The power of voice: VCs, others consolidating speech-recognition companies

Even in the confusing, clinical world of mobile software terms, the term “speech-recognition technology” draws yawns. Lately, though, it’s also drawing a flood of investment money.
Speech-recognition developers for years were the Eli Mannings of the wireless world, failing to live up to loads of hype and promise.
Mass-market wireless users simply didn’t see the point, it seems, in saying a few words into their phones instead of pressing buttons to make a call or check voicemail.
But as phones evolve from voice devices to portable messaging devices and multimedia players, industry players increasingly see voice as a way to let consumers cut through cumbersome carrier decks and access online content from their handsets. And nowhere is that more evident than in the mobile search space, where Internet behemoths are squaring off-or swallowing up-their pure-play competitors.
Microsoft Corp. jumped headfirst into the space in March, making headlines with the acquisition of Tellme Networks Inc. While terms of the deal were not disclosed, analysts speculated that Tellme’s price tag may have been as high as $800 million. The software giant said it will integrate Tellme’s wares into its line of familiar products including Vista, Office and Windows Mobile, as well as adding voice capabilities to its nascent Live Search services.
Microsoft’s move was preceded by Nuance Communications Inc.’s $140 million acquisition of BeVocal, a competitor on the speech-recognition field. Nuance, formerly known as ScanSoft, is a 14-year veteran of the space and has deals with Verizon Wireless, T-Mobile USA Inc. and Virgin Mobile USA L.L.C., among others. BeVocal was founded in 1999 and had snared more than $80 million in venture capital. Nuance, which in December acquired a startup called MobileVoiceControl, said the BeVocal buyout will add between $21 million and $23 million in revenue in 2007, and as much as $70 million next year.
Promptu, one of the more interesting players on the playground, recently pocketed $5.6 million in a round of funding in a round led by existing investors Steelpoint Capital Partners and Lauder Partners L.L.C. The company, which was founded as AgileTV in 2000 before shifting its focus to wireless last year, has raised more than $17 million since its transition from set-top boxes to mobile phones.
Meanwhile, Google Inc. last week began showcasing its latest speech-recognition technology with a voice search service that delivers local business listings to fixed-line and mobile users. The offering, which is available on the Google Labs site, invites users to dial a toll-free number from any phone to search for businesses by category or name. Users can opt to be connected directly to listed businesses, and wireless users can say “text message” to receive information via SMS.
“Using this service, you get fast access to the same local information you’d find on Google Maps,” according to a description on Google Labs. “But you don’t need a computer, you don’t need an Internet connection, and you don’t even need to use your cell-phone keypad. It’s voice-activated, and you can access from any phone (mobile or landline), in any location, at any time.”
Indeed, the advantages of using voice instead of triple-tapping to perform mobile searches are obvious, if not necessarily intuitive. Not only does speech-recognition technology minimize the number of click-throughs needed to find a specific ringtone, for instance, or navigate to a particular URL, it typically can be used behind the wheel or while otherwise occupied. Also, like other mobile search offerings, voice-driven services can be ad-subsidized through pay-per-click or pay-per-call models, saving the consumer a directory-assistance call than can cost a few dollars.
And while speech-recognition software may not yet be accurate enough for wide-ranging applications such as voice-to-text, mobile search technology can be confined to specific parameters, leaving less room for error.
“I know where you are in the application, so my accuracy is really high,” said Craig Hagopian, CEO of V-Enable Inc., a San Diego-based developer. “If you’re asking for airlines, and you ask the name. I know there are only 200 airlines in the world, and I’ll nail it every time.”
V-Enable is readying to launch Mobile411, a white-label, voice-driven local search application that combines business listings with maps, driving directions, weather and other information. The 50-employee company plans to announce at least one tier-one carrier partner in the next few months, and is gearing up to do battle against the deep-pocketed newcomers currently buying their way onto the field.
“The year of voice is finally here,” Hagopian said hopefully. “Our company has been around since 2001. We were hoping this day would come many years ago, but better late than never.”