Network Insight Institute

Chris Atkins

Chris Atkins, Partner, Comms & High Tech Strategy, Accenture

Today I am going to talk about what the numbers are saying, and give you just a bit of a perspective on how to think about the implications of some of those things. 

To summarise my main points:

[SLIDE]

First of all, data from the OECD shows the uptake across the first 10 years of a range of communications technologies. What the red line shows is that the uptake of broadband has been faster in its first five or six years than mobile, than narrowband internet and ISDN.  Incidentally, neither the OECD nor the ITU class ISDN as broadband technology.  The sorts of numbers we are seeing here are penetrations at around 10 per cent of internet subscribers now.  The growth in 2002, the last year the OECD data runs to, was 72 per cent annual growth.  So it is growing very fast, is available in 82 countries, and to roughly 80 million subscribers around the world.

[SLIDE]

Australian broadband penetration is relatively low.  What we see here is an OECD graph of penetration expressed as a percentage of population.  Australia is sitting down at the bottom end of that set of data, which are broadly OECD countries; although there are some non-OECD countries in that data. So the uptake is relatively low. Most people have attributed that to the price.

In terms of data on price, it is very difficult to do good price comparisons because of the different speeds available in the packages, the different throttling mechanisms and so on.  There are two different ways that people look at price. One is a normalised 100 kilobit per second access, and the other is a normalised one megabit per second access.  This is the ITU's definition. The only thing that I can say about it is that at least in this regard we are trying to compare apples with apples. Australia is in the middle patch in which we are very comparable to a whole load of other countries, so I don't think price per se explains the penetration.  This is normalised by purchasing power parity (PPP) exchange rates.  So it takes into account the ability to pay in a given country:  it is just not exchange rate driven here. 

What else do the numbers say?  As I said earlier, I think they don't actually explain very much.  I did a series of correlations on the data.  I could get data for GDP per capita at purchasing power parity, telephony penetration, mobile penetration, internet penetration, monthly subscription levels, download speed, maximum download speed, price per 100 kilobits, prices as a percentage of income, and the population density in urban centres. The last item was considered because one of the explanations of the high penetration in Korea is the fact that the population is so concentrated into a couple of big cities, and then within big tower blocks in those cities. I couldn't get the sort of tower block concentration for other countries, as the data doesn't exist. The only statistically valid correlation that I can get out of all that data set was actually against internet penetration.  It explains about 24 per cent of broadband penetration, and the residuals are pretty well distributed. So it is statistically sound Is it causation?  Who knows, because this is statistics, but of all that data that was the one where I could get a correlation.

One of the things that I would have liked to have tried but couldn't get data for in time was first year costs as opposed to monthly ongoing costs.  That is because there is a big up-front cost to get connected to broadband.  It would be very interesting to accumulate the connection costs plus the ongoing costs in some sensible way and to have a look at that.  One of the things that I think would be very interesting to look at in there would be the differential price between narrowband and broadband. If there were a small differential, then my hypothesis would be that it would be much more likely for people to connect to broadband.  I haven't seen a very robust analysis of that anywhere around. It would be quite an interesting piece of work to do. 

So that is what the numbers say: not a lot.  However, there is a little insight you could take from this correlation: that if you do want to drive broadband penetration there is a sort of indirect route to drive it, which is via narrowband. That is quite a profitable track, because narrowband access has almost certainly got a higher margin at the moment than broadband access.  You could actually drive your narrowband population profitably, and then migrate people over onto broadband possibly through some sort of pricing arrangements if the hypothesis on that is right.

As I explained, a lot of questions are being asked, and they are fairly basic questions, around the world by a whole load of content providers and subscribers. We are putting together a team that is having a look at this around the globe at the moment.  Should you actually drive broadband adoption? Or should you actually manage the growth of broadband a little bit more carefully? The thinking behind that is obviously about the profitability of narrowband compared to broadband.  Then, should you, as a service provider certainly, adopt either an access-only or access-plus-content strategy? If you adopt a content strategy, how should you do that, and what should your content strategy be? If you are then in the broadband game, what is it you should do to drive uptake?  Do you do flat rate packages or a lot of extras, or value added services that you can build onto it?  How to approach device convergence involves a lot of technological questions, such as: should you have a device as effectively an independent service so it can go on wireless or via cable modem, or via DSL? Certainly a lot of people are asking that.  A related technical question is around what is going to follow DSL.  And a less recognised factor than most people pay attention to is the huge complexity and cost of providing service in broadband at the moment, and if you are to drive broadband growth profitably how you deal with that complexity and cost to make a service that you can make money out of.

That is just a quick view of the sort of questions that are being asked around the world. How you think about those questions depends on the perspective you take, because it is different if you are an incumbent, or if you are a challenger telco. It is different again if you are essentially a broadband provider as opposed to a telephony service provider; and I can think of a whole load of other things it would be different for than some other things. So how can we think about that? The two biggest unknowns in this are the following two points: should you drive broadband growth or should you manage it?  and should you adopt an access strategy or an access-plus content strategy?  The perspective I have taken in the next slide is one of a service provider who actually has that option. 

[SLIDE]

On the vertical axis you can see just a slow drive of broadband, a well-managed penetration of broadband, and at the top I have a more driven broadband adoption. On the left I have an access-only strategy and on the right I have an access-plus-content strategy. I have given them some names that I hope try to get across the picture of where I think we are in each of those quadrants.  What I am trying to do here is, if you believe the world is like one of these sets of bullet points, then you are probably in somewhere where you want to either drive or not drive broadband and adopt an access or access-plus-content strategy. 

Just take the bottom-left for example, which I have called 'Managed Profitability'. If you think there is no single 'killer' application for broadband, if you think that broadband roll-out costs and operational costs will remain high (and the cost of complexity is going to be very difficult to bring down) and if you think as a service provider that broadband will have some positive impact on churn in a bundled offering; then probably what you should try to do is manage the growth of broadband fairly slowly and adopt an access-only strategy.

Looking at the top left, if you think there will be a demand for specific broadband content, if you think that broadcast and interactive contents are going to converge or are converging, if you think that content delivery and management require very tight integration between the content side and the access side, and if you think you can bring the roll-out and service costs under control over time; then you are probably in an area where you would like to drive broadband adoption, and you probably have to think about your access and content strategies and therefore understand how you are going to do content. Such a position raises a whole lot of questions like: should you go for exclusive content deals or not; and what sort of content should you be providing?

On the other diagonal, if we are back on the left hand side again, now consider speedy access. If you think there is no single killer content application, if you think there is a much more positive impact on telephony, and if you think you can manage broadband roll-out and operational costs, then you are in a space where you probably want to drive penetration but probably not get into the content game.

Finally, looking at the bottom right, if you think that most broadband content demanded will be more broadcast-like or entertainment-driven and delivered that way, if you think that most interactive content is going to be inherently narrowband-ish, if you think that the roll-out service costs will remain high, or if you believe that there will be some dramatic development in encoding and compression technology; then you are probably in a broadcast world where you would manage your broadband adoption fairly slowly, but have to adopt some sort of content strategy.

«  Back

^  Top of page

Network Insight sponsors include: