GigaNaki

My trip to Taranaki last week gave me a chance to put faces to the gigabit speed test results I've been posting. I took the time to visit Primowireless and see what these guys are up to, what I found is a seven year old business that is thriving by filling the niches that others find hard to fill.

At the moment those niches are rural wireless, business fibre, residential broadband and hybrid solutions that are a digital version of the 'Taranaki Gate'. Kind of sounds like the ideal future oriented RSP for regional NZ, actually I think they are.

Primo is located a couple of hundred metres up a side street in Inglewood, from the outside there's two double cab utes and an office located in an ordinary house. Inside is busy chaos, there's account being done, an antenna being assembled for an install and all sorts of gear lying around.

Talking to founder Matthew Harrison, I hear a typical story, he and is mates were LAN gamers who started building ad-hoc LAN's for gaming parties at each other's houses. This is a common starting point for entrepreneurs in this space. He started out improving rural connectivity for people who wanted what the rest of us take for granted.

They now have over 50 sites all round Taranaki and provide an impressive service which has seen customers leave to try RBI services and then come back because the 'official' solution doesn't work as well as what they already got from Primo.

They are also keen to mix and match technologies and are free of the 'rulebook' constraints found in our larger operators. I was impressed with plans to bring historic Parihaka into the digital age.

But what I really wanted to learn was what are they doing with UFF's gigabit residential service and what might a final service look like?

I was wondering how they were doing this when there is no UFB in Inglewood itself. The answer is simple, one of there tech's Hanan already had 100mb/s UFB to his home, so Primo upgraded this to UFF's 1000/20 'GigaNet' service.

Then the testing began by testing the 'SpeedTest' results from hitting servers all over NZ, the results have showed a few interesting things, obviously local is fastest, the next is that interestingly 'Spark' servers produced the best results at other NZ locations, this I think is indicative of the fact that Spark (not Chorus) have provisioned the network that you would expect is required to service 50% of the broadband subscribers in NZ.

Primo themselves have access to multiple backhaul providers and are confident that Taranaki customers really do have access to the best broadband in NZ.

I was interested in Matthews thoughts on provisioning and pricing a residential gigabit service, something I hear from the big guys is really, really difficult. So it was refreshing to hear Matthews thoughts and how his innate knowledge of the industry gives him the ability to create a product model in his head in about 20 minutes.

I need to stress that this is a ball park, blue skies conversation and in no way reflects what Primo may finally decide to launch.

First of all Matthew understands the cost of his inputs, installation, UFB service, handover and backhaul plus provisioning of data.and having good enough user level equipment to actually be able to use a gigabit.connection.

First of all Matt recognises that his handovers will need to be upgraded to 10gb/s, remember this is a shared handover at the point of presence. He will then monitor the need to provision more backhaul as required which I think is sensible. 

What I found really interesting was Hanan's observation that other than the SpeedTest results you can't really tell the difference between 100mb/s and 1gb/s, the other issue is that most people then use WiFi that probably tops out at 50mb/s at best. So the case for 1gb/s is mainly 'overhead' and that is what you'd get with a house full of device equipped teenagers or you've got Hanan and his mates having a LAN party.

I also found Matt's view on data interesting he'd like to separate plans into the connection and the data, rather than offering 'unlimited' plans as at 1gb/s it is possible to go through a staggering amount of data. So options could include daily data plans or buying into quantities in the 'terabytes' rather than 'gigabytes'.

I found it fascinating that each speed test consumes about a gigabyte of data, this is due to how Ookla have designed their speed test, I've heard similar tales with people showing off their 4G cell phones and blowing their monthly data cap while trying to get bragging rights at the pub.

The other observation is about customer expectations as to what a 'gigabit' really means in that the fastest test to date is actually 975mb/s and you will not ever actually achieve the magical gigabit. So is it better to call this a 500+ or 750+ service, the same observation is true for 30, 50, 100 & 200mb/s service.

I'll blog separately on Matt's views on rural broadband.