SAN FRANCISCO (CNN/Money) -
Almost lost amid all the hoopla over the European Union's ruling against Microsoft was some interesting news: On March 24, Redmond unveiled its Speech Server product line, which it expects to begin shipping in the next few weeks.
The launch, announced last week at the annual SpeechTEK conference in San Francisco, is the first major speech-related server product release for the company, and reflects an interest in such technology that "goes all the way up to [Bill] Gates," says Steve McClure, an analyst at IDC. Gates himself introduced the product in San Francisco.
It seems that every year is touted as the year that speech recognition will finally break through and go mainstream. That, of course, hasn't happened yet.
It's come a long way
But speech technology has come a long way since I first heard about it in 1996, and Gates himself was almost cautious when he sized up the future of speech recognition at Gartner's Symposium/ITxpo on Monday.
"Many of the holy grails of computing that have been worked on over the last 30 years will be solved within this 10-year period, with speech being in every device," Gates told the crowd, according to News.com.
A small but growing industry has developed around the technology and includes public companies such as Nuance and ScanSoft. IBM has pushed speech recognition for years as part of its WebSphere offering. According to a recent Gartner report, Nuance is the largest company in the sector, with ScanSoft close on its heels.
Gartner analyst Steve Cramoysan estimates that the total U.S. market for speech-related products in 2003 was between $750 million and $800 million. That market is expected to grow for the foreseeable future, in part because of Microsoft's (MSFT: Research, Estimates) entry but also because more applications for speech are appearing, most notably in the cell-phone industry.
More vendors are looking to add voice services, such as voice e-mail and voice-based traffic data, as a way to retain customers in this age of number portability.
An emerging market
As is the case with any emerging technology, Microsoft's entry "validates the market," Cramoysan says.
"It shows that speech is a significant interest to a major player. Second, Microsoft's ability to promote it and educate the market is substantial compared to the incumbent players."
"The [speech] vendors out there now have done a great job," says Bob Clough, a general manager for Microsoft. "But what's broken in the marketplace is the size of the marketplace: It's much too small. We're approaching it in a high-volume, low-cost way...expanding the market downstream. No longer do you need a speech scientist to come in and sprinkle the magic dust."
"We like to say you get what you pay for," says Chuck Berger, CEO of Nuance.
|Recently in Tech Biz
He says his company's experience in the market, its blue-chip client list (which includes such names as Cingular, Sprint (PCS: Research, Estimates), T-Mobile, and Verizon (VZ: Research, Estimates)), and its products' reliance on open standards such as voiceXML will help Nuance withstand the Microsoft juggernaut -- even though Microsoft's prices will undercut the market by about 40 percent.
Nuance has seen respectable revenue growth of late, but is not yet profitable. Berger told me he expects his company to be profitable in the second half of 2004.
Microsoft claims that it will have more than 60 partners for Speech Server when it starts selling it. ScanSoft has already hammered out an agreement.
Berger says he's searching for partnership opportunities with the Redmond crew. "We're in regular conversations with them to find points to work together; we haven't found those points yet. But we're still looking."
Gartner's Cramoysan says he'd "hate to have to bet which companies will survive Microsoft's entry into the market."
In addition to ScanSoft, some other companies are hedging their bets by partnering with Microsoft. Cramoysan says that Nuance's reliance on voiceXML should help the company in the short term, since it is a proven, open standard with a large developer community.
Microsoft has chosen to go with the SALT (speech application language tags) standard, which Cramoysan says is "interesting but unproven." Given that, companies such as Nuance and IBM (IBM: Research, Estimates) that rely on voiceXML have a window of opportunity to shore up their positions before Microsoft's marketing muscle and vendor relations can push the SALT standard's adoption rate north.
Sign up to receive the Tech Investor column by e-mail.
Plus, see more tech commentary and get the latest tech news.