Where is telecommunications going?

Now here's a heady topic! I don't intend to write up a "take over the world" strategy here. Although maybe over time, that is what will emerge. I'm using today's blog to get some thought down in writing. As no one is reading my blog yet, who cares anyway?

VoIP
There seems to be a general consensus that VoIP is the way of the future. Companies like Skype have brought cheap (read free under certain circumstances) telephony to the masses. Whether the Skype model is workable long term is debatable in my opinion. Along with a few other people, I see some major problems with the Skype model:

• You can't make money from it, long term. You can't support millions of people for free, that requires specialist infrastructure;
• It's too hard for non-IT literate people to handle. For my grandma to use it, it's got to look, sound, feel and bill like a phone - i.e. it has to be a phone;
• It requires your PC to be on all the time and connected to the internet (not good in the security conscious, capped DSL world). Although voice codecs are always improving, my 1GB limit is not large;
• You can't carry your PC with you and wander around the house like a cordless or mobile phone - Bluetooth headsets are a start, but are currently limited to a 10m range.

An alternative to Skype that looks a little better is called Teleo. I picked up on it from Scoble's blog the other day. By redirecting from PC to mobile, they cut out one of the problems above. If they can redirect to mobile, why not a landline... etc.

Incumbent responses

Here's the kicker. What are the large Tier 1 telco's going to do about all this? They're not stupid, they know this is here and they are developing their own VoIP strategies. You can get IP PBXs and various enterprise level VoIP offerings already, but nothing mass market. The problem many face is that the per call cost of VoIP is perceived to be zero. i.e. the IP infrastructure is already in place and voice is just another form of traffic.

Note: So long as the codecs can handle this that's true. In high latency environments, voice traffic is still relatively good. High packet loss is more of a problem. The corollary of this though is that the conditions under which a simple VoIP service won't work are the same as those under which general internet performance degrades. i.e. if you try and break voice, you break everything else. A way around this is to offer Quality of Service-based (QoS) services to retail customers so you can differentiate voice from data, but then you come back to the whole cost question.

Commoditisation of bandwidth

Another challenge facing those with networks is that bandwidth is becoming a commodity. With the major fibre rollouts that occured in the late 1990's and early 2000's, there's a heck of a lot of dark fibre around. That means that if I want a 1Gbit connection from A to B, I can buy it, relatively cheaply. Third parties can also rent bandwidth for next to nothing. Who knows - maybe we'll see bandwidth on the futures markets before long? That said, it therefore costs large telcos with all this infrastructure lots of $$$ to maintain it for little gain.

So what's the answer?
Buggered if I know. If we assume that VoIP is the way things will go and that the large telcos want to hold onto this business (otherwise why are BT building the 21CN and migrating all their PSTN customers to it) then they need a business model for VoIP that works. I can't see them passing off a VoIP offering as a direct replacement for PSTN. On-net calls would need to be free and there goes a major revenue stream. However, maintaining that customer relationship is all important. Perhaps the answer is (once again) in services. i.e. "the profit's in the ink not the printer." Give them the capability for next to nothing and call charges go out the window. On the plus side, that simplifies the rating and billing side of things no end, which should bring some major opex reductions. Then partner with the content providers and charge them for that.

Services

I like the idea of pay-as-you-go video-on-demand over IP (PAYGVoDoIP - wow what an acronym!) I don't want to pay lots of money for the priviledge of not watching a whole lot of Sky channels. I'd prefer a service where I can buy a single movie for more or less what I pay at the video shop, but do it from my remote. I'd prefer to buy a program schedule that I set up myself (kind of like Tivo) and only pay for what I watch. DYNAMIC is the word.

If my telco could give me that on my phone bill, I'd be interested.

More later this is long enough.