Much was said at the Forum about various aspects of the relationships between localisation service companies and product developers. It is apparent that these relationships are often strained, but that the tensions on the service provider side, for example, may well be hidden to the developer.

Cost is, of course, a major concern for both sides, and they are facing the need to cut costs without having an unnecessarily damaging impact on margins. However, product developers are generally in a better position to absorb short-term drops in margins than service providers; in an industry still dominated by small and medium-sized private companies, they tend to finance investment – including capital spending to secure future projects – from cash flow, rather than substantial retained earnings. Moreover, some service providers evidently feel that they are not infrequently being pressured to accept squeezes on cash flow to make good scheduling slippages and quixotic project management on the part of the computer product developer. On the other hand, there are product developers who feel that once they have selected a particular localisation service vendor, there is not much they can do to influence the course of a project.

There is probably an element of truth in both positions, but it is hardly a good advertisement for the supposed maturity of the localisation industry (meaning both sides) that these problems are still being voiced, not by newcomers, but by players who have been active in the field for some time. The reason for this is probably not so much that the two sides have been unable to talk to each other, or that there has been no will to do so – LISA has been providing this opportunity for a sufficient time now – but rather that the overall process of change in the industry is accelerating at such a rate that new problems emerge to dominate the minds of business and technical decision-makers before the “old” problems have been properly dealt with. After all, information overload affects the software industry just as much as – if not more than – the “downtrodden user”.

Shrink-wrapped localisers

At the Forum, we heard David Brooks of Microsoft talking about the “ideal Microsoft vendor”: basically a Microsoft in miniature, able to view localisation activities in terms of a long-term business investment strategy, rather than having any notions of short-term profit. This paragon of linguistic expertise is also able to handle any and all languages and uses no freelances, temps, or contractors. Later on in the Forum, Bernard Gateau presented the results of Workshop E, “Localisation Cost Structure”, with a list of the criteria perceived by computer product developers as being essential when selecting a service supplier, including “Ease of communication” (max. 2 time zones away from the development site) and “Either ISO certification or at least well documented processes which can be replicated.”

But how realistic are these target specifications, are they applicable across the localisation industry, and what are the implications for input service suppliers such as translators?

Taking the last requirement first, ISO certification or equivalent procedures is a matter which is being examined with some haste by the localisation and translation industries. Whilst ISO 9000-series certification has been obtained by a number of larger localisation service providers and translation agencies, there is also a belief that the procedures may too inflexible for much language-based work. This is the background to the development of the LISA QA Model and the draft DIN standard 2345. The ability to base processes capable of being audited, repeated and replicated on such quality paradigms will satisfy client requirements for quality-assured processes equivalent to ISO 9000-series certification, a development which will find widespread application throughout the localisation industry.

Teletranslation and the Extended Workbench

The question of vendor proximity to the client is one which has stimulated much debate and a great deal of unease in both the localisation and translation industries. Whilst the ideal Microsoft view of what might be termed the “localisation vendor in the office cupboard” is obviously not going to be installed for a while to come, the more general preference amongst clients for some degree of geographical proximity is causing headaches for many service providers. This is particularly topical in the light of the current debate in the translation industry on “teletranslation”, predicated on a vision of “a world where the world’s economies are moving towards a single integrated system connected by information networks”, with language services supplied by a “teletranslation service … totally integrated into a highly information oriented society driven by telecommunications. By harnessing human expertise and machine power … the teletranslation industry will overcome the one remaining barrier to effective global communications” (Zeger Karssen discussing teletranslation in the LISA Forum Newsletter, Volume V No. 2, May 1996).

Without wishing to be excessively sidetracked into a discussion of the notion (hallucination?) expressed above, there does appear to be something of a gap between what some service providers would like to achieve and what their potential customers are actually looking for. Not to mention the “one remaining barrier to effective global communications”; does this mean that we are really going to see cheap, reliable telecommunications and sufficient broadband capacities across the globe in the foreseeable future? Will cost no longer be a barrier between service providers and their clients? And will the clients be able to install the technical infrastructure needed to communicate with these leading-edge (bleeding-edge?) technotranslators? Perhaps these claims need to be asserted with a slightly greater degree of realism before they can be taken more seriously.

For localisation vendors providing direct services to their clients, concentrating wholly on the “teletranslation” paradigm is not an option at present. These “Tier I” vendors will still need client proximity for some time to come. One of the problems is not that the service companies themselves could not already provide almost 100% remote services, but rather that the client side is not yet ready for such a dramatic restructuring of the vendor/client relationship. A relatively large proportion of communication already takes place remotely in any case, by phone, fax and off-line/on-line file transfer. But there is frequently no substitute for on-site discussion and cooperation, although this requirement will diminish somewhat if efficient teleconferencing and collaborative processing systems take off (although according to the forecasts of several years ago, teleconferencing was already supposed to have revolutionized business relationships).

The Tier I vendors are already shifting towards what has been called the "extended workbench", i.e. “customer process integration”, a blend of remote and local services. In addition to more active consultation and feedback loops at the product design stage, there are moves to integrate service providers into customer enterprise messaging and intranet systems, enabling more rapid communication, real-time collaborative processing, shared resources, etc. Again, however, it will be some time before reality catches up with technological potential. We are regularly amazed – and frustrated – at the lack of technical (and in particular data communications) facilities at key players in many different industries, including major software houses. Perhaps this is a particularly European problem, and it would be interesting to hear the experience of vendors in other parts of the world, but one sometimes feels that there is little point in asking companies where they want to go today if they haven’t much of a clue where they're starting from.

Although many large companies, especially the global multinationals, are pressing on with the implementation of global, enterprise-wide networks, it is depressing to discover that these are frequently designed to keep information inside the company, rather than allow interchange with suppliers, with the exception of simple gateways, for example for EDI messages. In many cases, the concept of integrating suppliers is not even discussed.

“Teletranslation” is surely more attractive for the link between the “Tier II” service providers and the direct localisation vendors. Indeed, remote translation supply is already the standard model, allowing Tier I vendors to source translations on a worldwide basis as input data for the value-added services they themselves provide. This being the case, perhaps “teletranslation” is more a description of what is already in place, rather than Zeger Karssen’s “brave new world” (not that Huxley’s parable is particularly appealing in the first place).

Translators R.I.P.?

One aspect which did not come to the fore at the Forum was the need for subject area skills in addition to language ability. Although mass-market software, for example standard office programs, may not demand much in the way of subject area expertise, this is not the case in many other areas such as, for example, industrial process management, banking, the securities markets and other financial applications. In these cases, it is probably true to say that subject area knowledge on the part of the translator is even more important than “perfect” linguistic skills, a requirement which will only grow with the increasing complexity of the applications themselves to satisfy user needs in rapidly evolving markets.

Such expertise cannot be bought “off-the-shelf”, and most developers realize that they will inevitably have to pay a premium to buy application localisation services offering the requisite degree of domain-specific competence. In addition, they will generally have to look to smaller, specialized service companies who will rarely be able to meet the “all languages” requirement preferred by some vendors. In turn, these vendors will have developed their business strategy on the basis of a mix of in-house and external specialists offering the flexibility to cope with shifting customer demand. There simply is no other feasible option.

The relationships which are established between service suppliers and developers in these specialized fields are often radically different to those predominating in the shrink-wrap software business. Firstly, vendors do not tend to offer localisation services for a wide range of applications. The investment requirement in subject area expertise is massive compared with the skill levels needed in the mass market sectors, and the choice is therefore normally one of depth rather than breadth.

Secondly, the relationships tend to be much more intimate. In many cases, the customer will know the localisers and translators in person, and will have been persuaded by their domain-specific knowledge that they are the right people for the job. How many developers and localisation project managers at industry giants such as Microsoft, IBM or Hewlett-Packard have actually met the translators who produce their foreign-language software versions, normally via third-party localisation vendors? It does not appear to be a particularly vital issue.

Finally, the project management skills highlighted throughout the LISA Forum are frequently not as advanced in the specialized software localisation sector, although curiously enough, this seems to affect developers as much as localisers. The lower level of formalized project management procedures is coupled with an informality and flexibility rarely found today among the industry leaders.

In his closing address, “Sunrise or Sunset? Localisation 1999”, Alex McDonnell said that he was waiting for the emergence of the “EDS of Content Publishing”, offering one-shop turnkey services on the basis of an equal partnership with the client. Whilst this vision may take some time to become reality in the shrink-wrap software sector, there are already more than a handful of “micro-EDSs” in the specialized software field. This does not mean that these vendors necessarily have a near-monopoly on a particular product or service segment, but rather that their relationship with the developer is based on something more like equality. It is a partnership in the truest sense of the word, and the one side would find it very hard to exist without the other.

It is difficult to forecast whether the decline in translation-related cost in the mass software market (25% and falling, according to David Brooks) will be reflected in the more specialized sectors, where "translation" costs have traditionally been relatively high, and are likely to grow with increasing application complexity. Greater use of tools will undoubtedly result in the lowering of purely language-based costs, but this trend will probably be offset by the cost of continuously updating subject area expertise.

However, there may come a point at which developers will wish to see vendors passing on tool-driven savings, and the latter will have to be in a position to present detailed cost statements to justify their pricing models. Above all, vendors cannot afford to be complacent: they must expect a (for them) negative trickle-down effect from the shrink-wrap market, and even the closest vendor/client relationship may become strained. It won’t happen tomorrow, but specialized localisation companies should never forget that the large industry players tend to have an impact even on those market segments in which they do not operate.

Equally, since much of routine localisation work is ultimately capable of automation, the question of what precisely distinguishes Vendor A from Vendor B will become increasingly important over the next few years. The answer lies in our opinion in what may be called the market/marketing aspects of the job – i.e. locale, target group, and subject area knowledge. This development, too, will tend to blur the current de facto boundaries between the "high volume / low margin" and "low volume / high margin" paradigms currently in force.

One for all…

Is David Brooks then right when he predicts that the localisation industry will be dominated by a small number of vendors in future? Again, this is very much a Microsoft view of the world. It is quite possible that in future, the market for shrink-wrap application and multimedia localisation will be serviced by a relatively small number of localisation vendors, but this is not necessarily the case for the specialized software described above. There is also the legitimate question of whether such a concentration is in the best interests of the developers themselves. Given the volatility of both the software and the localisation industries, the well-known attractions of one-stop shopping have to be weighed against the potential disadvantages. For the product developers, there is the risk of over-reliance on a particular vendor or small group of vendors. Given the amount of know-how which would necessarily be tied up in these service suppliers, the phenomenon of “corporate amnesia” at a vendor could have a serious effect on projects. Secondly, in an environment without free competition, there would also be an inherent risk of price-fixing between vendors. Either way, developers should think very carefully before adopting policies which could encourage monopolistic trends in certain segments of the localisation industry.

Equally, there are serious risks for the vendors. The “corporate amnesia” problem can hit them just as hard as the developers, in fact probably more so. Experience elsewhere in the software industry shows that strategic partnerships can become seriously unstuck simply because of a change of line manager at the client. Moreover, the age-old principle of not putting all your eggs in one basket will not lose any of its validity: it’s not so much that all your eggs will break if the bottom falls out of your basket, but rather that – to put it bluntly – “baskets ain’t what they used to be”. In a global business environment characterized by rapid change, unpredictability and deep-rooted insecurity, the cost of managing a diverse customer portfolio will always be less than the loss of a single customer accounting for the bulk of a company’s revenues.

To sum up, the lessons of the 1996 LISA Forum USA are perhaps that the localisation industry is constantly evolving and reacting to external circumstances. To borrow Alex McDonnell’s metaphor, it is at one and the same time both a sunset and a sunrise industry, able to discard old models whilst simultaneously attempting to meet new challenges. Indeed, the localisation industry is almost too heterogeneous – in terms of its members, its clients and its services – to be viewed as a “conventional” high-tech industry, and what holds true for its global players does not necessarily apply to its Mittelstand. What is most encouraging and impressive, however, is the flourishing debate and active searching for new answers which characterizes LISA and its members, who play a driving role in this process.

---

Robin Bonthrone
Fry & Bonthrone Partnerschaft
Language Consultancy and Services
Rochusplatz 10
D-55252 Mainz-Kastel, Germany
Tel +49 6134 22504
Fax +49 6134 22860
E-mail 100277.3467@compuserve.com