That's key stuff Bowden - the difference between the headline/marketing speed, and the actual speed. The highest "up to" figures sound nice, but it is the median and bottom-end figures that actually tell us more about where the nation's infrastructure sits.
The first point I'd like to make is that while upto 80 Mbps is advertised and generally offered in all fibre-enabled cab areas, quite a lot of lines aren't actually capable of receiving 80. Some arent capable of passing the 40. So though advertised in an area upto 80, the customer isn't going to order that if the predicted speeds will be under 40.
So in this situation, if taking a general look, it looks like the EU could have ordered 80 but ordered the 40 service instead. Then built on that general information it could be drawn that maybe the EU didn't need up to 80 speeds. Which as you can see in the example I've mentioned would have given off a wrong impression.
Yes - such choices do skew things, so you have to take account of these things.
The reverse can be true too, because ISPs don't just increase the download speed on higher-tier packages. For example, TalkTalk and Plusnet's current bottom tier only has 2Mbps upstream. If you want more upstream (and most lines can indeed get more), you'd need to upgrade to an 80Mbps downstream, even if you would otherwise be fine with a 40Mbps downstream.
I would be interested to see the average connection speed most customers get. I remember looking at one average speed chart and noticing that the average speed was below mine. I'm connected at 65 Mbps. So it must have been between 50 and 60. I can't remember exactly but it was on the thinkbroadband site.
So unless we can get data on the actual connection speed a line is capable of then we won't know the reason why he chose 40 instead of 80.
The average connection speed they actually get? Or the average they could get if they hadn't purchased a cheaper tier?
I've got a few charts that can help here; the first one (attached) is from Ofcom, detailing actual speed tiers for all technologies in 2014. It tells us that:
- the median
actual FTTC speed is 40Mbps, through package choice. Almost 50% of FTTC lines had both chosen a 40Mbps package and were getting just under that.
- the median
attainable FTTC speed may well be 60mbps (using simplistic extrapolation of the curves)
- the median cable speed was at the boundary between 30Mbps and 50Mbps packages (and can be assumed to be 50Mbps with current packages)
One of the BDUK projects (perhaps Staffordshire, or perhaps Shropshire) said that the average speed in their intervention area was now 50Mbps; if you assume they are talking about "attainable" values, and assume that rural BDUK lines tend to be longer than commercial ones, that BDUK 50Mbps figure could correlate with a national average of 60Mbps.
The second chart gives you the spread of D-side line lengths, which puts just shy of 50% of lines in the first two categories - ie less than 400m long - so a median length very close to 400m. Experience says that FTTC can get 80Mbps out to 400m ... but not with the higher crosstalk that comes with higher take-up. But actual speeds in the region of 60Mbps are not unlikely.
If BT were choosing to put vectoring in place, they could probably manage to get 100Mbps to 400m with ease - and probably to 500m, if the Eircom results can be extrapolated to the UK. Shame this isn't happening...
I think what holds up the deployment of FTTP technology in this country is money. But as we have seen from the non-BT Fibre companies, Hyperoptic and Gigaclear in particulay, that there is profit to be made when focusing solely on FTTP. As I've mentioned on this forum before, Openreach are majorly underfunded. In my opinion there needs to be a solution to increase funding, whether from BT being made to invest more in Openreach or that funding comes from somewhere else.
I think you're right, but I think there are a couple of facets that undermine the chances of investing more.
Openreach would be more likely to invest more per line if they could be sure of getting that investment back over time. Two years of £10 per month is all that Openreach can budget as income at the moment - or around £250. Ofcom won't allow for any longer minimum contracts; for Openreach to invest £1,000 per property, they need to be very sure that you will stay with them for much longer - and not suddenly swap to an LLU provider instead. The risk of not even returning the investment is too great if the period of return gets too far off into an unknown future. Risk is also added from competition from the likes of Sky - especially when they want to give away broadband aside their premium TV content - as does the shifting sands of Ofcom regulation.
I suspect the reason that we're seeing Openreach rollout in steps is because it lets them break the risk down into manageable steps, each with relatively limited time exposure.
