Speaking Engagements & Private Workshops - Get Dean Bubley to present or chair your event

Need an experienced, provocative & influential telecoms keynote speaker, moderator/chair or workshop facilitator?
To discuss Dean Bubley's appearance at a specific event, contact information AT disruptive-analysis DOT com

Tuesday, October 25, 2011

Implementing native RCSe in devices brings operational risks from emergent behaviour

I'm at the Rich Communicatiosn conference in Munich today & tomorrow, listening to presentations on RCSe (asd asking critical questions).

I'll do a full summary another time (thus far: operators are more convincing - and more restrained - than vendors). But one point has struck me:

The big push from GSMA and the "Group of 5" lead operators is around natively embedding RCS capability into future mobile phones so that it "just works", and appears to the user in a similar way to today's SMS and voice diallers. On other devices, aftermarket apps for RCS may be available for download.

I can see the "elegance" here in native capabilities, but I think there is a huge risk which has not been identified. Downloaded apps can be updated "in the field" relatively easily. HTML5 apps can be updated even more simply as the functionality resides mostly in the cloud, or through browser plug-ins.

The risk is that of "emergent behaviour". Consumers and companies tend to find unexpected ways to use services - sometimes providing benefits and value, but sometimes creating problems. For the voice dialler, many people have started to use "missed calls" to notify friends of things, creating signalling load and congesting voice switches, but creating no revenues. SMS usage was essentially an "emergent" behaviour which drove huge revenues. SMS spam, however, has been a downside. Consumer use of BlackBerry Messenger has been largely emergent.

Nobody knows what the emergent properties of RCSe might be. They might be hugely addictive and valuable services, or they might cause huge problems. They are unpredictable and essentially untestable. There may be unexpected bugs or weird effects when users start behaving in a particular way.

One of the things we've seen in recent years is similar issues with Facebook and other online services, which have created risks such as privacy leaks, issues around personal safety / stalking and many other undesirable side-effects. When these occur however, they can be quite rapidly rectified, either by changing the web application, or perhaps by issuing an OS patch, or an amended or bug-fixed smartphone app.

It is far from clear how a "bug-fix" or emergency upgrade / alteration of RCSe clients could be achieved if the functionality is hard-coded into phones. Some might be updateable via FOTA (firmware-over-the-air) upgrades, but that is unlikely to be feasible across the board. Apple can update iOS via iTunes - but no similar mechanism will exist for RCSe.

The problem is that this risk is essentially unquantifiable - emergent behaviours are inherently surprising. I think that RCSe needs a clear and well-articulated strategy for "rapid response" if something untoward happens. This is not just the risk of creating network load either - there is the potential for reputational and legal risk as well. This needs to be considered and managed much more effectively than I've seen so far.


No comments: