources development issues connected with this, particularly in small enterprises, can become real challenges. How do you create a culture in which you consistently expect the best from people without demotivating young employees who are still in the learning curve? And how do you retain the specialists you have spent three years training, especially when your clients can outgun you almost every time on salary, titles and other traditional HR benefits?
  

• Scalability
  Even if translator productivity can be improved, however, the fact remains that capacity is and will always be limited. Scalability is the name of the game, and the lack of it is why many investors are currently steering clear of consultants and other professional services organizations. The reason is simple: since their activities are based on highly skilled people, growth is limited by the availability of the latter. This applies both in absolute terms (as everyone keeps saying, in today’s rapidly expanding global economy there simply aren’t enough really top flight people with all the requisite skills around) and in relative ones. To put it in a nutshell: there are only 24 hours in a day and once you have billed your clients for all of them you have to change your model to continue adding value. Of course, there are a number of ways of doing this, such as adding new people with different skill sets (e.g. consultants, to the extent that you can find them), and/or developing new services or products.

Falling in love again

What, then, will the effect of these developments be on the localization industry?

Firstly, we are likely to see even more mergers and acquisitions, but also more formal, semi-formal and informal alliances, many of them project based, as companies seek to tap the knowledge and resources they need for specific work or entry onto specific markets. Since under the new model the content change risk (having to keep abreast of the latest domain-related developments in two or more languages/cultures) passes to the translation/content adaptation providers, a tight focus becomes necessary. Specialist translation and localization companies are repositories of deep and wide multilingual, multicultural knowledge that is tightly focused on specific industries, domains, technologies and/or markets. By allying themselves with other like-minded specialists, they can effectively manage both one-off projects and continuous delivery models, as required.

Secondly, within service (sorry, solutions!) providers, we shall see a change in the status of “wordsmiths” of all kinds, be they translators and/or content creators. With some of the newer e-transformation companies already giving traditional ad agencies a run for their money in the area of multilingual Web sites, we are likely to see more aspects of “creative” behavior, working environments and remuneration policies. (Of course, this also presupposes similar levels of target language skill, as well as the ability to successfully marry such an ad hoc style with the process-dominated localization environment).

The upshot of all this is that specialist service providers, whether they originally came from “straight” translation or localization, will adopt the “high-touch” model described by Claudio Pinkus in Budapest, which is “founded on knowledge, customization and service. They put together solutions for clients founded on the belief that they (know) more than the client and (can) offer them this knowledge at a premium.” In this case, it will be interesting to see whether the localization industry as we know it will more or less disappear, subsumed into the global content delivery segment on the one hand, and the global content creation segment on the other.

---

Deborah Fry
Robin Bonthrone
Fry & Bonthrone Partnerschaft
In der Witz 29
D-55252 Mainz-Kastel
Germany
Tel +49-6134-2046-0
Fax ++49-6134-2046-47
E-mail: info@fb-partners.com