Relentless Specialization, Jacques Ellul and Schnabel's Pit Bull
Two recent conversations about voice recognition in telephony networks have me musing about the risks of technical evolution, captured by the idea of "relentless specialization." An insight into this somewhat antiseptic phrase is found in French philosopher Jacques Ellul's work La Technique.
According to Ellul, "technique engenders itself." We analysts euphemize this, calling it "focus." Some might call it technology in search of a problem, but the search is really for more technology that looks like itself. We call that "interoperability."
The scary part of standards evolution is the quest for end-to-end purity. The constant danger (I'll get to the pit bull later) is that focus becomes myopia. We can't see the forest for the trees and perfectly obvious opportunities become hidden in plain sight.
A recent conversation with our friends at Tellme provides a case in point. This is a company that from day one has been forging a new vision of telephony. They call it "Dialtone 2.0," a phrase that captures the convergence of black-phone telephony and software-driven Internet technologies. The odyssey has been a fascinating exploration of new caller experiences, new business models and an intriguing assemblage of performance-enhancing procedural and platform solutions.
So what's the "big new thing" at Tellme? It turns out to be a 20-year toll-free feature set generically referred to as "network announcements." AT&T introduced this as Call Prompter in the mid-80s. When toll-free portability was announced in 1991, MCI went into the equivalent of the touchtone space race to introduce more advanced DTMF-based toll-free call routing, which it christened ECR (Enhanced Call Routing). I specified ECR for a client once, way back when, because we wanted to deflect customer-service calls from HQ to a service bureau and weed out order calls for a different service bureau. There was no integration with any of our PBXs and interpreting call detail was the equivalent of reading the I Ching. Also, changes took a few weeks provided the stars were in alignment.
Network announcements address a fundamental need by heading off calls before they are misdirected and dropping them off at the best of multiple locations. The last time we checked, better than 35 percent of the call center organizations we surveyed reported using some version of network announcements. That's big money, because these are expensive minutes and the toll-free product management organization—one of the biggest and most ossified in any IXC—has made these features incredibly complex to use. And there haven't been any significant innovations in this category for years.
What Tellme has discovered—hidden in plain sight—goes something like this. Tellme has customers that are buying into the voice applications network scenario: outsourced voice recognition telephony platforms in the network and homegrown VoiceXML applications running in Web servers on premises. But before callers reach this new-age solution, they pass through the legacy network-announcements layer, which features antiquated touchtone technology and requires a team of six people to update a 15-second bit of audio. There's also zero integration with the outsourced Tellme platform.
What's wrong with this picture, they're both in the network, right? Tellme (presumably others as well) looks at this situation and says, we can do this better, cheaper and totally integrated with the rest of the call processing and business logic behind it. (By the way, current expenditures on this piece of the toll-free market are running in the hundreds of millions annually.)
This is a perfect example of a market opportunity in plain sight that relentless specialization will not capture. From the specialist's perspective, this is just legacy that needs to be replaced. By taking a larger view, however, Tellme is using VoiceXML technology and its infrastructure to deliver real value to multi-site call center organizations. This is good business.
Contrast this with a recent conversation involving a well-known SIP specialist, who responded to a question about scalability by saying it was an "edge problem." OK, I can deal with that I guess.
That brings me to the pit bull. A friend of ours has a well-tempered, medium-sized dog; let's call him "Legacy" to protect his identity. Legacy is on his leash, walking along and minding his own business when a pit bull named (you guessed it) "Relentless Specialist" slips its collar and turns Legacy's front leg into hundreds of little "bonelets." Several months later Legacy is starting to walk again and the vet's bill is approaching $10,000. The good news is that Relentless Specialist belongs to the incredibly successful SOHO artist, Julian Schnabel. So, in this case, money's not an issue.
Will your call center implementation be so lucky? We're not against relentless specialization (my colleague, Rachel MacAulay, informs me pit bulls are very sweet, so no mail please). But if you're going to unleash this kind of technical intensity, first take a minute to look up the street.
Mark Plakias is a senior vice president and managing director of communications and infrastructure for The Kelsey Group. He directs research relating to speech, operator services and wireless media technologies and applications. He may be reached at (609) 921-7200.