Adventures in Speech Mobilization

What's impressed me most about recent "wins" for purveyors of CAT-based solutions is that mobile service providers are finally buying solutions that bridge the gap between customer care and service delivery. It's more than "call steering," which automates call processing and call routing to support single-number access to customer care, billing and trouble reporting.

These new deployments field calls that cover the constant variety of subscriber needs and wishes. Time was, not so long ago, that the distribution of automation was skewed heavily toward complaints about dropped calls and requests for refunds. Today's mobile callers dial an abbreviated access code (akin to 611) to order ringtones, streamed content, top off a pre-paid plan or even get a phone number for a local business (a la a directory assistance).

A Better Way to Search

An array of Nuance partners and customers deserve credit for advancing the cause of speech-enabled mobile and local search. Promptu (formerly AgileTV) has successfully adapted the resources it developed to support plain English search of video guides so that it can support search for ringtones and streamed content to mobile devices. Likewise for V-ENABLE, whose flagship product has supported the search for ringtones for Leap Wireless' Cricket® MVNO for a little less than a year.

When the United States Patent and Trademark Office (USPTO) granted Google the rights for a voice-enabled search patent, the whole idea took on greater legitimacy. Google's patents are so narrow and specialized that they may not lead to a significant revenue stream for the search giant - either as the basis for a new business or as a source of license revenue. Yet its very existence is akin to a virus that spreads by "jumping species" and makes Web searches of a mobile or local nature a more legitimate business opportunity.

Before year's end, you'll see search specialists like InfoSpace, Yahoo! and perhaps Google debut speech-enabled, mobile products.

Interactive Mobile Care

One of the most impressive services I've seen lately was showcased at the 3GSM Forum last March. Ydilo, a very progressive ASP based in Spain, showcased a set of customer-care and service delivery applications that foreshadow what's coming up for mobile subscribers around the world. The service amounts to a true mobile portal. It combines automated speech, text-messaging (SMS), multimedia messaging (MMS), games and streamed media (including video messaging) to better respond to emerging customer needs in real-time.

Ydilo had already earned its speech-enabled merit badge by hosting Vodafone's speech recognition-based customer care portal over the past couple of years. Its platform, based on Nuance's ASR and TTS, was handling something on the order of eight million calls per month solely for customer care over the telephone. The new platform is a major upgrade designed to support the mobile subscribers ability to "talk, tap or type" their way into the system to initiate queries, provide instructions or order services.

Vodafone maintained an aggressive schedule of service introductions and product promotions, requiring constant updates of the applications and scripts running on the Ydilo platform. To keep pace with demand for frequent changes and enhancements, the hosted service provider assembled a platform that took great advantage of well understood building blocks: VoiceXML - to support simple dialogs and SpeechObjects, reusable Java Applets developed by the west coast Nuance prior to the merger, supporting rapid deployment of more complicated or transactional services.

Combining Customer Care and Service Delivery

For Ydilo, its "transactional layer" became just as important as speech enablement when Vodafone Interactive Care made its debut. In addition to fielding the traditional subscriber complaints (regarding dropped calls and billing issues), Vodafone closely linked the automated speech applications to delivery of streamed content and enhanced messaging services. The ability to add new services and content quickly is crucial both for justifying investment in the IP-Multimedia Subsystem (IMS) and to revenue maximizing strategies that include customer retention and lock-in by fulfilling the promise posed by handsets that include larger screens, MP3 players and video cameras.

It will call for combined efforts of device makers, content providers and speech technology providers, but it is clear that, before year's end, mobile service providers are poised to deliver compelling content and services on spoken command from their customers.

Dan Miller is senior analyst for Opus Research. He founded Opus Research, Inc. and published Telemedia News & Views, a monthly newsletter regarding deployments in voice processing and intelligent network services. He can be reached at

SpeechTek Covers
for qualified subscribers
Subscribe Now Current Issue Past Issues