Total Pageviews

Friday, April 04, 2008

Voicemail-to-Text Services Picking Up Steam

Copyright © 2008 Unified-View, All Rights Reserved Worldwide

April 3, 2008

Will Voicemail-to-Text Messaging Services Close the Recipient’s Loop For Enterprise Unified Messaging?

Art Rosenberg, The Unified-View

The experts at just published a comprehensive UC eBook that focuses very heavily on defining the various components of UC, which are primarily:

  1. Real-time calls and conferencing (Person-to-person, customer contacts)
  2. All forms of asynchronous and real-time multimedia messaging
  3. Communications-enabled applications (Automated applications that act as “contact initiators” to individual people)
  4. The use of presence management technology to intelligently initiate a mode of contact and communication based on the accessibility and availability of the recipient(s)
  5. The ability for users to dynamically and easily move between the different forms of contact as the situation requires and allows.

From an end user perspective, UC’s objective is to enable complete user flexibility as either a contact initiator or contact recipient/respondent by enabling them to individually and independently choose any mode of communication at any time. If all parties want a real-time conversation, they can immediately activate or schedule a voice/video conference call, or initiate what I have termed an “as soon as possible,” (ASAP) call. The practical alternative would be any form of messaging that the contact initiator chooses, without worrying about how the recipient will get the message. That would bring into play the practical flexibility of unified messaging capabilities.

For business and service users, automated applications will also act as contact initiators by generating asynchronous messages that need to be delivered in the medium of the recipient’s choice (voice, text, image). Just as email and Instant Messaging enable online interactions with online “click-to” options, messages from a business process or service application can also provide access to either self-service applications (voice or visual), a simple structured message response in voice or text, or, if necessary, to live assistance (“click-to-call,” “click-to-chat,” “ click-to-message”). Such flexibility is geared to quickly notifying a user about an issue and minimizing “human contact latency” wherever possible.

The Problems With Voice Mail Messaging

UC technologies are still evolving by looking at the operational problems that end users have with legacy communication applications and eliminating or minimizing such problems by exploiting converged interfaces and functions. Even though legacy voice mail systems helped eliminate those paper “pink slips” that “message desk” staff used to transcribe an error-prone and often delayed telephone message (name, number, and cryptic “message”) user voice message retrieval and management was not very efficient with only the traditional Telephone User Interface (Touch-Tone command inputs and voice outputs).

Unified Messaging addressed the need converge the management of email and voice mail messages in several ways, including combining all forms of messaging into a single mailbox, enabling message retrieval to convert text messages to voice for retrieved with a telephone interface, and, with extending a screen-based email interface to include “visual voice mail” to selectively listen and dispose of specific voice messages from a screen GUI, rather than learn to navigate TUI Touch-Tone commands.

However, “visual voice mail” doesn’t really address the basic inefficiency of navigating through voice message content, i.e., having to skip backwards and forwards to repeat certain parts because they were hard to understand or needed to be transcribed as text notes. These are problems that text messages don’t have because the full message is immediately displayed, can be scanned quickly on a screen, stays visible as long as necessary, can be printed, copied and forwarded easily, can be archived for easy search and regulatory compliance, and doesn’t require the storage or the network bandwidth that voice messages do.

Voicemail-to-Text Message Transcription Services to the Rescue?

To minimize those problems, the next phase of unified messaging technology has already appeared in the form of voicemail-to-text message automated transcription services. Using the power of mature speech recognition, coupled with human transcription services to handle “exception” cases, voice messages can be quickly converted to text email or SMS messages before delivery to the recipients. This will typically be a service subscribed to by a voice message recipient, not a voice message originator, although the latter may also be a useful service option in the future.

Because the convenience of speech input is particularly important for eyes-free/hands-free input with handheld mobile devices, there have been a number of noteworthy announcements about voicemail-to-text messaging in conjunction with the big CTIA show. These included:

· New partnering announcements with telephony service providers from voicemail-to-text message developers Spinvox and SimulScribe

· A Senior Cisco executive joining the board of CallWave, which offers Vtxt services that delivers the “gist” of a voice message as a text message

· Last, but not least, the entry of leading speech recognition developer, Nuance, into the Voicemail-to-Text Message services market through major carriers

Voicemail-to-text technology can’t be perfect because of the difficulties in understanding speech in different dialects and varying voice qualities. So, although automatic speech recognition technology (ASR) is used for the bulk of the transcription, manual transcription is brought into play when ASR runs into any kind of problem. In addition, the original voice message recording is also available for selective access by the service subscriber.

Underscoring Nuance’s entry into the Voicemail-to-Text service business, their announcement highlights their use of over 3,000 Nuance transcriptionists in a centralized Nuance facility. Since this clearly a responsibility that enterprise organizations won’t want to do themselves, it is clear that this chunk of unified messaging is another piece of the UC puzzle that will exploit outsourced, hosted services.

What Do You Think?

You can contact me at: or (310) 395-2360.