Friday, October 20, 2006
Operators are clearly taking a very pragmatic line. Their story is less about an all-embracing architecture and vision of the future, and more about solving practical networking issues for voice evolution and FMC. There still remains considerable tension between the IMS standards vision being put forward, and the commercial reality of metered voice revenues heading down fast, sometimes to zero. Peer-to-peer networks are "crossing the chasm" and equally significant is "give it away in return for eyeballs" from Google et al. Many participants picked up on this, and the vendors appear to be coming in well below the bar being set by their operator customers. Without a vision of life beyond the walled garden and closed services, IMS faces an eternal uphill battle.
As always, the definition of new services remains a problem. This begs the question of whether the industry should be looking for new IMS services. If there was market pull, surely the services would already have been defined, with pre-IMS (or non-IMS) implementations widespread. The telecom industry may have traditionally been technology-push, but this is the first investment cycle where domestic broadband is widespread in developed markets. Standards for applications (rather than the technology platform) are even more fragmented, where they exist. Who should really define how a multimedia ringback tone should work? And what's the incentive for every carrier to support it?
The conclusion is that IMS is best deployed offering additional features to existing voice and messaging services. The accusation that IMS is technology pull can be answered if its scope is limited to existing successful products (voice and SMS), and their evolution. Some attendees were confused by RedKnee's pitch, which focused on identity and profile (accessed via a service architecture) rather than session control. Perhaps RedKnee and their customers are hinting that the value lies somewhere other than a "control plane" which no longer offers competitive advantage in the mass market? Or, indeed, that identity services are the new control plane?
The industry seems to lack a clear insight as to what the user needs are, or a philosophical compass that will help understand why certain services are successful. A good example of this is presence, which is normally part of an availability solution. The naive approach is that the fixed IM presence should translate straight to mobile. But if presence is really a sign of availability, and thus a signal of "I'm at my PC, not eating dinner or bathing the kids", there's no certainty that the model will translate to mobile. Instead, you would build a different availability model -- maybe a "request to call me" function instead.
A deeper concern, raised by Dr John Waclawsky, Chief Software Architect of Motorola, is that the real value is shifting from connecting people (which the Internet does very well too) to connecting personal devices (where the Internet is relatively weak). If my light switch wants to talk to my lightbulb, or my thermostat to my boiler, why does an ISP need to get involved? If he is right, IMS will arrive just in time to miss the next great wave of computing technology.
IMS standards are already fragmented, with different initiatives coming from the fixed and mobile worlds, as well as specific vendors and customers pressing ahead with partial implementations to solve immediate problems. The situation on handsets is cause for concern. The network may be ready, but it takes time to penetrate the handset base and software configuration management of an unproven technology is hard. How will carriers manage handset software updates? What are the real advantages over "pure SIP", which is already widely deployed? What if the user is roaming onto their home broadband connection and QoS guarantees no longer apply?
There is considerable uncertainty and disagreement as to what the role of IMS is in the Enterprise market (if any). The vendors are pushing this as a possibility; members of the audience were pushing back equally hard.
Perhaps the biggest problem is simply one of marketing and perception. IMS needs to be more clearly positioned as what it is: delivering the IN vision on IP technology in a way that more plausibly enables feature development using commodity IT tools. Some parts of the voice and messaging business, like premium rate calls and short code SMS, are doing very well. No Internet player is coming close to threatening these revenue streams. IMS looks likely to be the underpinning technology for these in future.
One of the most telling (if brutally blunt) comments came from the Voice & Messaging breakout of the parallel Telco 2.0 brainstorm:
“It is interesting to see the huge gap between the people in the IMS conference and the Telco 2.0 conference. We need to ensure the engineers get brought down to earth and listen to the market - I think many are them are still on planet Zog.”
Ultimately, IMS architects and network equipment vendors can't be expected to solve the structural business problems that operators face. As the sages in the room pointed out:
"I think the main advantage that internet start ups have is that they don't have the money or the time to continue talking without acting"
"This is about on industry that need to re-invent it self. ... IMS alone can't address a systemic problem with Telcos as currently structured."
The Editor, IMS Insider