I was chatting to a guy from a Big Operator yesterday, and during a chat about SMS and instant messaging he mentioned something I'd never considered: the impact of presence on radio networks.
Consider a typical user who has 10 calls a day and receives 10 text messages; this activity results in his handset being contacted on the local network 20 times in a day (my operator chum referred to these contacts as "pages"). Now consider what happens if this user has 10 buddies in an IM list and is alerted whenever they drop offline or come online: suddenly the number of pages rises to hundreds per day.
This rise impacts both battery life of the device itself, and traffic on the radio network. The operator in this case did a couple of studies and experiments which confirmed this scenario, and in fact painted a gloomier picture around networks collapsing once IM-enabled users hit a certain (low) threshold.
This says to me that IM in its current form won't threaten SMS revenues for a good few years: a bit like Wi-Fi, whilst we can see it starting to creep in at the edges, it's not ready for mass adoption and won't scale up in the way that SMS on current GSM/3G networks does.
What are possible solutions? Limit the number of presence-aware contacts in individual address books, to reduce the network effects of all those presence updates; allocate a new radio channel for presence information (but that'll take 5-10 years to sort out); and IM vendors optimising their products to be sensitive to the needs of mobile networks could help a little - in the short term. And of course, if unchecked IM threatens operators abilities to offer their core services - voice and text - over their networks, then we can expect to see them clamp down on it (presumably taking a fair amount of flak from the lets-replace-operators-with-wifi brigade).