John Linton
I was mulling over the various bandwidth contract renewal options and new proposals over the last few days. As with all such processes you are inevitably reminded of the really bad decisions you made when you signed the previous contracts locking you in to spending far more than you should have over the past two years or so. What I thought were good decisions obtaining good pricing turn out to be really bad decisions causing significant financial pain over a long period of time.
So, I suppose, the good news is that Exetel can, in the next few months, begin reducing its spending on IP bandwidth by around $130,000 a month when all current contracts are either transferred to more realistically priced competitors or the current suppliers work out that losing $A500,000 of our current spend (and the obscene profits they have made while ensuring that Exetel barely managed to stay 'alive' over the past two years) is more problematic to them than providing the same bandwidth at 30% less than their current pricing and at least retaining a big chunk of business that is still very profitable to them - something we know for a fact having gone to the detailed trouble of agonising over whether we would commit to buying an STM16 direct from Southern Cross.
We are still favouring a 'joint venture' with another company who is interested in buying SX STM16s but needs a 'major' customer (amazing to think that term can be applied to Exetel) to commit to a large percentage of the first STM16. That pricing would deliver a cost to us of around $A150 per mbps within 18 months which is around the time that the new PIPE cable may become available at prices that will be interesting to see at that time.
Bearing in mind my less than brilliant decisions of the past in selecting bandwidth contracts I'm hesitating more than usual over these decisions and certainly won't be signing anything longer than a 12 month contract. I'm more than a little sick of Exetel's personnel working so hard and efficiently to ensure that other companies have an easier life by overcharging small companies such as Exetel. Four years ago it was very much a case of 'beggars (start ups) can't be choosers' but things are a little different in January 2008 than they were in January 2004.
One sobering thought is that the amount of bandwidth 'consumed' by an Exetel user today is almost 2/3 more than the average of a year ago. This is partly due to the increase in ADSL2 users and the general increase in usage of all types of customers. We have managed to ameliorate the effects of this growth with the Allot 2250 smoothing P2P demand which, along with the off peak strategies of the ADSL plans, has allowed an incredible level of efficiency in band width usage. Similarly the P2P caching has made a difference to the peak usage period by providing another 200 mbps at the most heavily used peak times and up to another 400 mbps in a small part of the off peak time.
Even so the usage per customer is way up on a year ago and continues to climb. The peak delivered bandwidth in January 2008 has been 1.950 gbps compared to the peak in January 2007 of around 1.0 gbps. Apart from any other consideration we are having to add a third GigE link to the border routers as we will need more than 2 gbps before the end of February.
The decisions aren't becoming any easier and the temptation to 'play it safe' is significant. On the other hand 'playing it safe' will cost around $A80,000 extra a month and I'm not sure I want to do that for another year. $A960,000 in Exetel's bank acount would be money that can be spent very usefuly.
.......but then there's the risk of a 'joint venture' with a company smaller than Exetel........though, come to think of it, Exetel was a company with no track record four years ago so have I got some false views here?
There must be a better way to spend Australia Day - and I notice I missed out in the honours list again!