Australian broadband speeds slow: Akamai Report

Australian broadband speeds slow: Akamai Report

Summary: A new report shows Australia is in the bottom seven OECD countries for broadband speed, and heading backwards.

TOPICS: Australia, Broadband

Frankly, it's an embarrassment: The latest Akamai State of the Internet Report (just released, but for Q4, 2012) shows Australia with average broadband speeds of just 4.2Mbps. Out of all 34 OECD countries, only New Zealand, Italy, Greece, Chile, Mexico, and Turkey fall beneath us.

(Image: Phil Dobbie/ZDNet)

Worse still, we've seen average speeds slide over the last couple of quarters, with a 23 percent year-on-year drop. Meanwhile, most of our stable mates at the bottom of the table have at least made some headway.

It's this lack of progress that, sadly, makes Australia stand out. Countries like Spain and Russia have leap frogged us, with annual growth of 24 percent and 34 percent, respectively. The danger is, of course, that with the National Broadband Network (NBN) still years away — in whatever form it will take — we'll stay behind the pack for some time to come.

In fact, it's our lack of fast broadband that makes the difference. We keep the same company at the bottom of the list when it comes to OECD countries, with the lowest proportion of connections greater than 10Mbps. Luxembourg pushes us one spot up the ladder, but we're still 28th out of 34.

(Image: Phil Dobbie/ZDNet)

It seems that most of the rest of the world is making good headway in getting more people connected to faster broadband. And we're not. The UK has seen a 129 percent annual increase in the proportion of high broadband (greater than 10Mbps) connections; in the US, they're up 90 percent. In Australia, penetration has fallen by 56 percent.

Yet, we are as keen as anyone to use the internet. The report recorded 8,631,783 IP addresses for Australia — 0.4 per person. That's more than the UK or the US, and only a little behind South Korea, which, with an average speed of 14Mbps and 49 percent of connections over 10Mbps, clearly leads the OECD.

Tucked away in its own corner of the world, Australia can only dream.

Topics: Australia, Broadband


Phil Dobbie has a wealth of radio and business experience. He started his career in commercial radio in the UK and, since coming to Australia in 1991, has held senior marketing and management roles with Telstra, OzEmail, the British Tourist Authority and other telecommunications, media, travel and advertising businesses.

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.


Log in or register to join the discussion
  • NBN the cause of stagnation?

    Surely the reason that Australia is standing still (and therefore going backwards on world rankings) is because investment dried up on 9th April 2009 when the NBN project was announced, and it was clear that the NBN would be the monopoly wholesale network. With NBN fibre passing only about 0.5% of households the higher speeds of the NBN are unable to influence the statistics. It's going to take at least a couple more years before we make any headway on those rankings (and then only if the FTTP version goes ahead AND the rollout finally gets up to speed).
    • in the short term

      the NBN has done the opposite of its intended effect. the slow speed with which it's gotten moving has made it into a double edged sword. if the latest news is right, hopefully it won't be too much longer till it really starts to roll out.
    • Don't get reality get in the way of bagging the NBN

      Since the announcement of the NBN rollouts have actually increased. Telstra in it's greed to get paid for more customer migrations has finally been rolling out tophats all over the country. Tophats that people had been screaming for them to rollout previously.
      For our speeds to be dropping even as we are rolling out fibre the copper must be deteriorating in Australia at a rate higher than other countries. Maybe the 5 minutes to midnight Testra said of their copper network is true no matter how the Liberal fanbois say it isn't.
    • Monopoly

      Networks of this kind and size are a natural monopoly. No company other than Telstra (really the government, back in the day) has built a last-mile network in Australia for a very good reason: there is simply nowhere near a business case, nor will there ever be, to build a competing network like Telstra's. This is why the Internodes and AAPTs have stayed on the periphery. Hence the decision taken by Labor for the NBN to be government-funded like its predecessor, the copper telephone network. The Liberal Party ideology of "just get private enterprise to do it" muddies the waters because they know this will produce local monopolies like in the USA and Canada. The only alternative is forced competition at the retail level like what we have had since the Telstra sale in the 90s. For the past 20 years there has been a lot of bashing of heads against brick walls and it will only end when we acknowledge the monopolistic nature of this industry.
    • You really need to check the graphs

      According to the Akamai report, Australian speeds have continued to increase gradually. There might be a minor stall in 2012 but nothing that explains the horrible standings internationally.

      The main failing of Australia is the meager 3.84% "High broadband adoption (>10 Mbps)", which compares to first world countries like Korea (49.41%), the US (19.50%), Russia (7.15%).
  • Copper to blame

    NBN roll-out cannot make speeds DROP.
    Slowing investment (not there was any before) could lead to us fall behind in the rankings, but this cannot in any explain an actual drop of 23% in average speed.

    The simple fact is that there are more users per household sharing the same congested copper cable pair.
    But Tony will fix this after September by cancelling the 1Gbs FTTP roll-out and replacing it with a 25Mbs FTTN record which will perform so much better than the highly congested 24Mbs ADSL2+ system we currently use.

    Yes sirree, that extra 1Mbs will make all the difference. Of course, to be fair, at some time in the future that is planned to be doubled to 50Mbs, probably by doubling the number of nodes I expect.
    • As Well

      Note data Volumes have been decreasing
      3/4G were delivering good speeds untill customer numbers grew, congestion reared it's ugly head.
      Same with HFC
      More ADSL2+ customers on steadily degrading copper, more interference, poorer performance.
      Abel Adamski
      • Sorry

        Data Volumes have been INCREASING
        Abel Adamski
  • Copper is the culprit

    No surprises here, Phil. Copper is the culprit, fibre is the answer.

    Turnbull's deliberate, taxpayer-funded delaying tactics of the NBN enabling legislation forced the Telstra shareholder proposal back six months, then the ACCC, instead of being ready to endorse it immediately in October 2011 after the 99.3% shareholder vote, took until 28 February 2012 to do so. NBNCo could then launch its three year construction plans, and the industry began upgrading ADSL to areas not imminent for fibre deployment, where they can recover their investment.

    Our marginal ADSL has gotten worse in recent years, and the third Telstra technician to swap us onto a different degraded copper wire pair, tracing it back several pits along the street, has declared the street at end-of-life, with no more options. Turnbull would have us lay new copper, in fact, four copper pairs to each premises if his "upgrade" promise is to be believed.

    Right now the ACCC must quickly amend its regulation to adopt the Telstra-NBNCo request to allow activating customers as the infrastructure to their premises is ready, not arbitrarily delay it until the entire FSAM is declared complete.

    ACCC head Rod Sims said on ABC TV's The Checkout last night that if a product or service does not deliver what you consider reasonable, then you have the right to complain. Well, the ACCC, ever since its pointless 121 POI decision in November 2010, is plainly acting against competition and consumers, so I hereby complain.

    Finally, the proliferation of new independent candidates and conservative parties (Clive Palmer's UAP and Bob Katter's crowd, for instance) should be worrying to a coalition that could romp home if it adopted the all-fibre deployment for urban Australia which cost it the 2010 election. It only takes a small number of disaffected coalition voters to vote cross bench for a minority government to be the outcome.

    Fibre to premises in urban Australia is the goal, and it is already proven to be self-funding in the first mainland areas. The coalition should adopt it, and our bandwidth will be in the top eschelons of the OECD where we belong, not the bottom.
  • ISP fail

    NBN is a double edge sword. Australian ISPs have stopped investing in broadband, they are just maintaining there existing networks until NBN comes around. I was told this by my ISP, when I complained that I can only get Dial-up fixed line broadband in a central suburb of Adelaide, because my copper line has so many ppl crammed onto it, that it is too noisy to allow ADSL of any kind to work. Another reason why Liberal's FTTN is not a solution. The issue is in the last mile copper network, which Liberals plan to leave untouched.
  • Global average Internet connection speed grew by 25% year on year

    Copper cannot support this level of Internet growth. Australia will be left behind with copper. Download volumes are increasing 61% per year. Congestion is increasing and average connection speeds are going down.
    Fibre can scale up for decades into the future. Fibre can support fast upload speeds needed for video conferencing, business collaboration and digital economy enterprises.

    What is the potential new productivity worth? $1 trillion AUD by 2050.
  • Anyone stopped to consider the basis of the statistics?

    Akamai provides the video platform used by services such as iView. Their statistics are based on connection speeds to that platform. If people start connecting to that platform using devices with slower than fixed broadband speeds, it will reduce the average speed as it appears to Akamai. With iView now being available on Apple devices and most recently the iPhone, people connecting to the service via 3G will do so with a speed a lot lower than 4.2Mb/s or whatever the average used to be, so will pull the statistics down. Even the use of WiFi in the home will reduce the apparent speed. As people move away from the PC in the study hard wired to the broadband modem and use WiFi to connect tablet devices, the weak link in the chain will become the WiFi link itself. In spite of advertised speeds of up to 300Mb/s, like all radio based systems, this maximum can be easily reduced as transmission becomes more difficult with 1Mb/s being fairly common.

    It really is childish to jump to the old 'it's the copper's fault' claims. If it was caused by the degradation by the copper network, then if we wound the clock back a few years, we'd have had stellar broadband speeds. We didn't. Overall, the copper network is no better or worse than it was last year when the speeds looked a lot higher. Clearly, it is something else.
    • Yes, but..

      ..the interesting thing about the statistics, is that the numbers for Australia have gone down while other countries' numbers have gone up. BRC, thank you for a thoughtful comment (all too rare in these fora), but I don't buy your argument entirely, because mobile devices and WiFi are just as popular in other countries where the average speed as reported by Akamai is increasing. However you are the only contributor to this discussion who stopped to consider eactly what is being measured. Clearly there are multiple factors involved. In most countries, where the average speed (even as measured by Akamai) is increasing, improvements in their bb infrastructure outweigh other factors. Contrary to what some commenters claim, I'd contend that there has been relatively little investment in Australian fixed broadband infrastructure outside of NBN for the last four years. I'm not knocking NBN as a vision, but stagnation for an average of at least six years, and up to 12 years (2009-2021) is a high price to pay, and one which I think was not factored into Labor's or NBNCo's planning. As originally announced the rollout was meant to be completed by around 2017; we'll be lucky if the (fttp) NBN is available to half of Australians by then even if it is allowed to proceed.
      • Fair point

        Thanks, Achilles. You do make a valid point. I haven't done so in detail, but it would be interesting to look at the services the Akamai platform is used for country to country. I understand it is used for Netflix in the US. This service lends itself more to streaming movies on large screen TV's via devices like XBox. In Australia, it is used for iView and would tend to be used to catch up on Four Corners and other TV shows. As such, it could be argued that the Australian application lends itself more to mobile devices whereas in other countries, to would be used to watch hours of late release movies in HD to a large screen TV. It would be interesting to see what would happen if Netflix was released in Australia. Then we would get a true like for like comparison.

        There has been an understandable decline in DSL equipment investment since the NBN announcement due to the reduced period to cover the return on investment. The slower than anticipated pace of the NBN rollout has caused this to be a greater problem than it should be, but that is a different story. With no large-scale improvement in capacity and a corresponding increase in download volumes, speed must suffer. That said, Telstra's 'top hat' rollout is probably an exception in that it is increasing investment in ADSL and should push the figures up at least for the DSL component of the statistics. Maybe it does not increase it enough to compensate for the general decline.
  • The NBN CAN and HAS given us lower speeds

    The NBN CAN make average speeds drop. Because its coming, investment in extra ADSL ports has stopped, and the number of people on ADSL has stalled. Meanwhile huge numbers of extra users have signed up to the slowest of all the main technologies, wireless. With the same number of users on ADSL, and lots more on slower wireless, the average speed has dropped. 4G wireless hasn't got many broadband users yet. And fibre still has virtually none despite Labor's NBN having been around for, what, 4 or 5 years, and nearly $10Billion having being spent on it.
    Gordon D
  • Your reasoning is sound

    I had not considered this and to me, it sounds like one of the best explanations I have heard.
    • Sorry, should have been attached to Gordon D's comment

  • I think 25 Mbps advertised = 4 Mbps actual

    Many people in Australia have ADSL2+, which provides connection speeds of 25 Mbps. However, in fact, the normal person who has ADSL2+ will only get measured maximum downloads of around 10 Mbps. Also, when considering international links, that gets decreased to 2 Mbps. Seems like the Opposition's FTTN will not shift Australia's ranking, because they keep hyping the connection speeds rather than the actual speeds that get experienced. Care to comment?
    • Advertised vs Actual

      Tom, networks can experience congestion at many different points, so the performance that you experience can vary widely, as you observe. ADSL2+ can achieve connection speeds of up to twenty something Mbit/s, but only if you are on a good cable pair and within a few hundred metres of the exchange. If you're further away the achievable speed drops off quite sharply, so most people experience 6-10 Mbit/s download at best. FTTP eliminates the problem of congestion in the access network, but there is still the potential for congestion in the RSP's backhaul, in the amount of bandwidth he leases from NBNCo (RSPs have to pay NBNCo for Virtual Circuits at each point of presence; they are charged by bandwidth), in his routers and routed network, and as you observe, in the international links. Whether people on an NBN connection experience a true 100 Mbit/s will still depend on all of the above factors. Initially the NBN experience should be good (because it removes the most obvious ADSLnn bottleneck) but RSPs will need to keep an eye on the dimensioning of their entire networks.
  • speed of connection vs speed of access to content

    do we want to start confusing people more? all these reports dump data on people who might not understand it, plus we are mixing speed of connection vs speed of access to content in with it all too... something most people don't understand. plus where are we actually getting are content from in australia? it's to be expected of course to be slower from overseas? where's those statistics? it's obvious we are behind globally not only on physical connection speeds, but actually how quickly we get content that's not local, Akamai, Google and other such companies are thankfully working on distributing content closer to us, but that's a long way to go.