The subject of instant messaging came up this week (and the week is just getting started) via a call from an editor who wanted to talk quite broadly about this topic, particularly with respect to how IM is being used in business. In general, I am not personally a fan of IM, and rarely recommend it beyond purely personal use, for two reasons: it's not secure, and, even more importantly, it's not auditable. This is why many enterprises ban the use of IM at work - there's no trail of what was communicated as there is with e-mail. Given that IM is only a little bit faster on delivery than e-mail, albeit with a slightly more user-friendly nature, and given that if I'm going to chat with someone that I'm temporally coupled to I'm going to do it by phone, IM just isn't a good idea - at least for me.

But there's a bigger question afoot here, and that is which IM is going to survive. I'm assuming that Gen Y will eventually force corporations and indeed all enterprises to adopt some form of IM, a decision may eventually be required as to whether the protocols involved and the applications supporting them will be those of the Internet world (AIM, Yahoo Messenger, Microsoft Messenger, etc.) or the cellular world (SMS, EMS, MMS). Granted, both of these broad domains could (and likely will) last a good long time, and gateways can handle interchange between the two. But if we are to have secure, auditable IM, then the client is going to become the most important part of the solution. And universal clients are seldom common in any form of communication.

It can be argued that the cellular branch of messaging has most of the mindshare for mobile users today. All cellular systems, including even the recently-departed analog services, have some form of messaging included, with the objective of enabling revenue-generating traffic that has some tolerance for latency on what is otherwise an isochronous network - thus maximizing revenue by filling in otherwise dead air. Pricing can be used to influence which type of traffic a given user pursues, and can thus allow a carrier to manage capacity while enabling a slower, ore economical network buildout as demand grows. The Internet world has no such constraints, as everything there after basic access is free. Want videoconferencing, for some the ultimate in IM? That's free on the ‘net, too. At least for now.

It can also be argued that, as most cellular handsets now provide a gateway to the Internet, the cellular world should simply adopt Internet messaging protocols and we can be done with this discussion. They will not do this, of course, for purely economic reasons, and also for a few relating to politics and control of the customer. And it can also be argued that, because gateways exist and continue to be enhanced, no choice will ever really be required. This put the focus where it should be - on the messaging client application, not the network, and on the server supporting that client. All we need is security and auditability, across multiple messaging protocols, on a good number of mobile devices (and PCs), and maybe integration with e-mail and other communications services, and my objections go away. Wouldn't that be cool? And useful, too.