Back when I was a young blood at Datacom, I was one of the crew that attended an expo to help out with the Microsoft pavilion and a few other odds and ends. This was right at the very infancy of Telecom releasing ADSL in to the New Zealand market.

At the time, I asked a Telecom representative at their booth just when Telecom was likely to release a flat rate DSL plan, to which all I got in return was laughed at. Telecom believed that flat rate was not what people wanted. People wanted to pay for what they use. I just looked at him with a gleam in my eye and asked him if that was the case, why most other broadband providers around the world were offering flat rate broadband?

Technically, flatrate is cheaper for the ISPs. Data traffic accounting takes up significant processing. You have to provision some pretty heafty equipment that can accurately track the packets used, and you have to make sure you’ve provisioned enough to cover all your customers usage. This puts the cost of providing the broadband service up.

Further to this, data is significantly cheaper the more you use. Consider that sending one solitary packet across the pipe is a huge waste of available bandwidth. You are far better off consistently filling that bandwidth as much as possible, and thus getting a return on the investment sooner. Whether I have 1 or 100 users on my network, I still have to provision bandwidth to accomodate not only current users, but also potential future users. The sooner I use that bandwidth, the sooner it pays off.

So with this little bit of intro, I want to point you to Monday’s Herald article that very briefly discusses a new report about the cost of broadband and how our data caps are a significant issue for New Zealand consumers as well as make a little observation of my own.

All up, the top country in the sample was Sweden, with New Zealand fifth from the bottom at 22nd.

Keith Davidson, executive director of InternetNZ which commissioned the survey, said data caps were a serious constraint on the way broadband internet could be used.

“Telecom and the rest of the industry need to give up data caps as soon as possible, just as they had to give up per hour dial-up charging years ago in favour of flat rate access,” he said.

Data caps ‘serious constraint’ to NZ broadband use – NZ Herald

And this is very true. I have a 10gig/10gig plan. That is, 10gig peak usage and 10gig offpeak usage. I don’t use that full amount because I find it very difficult to keep track of it in the space of a month. Worse still, I have no accurate way of measuring my own usage because I cannot afford the processing required for me to do so properly. On top of that, the usage statistics from my ISP could be completely false for all I know. I have a general idea of what I might use in a day, but there are 6 computers that use that connection, so I cannot account for all of the usage.

Thus, I find myself being very restrictive about what I do actually use my internet access for. I’d love to be able to sit there and watch streaming video of lectures and other things that I think might be useful to me, but I find that I don’t do so until off peak times (through the middle of the night) because I’m not sure how much onpeak data has been consumed for the previous day.

I would love to be able to update all my computers the moment an update is made available for them, but I find myself staggering the updates across multiple days at off peak times because I cannot guarantee that my peak usage will not be low enough in a month to not put me over my data cap limit.

ISPs in this country, not just Telecom, all need to get a grip and come to a realisation that bandwidth is going to cost them just as much whether they use 1Mb/s or 100Mb/s. They’ve still had to provision as if they were using that 100Mb/s continuously. They cannot accurately predict burst times and so need to provision as if every millisecond was a burst period of usage.

Data is exceptionally cheap. Once the hardware is bought and configured, there is very little management required for a network to run. Monitoring a network can almost be automated to the point now where only a small group of engineers could manage and maintain an entire nexus with minimal effort. I myself (and most other admins) have implemented networks in the past that were self monitoring and could reroute themselves when one node went offline. I myself have single handedly managed a core network of a rather significant and high profile company in New Zealand. Note I said network, that does not include application servers, just the network they communicate across.

Even more evidence can be found in the fact that data ports on a switch are exceptionally cheap now. I present the competition between Cisco, Foundry, F5, Alcatel, Juniper and the many others out there as a good example of how competition in an industry does truly benefit the consumer by providing high end hardware at a reasonable price. Problem is that the telcos don’t want to admit that this cost is minimal. They want a way to charge significant margins for a product that really doesn’t offer any significant margin because once the hardware is paid for, 80% of any income from a network is entirely a profit margin. The other 20% covers engineer wages, administration and copper/fibre maintenance.

Services on top of the bandwidth are where the money is made. Phone connections and video services and the like are the current money makers. In the future as the industry grows, there will be even more services that we’ll come to use and be willing to pay for. But the bandwidth itself is worthless.

Even ignoring the Foundry FastIron sitting in my home, I have several gigabits of bandwidth available to me in my home that cost me nothing more than the network cards and cables. I can buy a 100Mb/s switch for less than NZ$30 now and if I really wanted, I could get an unmanaged 1000Mb/s switch for not much more than that. A managed gigabit switch gets a little more costly, but it also depends on the bandwidth of the backplane on the switch as well. Still, its not so expensive that I couldn’t afford to wire every single room in this house up with dual gigabit ports and use them all at full capacity 24/7.

Telco’s need to understand that the population isn’t exactly clueless anymore. The internet has made it possible for people to learn about how their internet (or home network) connection works. People are learning this stuff because they often have multiple computers they want to attach to their broadband connection, thus need to learn the basics to be able to do that.

The time is coming when I’ll be able to ask that Telecom representative at a trade show just when they’ll finally roll out true flatrate broadband and instead of laughing at me, they’ll actually be able to give me a real answer. Slowly the wind is changing and one day we’ll actually be on a par with the majority of the rest of the developed world when it comes to telecommunications.

But first, that data cap must go.