My writeup of day 1 is here and my general summary of the rest is here. This is a post on one very specific subtopic which caught my eye at 3GSM - VOIP. Voice over IP is not something that looks like it would immediately threaten mobile operators since, at first glance, what it is doing is replacing the POTS line in homes and offices, but look a little deeper and it also begins to have implications for mobile vendors. As a result, while VOIP was not everywhere, there were plenty of VOIP bits and bobs on offer at 3GSM. The good news for the mobile operators is that VOIP is potentially an opportunity as well as a threat, this is better than VOIP has for fixed line operators where VOIP is simply killing off the voice cash-cow that most have been relying on. The bad news, however, is that it most definitely is a threat to the goal of increased ARPU that is the goal of every mobile operator.
The Opportunity
The obvious opportunity is the small one - this is the way that VOIP allows mobile operators to cut the costs on call trunking once a call has left the air interface. Sure VOIP does help operators cut costs, and no doubt cunning operators will be able to delay passing on some or all of those savings to their subsribers for a sognificant period of time, thus increasing the profit of each minite of talk time. However this is not, I believe, the major opportunity.
The major opportunity is for the mobile operators to be able to move in to the fixed line voice business. AT&T first attempted this when it bought up all those cable operators, but the current crop of mobile operators doesn't need to do this since broadband is already widely available. All they need to do is resell some VOIP service on top of the customers existing broadband. For example I saw a number of little players offering clients that, in combination with some cunning wireless LAN or bluetooth device, allowed a user to make and receive VOIP calls while at home and seamlessly make regular wireless voice calls when away from home. There were a number of varients on this scenario depending on the seamlessness required - at least one claimed to be able to manage handover transaprently without dropping the call, others were less transparent - the amount of special wifi/bluetooth/DECT gear required - at least one just required a 802.11 capable handset and any accessable 802.11 network - and so on. This is, in many ways, a very compelling story to subscribers. One of the problems with people wishing to bin their fixed line POTS service for VOIP is that in the event of an emergencey coupled with a power loss they have no way to contact emergency assistance. When the VOIP phone is also the mobile phone this problem goes away; for example if VOIP is down the handset just uses the regular wireless service as if it were away from home. Of course it is possible to argue that in a major emergency the wireless network might be down but that isn't much different from a tree taking out the phone line as well as the power line to a property. Other benefits for a mobile subscriber using the mobile branded VOIP is that other mobile features such as voicemail would be included for free, you really would have one device to use (and thus one quickdial list) and it would really be a phone - I have seen some people gripe immensely about using PC headsets with VOIP so that latter is bigger than you mught think. All in all this is a bonus to the wireless operator since he gets to add an additional charge per subscriber for very little if any equipment outlay.
The Threat
The threat on the other hand is fairly clear. As and when Wifi becomes reasonably ubiquitous and public wifi access costs drop to free or nearly so there may be a significant drop in the number of calls made over the regular wireless network because its cheaper to use the Wifi and VOIP than the cellphone. This of course will tend to lower ARPU. Perhaps worse many of the other dodges the operators wish to use to increase ARPU could be undermined - for example rather than using a carrier provided GPRS or 3G search or walled garden for web surfing subscribers could just hook into the nearest wifi access point and surf the full web instead. I believe that highish bandwidth data networking is likely to only serve applications that can be used while the subscriber is, essentially at rest and sitting down (at rest in an airplane maybe but relative to his surroundings at rest) which means that it is more suited to hotspot access than to the general coverage of the cellular network
For a whole host of reasons - simple geography for one - I can't see VOIP taking away the whole of the mobile phone addressable need and therefore I still believe that most people will continue to believe they need a cellphone, but what I can see is that the premium rate services, particularly the ones that supposedly drove the requirement for 3G being whittled away. A good example (TMF Nokia board readers will recognise this in part) are video phonecalls. Simply because of the fact that you want to set the camera up facing you at a fixed distance I believe that the vast majority of video calls will be made while the speaker is sitting down at a table. If the subscriber is sitting down at a table then all the overhead of seamless handover from one cell to another is wasted, on the other hand it is the perfect application to drop in to Starbucks grab one of their overpriced coffees and use their wifi hotspot to make the video call. Likewise for services like checking email, viewing movies and so on. Sure some passengers in a car might want to do these things and they would then need the cell network but I'd guess that 90% of the time when a subscriber wants to do this he will be in the sort of place that a wifi hotspot could serve - recall my pictures of laptop users in the previous 3GSM reports. The cost of deploying wifi hotspots is trivial compared to deploying 3G and you generally end up with more available bandwidth.