Wednesday, September 12, 2007

Is Voice the Killer App for IMS?


You have seen this story before: a new service rolls out and providers look for the "killer app." Then it turns out the killer app is something people already do, but the innovation allows them to do it in a new way, or maybe a better way.

To some extent, voice is a bit of that sort of thing for broadband Internet access, as email was something approaching a killer app for dial-up Internet access. Though the initial "killer app" for broadband was fast Internet access, voice becomes a very important incremental value.

"We are seeing a pattern in Europe of VoIP being delivered by companies that control the broadband infrastructure," notes Stephan Beckert, TeleGeography analyst. "It's an add-on feature to broadband."

So what is the killer app for fixed-mobile services? It's voice again, allowing legacy providers to hang on to more of their fixed-line business than otherwise; allowing mobile providers to displace landline traffic with mobile; or new providers to displace business phone systems.

So what is the killer app for IP Multimedia Subsystems? Wouldn't it be surprising if it turned out to be voice?

So what's the logic? Assume wireline carriers might lose as much as $13 billion in annual revenues by 2011, in part because 34 percent of U.S. households might elect to go "mobile only." So enter IMS, allowing mobile users to take advantage of cheaper Wi-Fi-based calling over their broadband lines.

Assume the landline carriers then lose just $8 billion in revenue to cord cutters. That's a $5 billion annual revenue stream. So put that in perspective. All U.S. multichannel video providers put together earn about $4 billion a year from pay-per-view and video-on-demand services.

So if wireline carriers just prevent landline erosion, they make more money than the whole U.s. VOD and PPV providers put together.

No comments:

Costs of Creating Machine Learning Models is Up Sharply

With the caveat that we must be careful about making linear extrapolations into the future, training costs of state-of-the-art AI models hav...