[AusNOG] Netflix, AWS and Softlayer vs. Australia

[Mark ZZZ Smith]

----- Original Message -----
> From: Oz Nog <oz.nog at yandex.com>
> To: ausnog at lists.ausnog.net
> Cc: 
> Sent: Tuesday, 2 December 2014, 8:51
> Subject: [AusNOG] Netflix, AWS and Softlayer vs. Australia
> 
> Hi all,
> 

<snip>

> 
> So far the only downside with Softlayer is that they do not have a Sydney DC as 
> an alternative.
> 
> By looking at my tea leaves, I predict the following:
> 
> 1/ All latency sensitive and all bandwidth intensive apps/services will start 
> migrating back to Australia,

I'd be surprised if any latency sensitive applications ever left.


> fueling the pent up demand of all connected Oz 
> residents. AWS already made a crack, Softlayer opened the floodgates and Netflix 
> is about to bust the dam wide open. Where will you be when your ports saturate?
> 

This reads to me like you think there is going to be a whole lot of extra traffic. I'm not sure that is going to be the case. Of the three services/service providers you've mentioned, only one of them could possibly create a lot of extra traffic and that is Netflix. However, bandwidth intensive Video on Demand has been available in Australia for quite a long time from Youtube, ABC iView and SBS On-Demand. A number of other paid-for VoD services have been around for at least 6 months. So I don't think Netflix entering the market will make a dramatic difference in traffic profiles.

Local services might shift that traffic from transit to peering, which is a good thing, and might cause some networks to increase peering link capacity, of which the cost may be offset or completely paid for by the reduction on transit traffic. With an extremely "well peered' network you may be able to absorb that load with your existing unused peering link capacity if it is distributed enough.


> 2/ Local hosting providers are going to have a hard time competing with the 
> likes of AWS and Softlayer that operate at scale and have direct connections to 
> Telstra et al. Perhaps they can all become resellers or find some small niche to 
> occupy?
> 

Finding a niche is a better approach.

It seems to me that to win in a market you have to be either better or cheaper than everybody else - and of course to be cheaper, you have to be *better* at controlling your costs. So, in other words, all you have to be better (!). So an organisation needs to know how it is better or how it is trying to make itself better than other players in the market.

The thing that these large players are primarily doing better is reducing their costs by leveraging scales of economy of purchasing and utilisation of equipment, utilising people and utilising software. They are the 'ready to wear' or 'one size fits most' in the market.

If you're a small player, you're never going to be able to achieve those levels of scales of economy to be able to compete on price. So stop trying to compete on price! Find something else that the other players can't do or won't do that customers will value. 


> All in all, it's going to be an interesting new year.
> 
> In closing, heads up everybody! The data tsunami is coming!
> 
> Cheers,
> 
> Baz
> _______________________________________________
> AusNOG mailing list
> AusNOG at lists.ausnog.net
> http://lists.ausnog.net/mailman/listinfo/ausnog
>