What's Your Value? It's Just A Phone, Stupid.
Recently, we had the exhilarating experience of undergoing a merger. In the interim, we were able to set up a fully functional intranet, a speech-based automated attendent, and a virtual Exchange server in one day. Take-away: this used to be hard, but things that are useful get commoditized.
If past is prologue, then consider the case of VocalTec Communications. VoIP veterans out there may gasp in nostalgia. Just eight years ago Vocaltec was an Israeli startup, beloved by Pulver, and in at the creation. It owned the equivalent of the media gateway, heck, tit was a media gateway, ('scuze me, still are). But back in the day, VocalTec was seen as the market leader, chased by other pure plays (Vienna, Clarent, Applio, some of which are still with us, some not).
Today, having given birth to the VoIP industry and with it's first-in advantage well behind it, VocalTec is a $15 million company. Its stock price as of this writing is around 50 cents.
Who sells the most VoIP ports now? Cisco is on the short list, Nortel and Lucent are up there. And there's a whole slew of new players feeding off of massive greenfield markets in China, India and elsewhere. That's the Rewind part.
Now let's hit Fast Forward, remember: past is prologue. By the end of this year, we predict both Cisco and Intel will be shipping really acceptable VoiceXML browsers, integrated with an incredibly rich mélange of fine-grained call control stacks, bundled as a middleware layer that sits on top of the industry's most ferociously cost-effective media server hardware at under $200 a port - make that $125. Protocols available (basically) for free include SS7, SIP, MRCP, and, when it's baked, CCXML. APIs for leading ASR engines will be there for the taking.
Because of MRCP, license managers will be a quaint anachronism. The word is commoditization. We all talk about it, we intellectually know it's as inevitable as Moore's Law, but all too often we still think like VocalTec. What are the symptoms? Well, if you use the term "ROI" more than once an hour, you probably have the disease.
Remember how much VoIP was going to save us in phone bills? Guess we missed the part about the PSTN getting commoditized. That is symptomatic of the conundrum facing the speech solutions provider today. Unlike the Web, say, five years ago, customers are not sitting around saying, "we got to get VoiceXML-based ASR up tomorrow, and we'll pay whatever it takes." Because if they were, they'd be taking out the checkbook to pay the phone equivalent of ATG or BEA or BroadVision ridiculous amounts of money for what today we all see as something we can get from Tomcat for very little money - a basic capability to run XML scripts.
There is no similar imperative with respect to speech, after all, it's just a phone, stupid. Nuance has been conducting awareness studies among enterprise respondents lately and finding that around 50 percent fail to understand speech-related ROI models. While I personally find this surprising, let's assume it's that bad out there. That means it's time to roll up our sleeves and get to work, explaining the value of lower abandons, higher resolution rates, lower TCO on application maintenance -- right? Maybe. Or, maybe those customers just don't want to think that hard.
If the argument rests on business calculus (or at least algebra), it starts with an ROI formula. The numerator (revenue lift and cost savings) requires substantiation in the form of financial reports that may or may not exist within the call center. The denominator (investment) contains a dizzying array of speech components and capitalization of professional services. Clearly a simpler approach is called for. Therein lies the appeal of Intel or Cisco bundling it all up and delivering as part of a router, app server or media server -- real cheap. It's just a phone, stupid.
Closing note: Hopefully, I have enraged some readers out there. If so, let me know.
Mark Plakias is a partner and senior consultant for The Zelos Group. He can be reached at (212) 366-0895.