Peter Mahoney, Vice President of Worldwide Marketing, ScanSoft
Q. One of the key themes at this year's SpeechTEK West was VoIP. Tell us how you see the market opportunity for the industry, and the role of speech.
A. There are two main reasons for the excitement around VoIP & contact center solutions: ROI improvements through cheaper calls, and enhanced services that benefit from the convergence of voice and data. Many Global 2000 businesses are already using VoIP to reduce the cost of internal communications. As broadband adoption becomes the norm in more and more markets, the number of everyday solutions that exploit VoIP will grow. The immediate baseline opportunity for call center operators is to deploy the same speech-enabled applications that they deploy today, but in a VoIP environment, which will improve ROI through cheaper calls alone.
The second type of applications that make VoIP exciting - particularly for speech applications - are those that exploit the convergence of voice and data. At the simplest level this could mean enhanced voice-activated dialing services, but there are nearly endless possibilities for data management, unified messaging and other services, and multi-modal applications in the network.
In order to deploy successful speech enabled applications in a VoIP environment, businesses will need to adopt robust speech technologies that are optimized to perform well with the variety of codecs used in VoIP calls. This particularly affects speech recognition and speaker verification because the inbound signal is affected (as opposed to text-to-speech, where the speech generation is not affected). The second area businesses need to focus on is network traffic: dropped data packets will affect application performance.
At ScanSoft, we have been focusing on VoIP for some time, and we're excited by the exceptional performance we've achieved with OpenSpeech Recognizer and SpeechSecure across all the widely deployed VoIP codecs. We also continue to make improvements in robustness to dropped data packets, although we expect market demands to drive VoIP service providers to ensure increasingly performant infrastructure to improve QoS (Quality of Service). We're very well positioned to deliver successful solutions in a VoIP environment.
Q. What is ScanSoft's VoIP strategy?
A. ScanSoft's VoIP strategy echoes our overall speech strategy. We help customers solve their business problems, improve their service levels and drive down their costs by developing speech-enabled applications that callers enjoy using. We achieve this through several means: our exceptional network of partners; our unique expertise in delivering successful speech solutions; our leadership in driving open standards; and through the relentless optimization of our market-leading speech and dialog technologies. We are already working - and have been for some time - with partners including Avaya, Cisco, Genesys and Nortel to deliver the same value proposition in VoIP applications, and we're already seeing businesses and end-users reap the rewards of this strategy. On a technical level, we've had explicit VoIP support built into OSR since version 2.0, and we worked closely with Cisco and others on the creation of MRCP, which is particularly well suited to VoIP architectures.
Q. Where do you see the market for VoIP in one year? Five years? What will be the rate of adoption?
A. The market is currently in the early adoption phases in the U.S. and probably a little further along in Asia. There is a proliferation of service providers, and we expect to see some consolidation over the next five years. This consolidation will not come at the expense of adoption however, and we expect to see the rate of adoption accelerate. Analysts predict anywhere between 20 percent and 30 percent of the U.S. market will use VoIP for their domestic telephone services by 2008, and that VoIP will be prevalent in intra-business communications. This is why ScanSoft and our partners have been aligning our organizations to serve the market - speech enabled customer service applications must be in place in the VoIP network as the market demands them.
Q. How can vendors maximize their opportunity in VoIP?
A. Prepare early - be sure to have an infrastructure that can seamlessly handle VoIP calls, and that your speech-enabled customer service applications are robust and flexible enough to handle calls from various sources - VoIP, PSTN, mobile phones, and so on. Talk to speech solutions experts, like ScanSoft and our partners, to make the right decision about how, where and when to deploy powerful speech applications that can adapt to the growing demand for VoIP. VoIP environments will also enable sophisticated applications that take advantage of voice and data convergence, requiring more advanced dialogs. To deliver these solutions businesses will need access to intelligent dialog technology that understands how to manage conversations - the brain behind the ears of ASR and the mouth of TTS, if you will. Intelligent products like OpenSpeech Dialog are just as critical as robust speech technology in supporting these more advanced and natural conversations and we expect VoIP environments to stimulate demand in this area.
Q. Please update us on the integrations of your latest acquisitions (Phonetic Systems, ART and Rhetorical).
A. Each of the acquisitions has officially closed, and we are excited to formally welcome the employees to ScanSoft. We are pleased with the progress in the integration of each of the companies, and are excited about the contributions and fresh perspective that many of the employees have brought to our organization. While there is still additional work to be done, we are making significant progress in each of the product lines that draw upon the combined strength of our organizations; in fact, we expect to launch a new product stemming from our progress with Phonetic Systems in the next month or so.
Again, we are very pleased with the addition of each of these organizations, and with the commitment and enthusiasm we've seen from our employees around the world. We continue to believe that each of these companies will help us to deliver on our strategy for a future where natural, human conversations will be the predominant way that people retrieve information and interact with automated systems. To achieve this, we recognize that speech applications need to become more dynamic, sound more natural, interact more conversationally, perform tasks on multiple devices and adapt to personal preferences. Each of these organizations will contribute to this vision with an array of technologies, customers, partners and talented employees to help propel the quality of speech applications throughout the world.