Industry Dashboard: The 2008 Forecast
There were more publicly announced customer wins for speech recognition solutions across network, PC, and embedded environments in 2007 than in 2006. The number of deployments and the size of implementations for network-based speech recognition solutions increased as companies became more familiar with speech recognition’s capabilities and limitations. The number of devices and automobiles with embedded speech recognition grew substantially in 2007, while speech-enabled dictation for healthcare transcription is exhibiting greater market uptake.
Is the industry nearing the hockey-stick growth that was talked about for years? Probably not. Speech is, however, becoming more of an accepted user interface. Consequently, we expect steady growth in spending for speech recognition during the next five years.
Having calculated 2006 and 2007 retro figures (shipment and revenue data) and combining those with new forecasts aligned to vendor and customer data, 2008 will be an important year for network-based speech recognition solutions. The following represents Datamonitor’s predictions:
• Increased uptake for identification and verification (ID&V) speech recognition applications: Companies that have deployed information provision and transactional speech applications are in the second or third phase of their speech strategy; many want to optimize their initial phone menus to improve routing capabilities.
• Growing interest in voice verification (voice biometric) applications, primarily among financial services institutions and government agencies. A handful of large-scale commercial deployments will serve as the crucible test for consumer-based voice verification applications.
• VoiceXML displaces traditional IVR: For the first time, there will be more VoiceXML ports shipped in 2008 than traditional IVR. This evolutionary step opens new possibilities for multimodality and video, the next phases in IVR.
• Sharpening focus on speech-enabled directory assistance and mobile search markets: New channels of phone-based advertisement revenue attracted Nuance (Bevocal), Google, and Microsoft (Tellme) to this burgeoning market. A more defined value chain for speech-enabled directory assistance and mobile search will emerge.
• Information provision applications will thrive in the hosted speech segment: Many midmarket companies’ first forays into speech should come in the form of simple information provision applications in a hosted environment to power campaigns and basic information access applications.
• Consolidation in the application and analytics space: Further consolidation will occur in the application and analytics areas as vendors try differentiating their systems.
• New deployment paradigms for speech in the enterprise: Network-based speech recognition should gain further traction in warehouses and the supply chain. Speech recognition’s importance is finally diffusing into the mindsets of IT decision-makers supporting back-end logistics.
This might be the year in which the industry further succeeds with improved speech applications and performance. Task completion rates for speech applications are expected to improve along with best practices and application design.
Source: Datamonitor. It is important to note that the information provided in the Industry Dashboard reflects only publicly announced customer wins. Datamonitor provides more detailed and comprehensive studies of the speech recognition market. For more information on Datamonitor’s research please go to www.datamonitor.com or email the author at dhong@datamonitor.com.