Europe Embraces Speech
“Give me a lever and a place to stand, and I can move the earth.” Ancient Greek philosopher Speech technology is fast becoming the interface or “lever” that enables the people of the world to communicate - anyplace, anywhere, anytime, on any device
and in any language. This is especially true in Europe, where geographic boundaries, multilingual and multinational requirements, and regulations pose a unique set of challenges, and where different, and sometimes more complex voice-enabled solutions are required.
“It is a mistake to look too far ahead. Only one link in the chain of destiny can be handled at a time.” Winston Churchill While Europe may have been a late-comer to speech technologies in some sectors compared to the U.S., their markets are quickly maturing and enjoying wide appeal. Even when predictions differentiate between licensing revenues and solution or service revenues, as Etienne Lamort de Gail of Elan suggests, forecasts for voice service revenues worldwide look very promising, ranging from $1.6 billion (Cahners In-Stat Group) to $3.5 billion (IDC) to $16.3 billion by 2005 (The Kelsey Group). Last year, HSBC predicted that the worldwide market for TTS software alone would grow to $6 billion by 2005. On that basis, Marc Moens, CEO Rhetorical Systems, says that the European market will grow to be worth $1.5 billion dollars or a quarter of the entire worldwide market. The Speech Recognition Company’s (SRC) analysis suggests at least $400 million will be invested in speech technologies in Europe during 2002, with about 25% spent on application development. Most (95% +) of the investment in Europe to date has been on speech recognition technologies (ASR).
“Aim for the highest.” Andrew Carnegie Projections vary from country to country, analyst to analyst, particularly within specific speech applications. In Europe’s wireless market, for instance, deregulation enabled the telcos to compete with one another, creating a huge number of mobile phone users, estimated to be approximately 44% of the worldwide market (Yankee Group). If predictions made by Pyramid Research are correct, the mobile penetration rates in (Western) Europe will approach 100% by 2006, creating an enormous opportunity for speech-enabled phone services in Europe. Furthermore, legislation in England, Italy, Israel and 21 other countries banning the use of handheld mobile phones in automobiles is virtually forcing wireless operators to embrace speech technology. On the telephony/server speech software market side, where customers are demanding service 24/7, tele-banking, tele-trading, tele-shopping, tele-payment, tele-healthcare, and content information management applications already reign supreme. These types of applications have spilled over into the leisure market as well, as evidenced in the UK, where taxation laws have removed the advantage of offshore betting. The demand for telephone betting operations, already beyond the point of cost-effective delivery by traditional call centers, has increased exponentially. SRC is one of the speech recognition vendors responding by working with major gaming companies to offer speech automated telephone betting for the April 6, Grand National 2002. This is the UK’s biggest horse race, which in past years, required at least quadruple the number of live agents to answer calls.
"In theory, there is no difference between theory and practice. But, in practice, there is." Jan L.A. van de Snepscheut All of these prognostications are a fine way to get a handle on the industry, but fail to take into account local and regional realities. As Loquendo’s CEO, Silvano Giorcelli, said, “Europeans tend to think of market size in terms of the number of deployed ports and don’t break it down into individual speech technology components. After all, buyers need ALL the speech components in order to deploy a complete voice solution. This is in marked contrast to the North American market which tends to have a more complex set of speech component suppliers, all competing for a piece of an installation. European customers value the convenience of one supplier who provides a complete integrated voice platform solution which quickly puts them in the voice application business, not the platform integration business.”
“Everything should be made as simple as possible, but not simpler.”Albert Einstein The European market is a practical one where, according to Nick Applegarth, managing director of Nuance, “simple is sexy”. “A deployment does not need to be complicated and flashy to be appealing and add value. If it is practical and works and people have a good experience with it, they are more willing to use those services again.” A prime example is in Italy, where almost everyone has encountered speech recognition through “12” White Pages (Directory Assistance and Reverse Directory Assistance) and speech technology is fast becoming a tool rather than a novelty. Increasing acceptance of speech technologies drives demand for voice services from both enterprises and end-users. As successful and user-friendly deployments become more widely understood and the financial benefits easier to project, speech-enabled services become cost-effective alternatives to conducting business. Once companies experience return on investment (ROI) and increased customer retention, the circle of acceptance is complete.
“A likely impossibility is always preferable to an unconvincing possibility.” Aristotle There are many characteristics unique to the speech technology market in Europe. One is the high penetration of mobile phones, estimated at 70% by 2003 according to DataMonitor. High penetration makes the telephone an ideal device for delivering speech services. They’re always available, easy to use and will probably remain the tool of choice for many people. Additionally, Europe’s lack of Internet penetration in homes, due to high telephone service charges and late attention to Internet education, is a unique phenomenon. Almost all telephone services in Europe – even local residential services - are charged on a per minute of use basis which further hinders Internet adoption because people do not remain on-line for hours.
“Off-the-rack solutions, like bargain basement dresses, never fit anyone.” Francoise Giroud The most obvious characteristics that drive the European market are the multilingual and multinational requirements of anyone using voice-enabled applications. Voice solutions must support multiple languages, especially when accessed by tourists and Pan-European business people. Europeans have a cultural tradition of wanting to speak to people in their native language. Automated voice products and services have evolved to meet this expectation, typically by prompting callers for the language in which they would like to speak or by employing engines that either recognize multiple languages simultaneously or utilize speaker verification technology that is language independent. Europe is, by its very geographic nature, a fragmented market, which, according to Elan, “gives an advantage to European companies de facto” because of the heavy investment required. However, the venture capitalist’s attitude in Europe of looking at short term ROI means less overall investment in Europe than in the states. Laurent Balaine, CEO of Telisma, emphasized the importance of human and legal factors when designing applications within such a fragmented market. While the design methodology may be the same, the customs and language differences of each country must be taken into account, requiring the use of local resources in each country in order to penetrate the market.
“How can one conceive of a one-party system in a country that has over two hundred varieties of cheese?”Charles de Gaulle Not only must voice-enabling products and services be deployed in more than one language, dialects and regional accents within each country must also be factored in, as well as the “way” in which people say things. As Roy Timor, marketing director for ART pointed out, Swiss French, France French and Belgium French are different dialects of French. In France, they also say numbers in pairs, for instance, an eight after a one means ten plus eight. Likewise in Germany, as Maryellen Cronin of ScanSoft said, Bavarian German is totally different than North German just as Boston English is different from Southern English. The key to success in deploying speech, therefore, is the ability to support cultural and geographic variances. These variations require very complex adaptations of the recognition engine to decide in near real time the correct format of the number. “These variations”, said Steve Chambers of SpeechWorks, “must be accounted for in the user interface design that usually translates in subtle differences in prompt phrasing. For example, German callers may perceive it to be patronizing to be constantly offered help, so a German speech service typically prompts the caller in the beginning to just say “help” if they need assistance and not repeatedly offer help throughout the conversation.” Clearly, the design challenges to building and deploying a user-friendly interface in a multi-lingual, multinational and very people-mobile environment are immense.
“The world is round and the place which may seem like the end may also be the beginning.” Ivy Baker Priest Europe has relied heavily on the experience of the U.S. in developing the speech business and for good reasons. The U.S. is primarily “uni-lingual” with millions of potential users and a willingness on the part of investors to try out new technologies and to push the boundaries. U.S. companies have also had a head start on research, developing valid engines that are not easily or quickly duplicated, and have adopted an attitude that mistakes are expected and acceptable. The European approach to deploying speech technologies has been more cautious, particularly when costs to design, research and develop voice-enabled applications have been prohibitive. Organizations must be convinced that speech technologies really work and provide real benefits before, as Nuance said, “dipping a toe in the water.” European companies watch to see what happens in the U.S. and apply those lessons to their own deployments, often leading to better designed and sometimes more effective applications. Europeans do not subscribe to the “Field of Dreams” approach (build it and they will come), according to Dr. Kenton Sanmogan of SRC. “Only the telcos have put any significant money into voice portal type services based on the view that it should drive up network traffic and revenue from call minutes. Almost all other investment in Europe to date has been justified by water tight business plans based on cost saving through automation and redesign of business processes. This has, for the most part, limited projects to those where usage for the services (calls from customers) already exists.”
“No culture can live if it attempts to be exclusive.”Mahatma Gandhi Repeatedly, developers of voice-enabled applications mentioned the cultural barrier in Europe, where Europeans just don’t like the idea of talking to a machine, particularly in public places. And as Marcus Klische, director business development, The BioID Group said, “Passive recognition, where the system simply listens to the user without their knowledge is not accepted either, because of privacy concerns.” And there is the additional taboo of talking about automation or processes that Europeans may equate with layoffs. As Mark Bannon and John McCready of Phonetic Systems explained, “the strength of the Unions has and is a factor for the speech industry which typically relies on personnel replacement or personnel augmentation as part of the strategy for a ROI story. The regulatory environment makes it difficult to replace people with technology since laws require six months notice and six months severance for employees being laid off.” Obviously, it’s much harder to make decisions about work, especially when it applies to situations that might eliminate CSR positions.
“The expectations of life depend upon diligence; the mechanic that would perfect his work must first sharpen his tools.”Confucius While it may be true that speech technology penetrated the U.S. market first, the collaboration and cooperation among the continents has certainly bridged any oceans that separate us. The global nature of speech technologies encourages existing speech vendors to develop partnerships with a growing base of companies throughout Europe, as evidenced by the number of U.S. based companies with subsidiaries in Europe. These alliances are, according to Rhetorical, a mutual learning experience, where “we are all developing novel applications and creating an ever expanding market”, which in turn is helping drive the European industry forward at a rapid pace. Speech is coming of age; people have confidence in the technology and it is mature enough to really be helpful to users and enterprises. Companies are earning more revenue and saving money – both of which are key business factors for fully embracing this technology. The future is bright, the lever is strong, and the possibilities to move the earth appear very real indeed.
Nancy Hapgood is executive editor for Speech Technology Magazine. She can be reached at 859.278.2223 or at nancy@amcommpublications.com.