The slow and steady increase in broadband speeds means we are using the Internet more often for more things. Statistical proof of this trend came via the latest edition of Cisco Systems Visual Networking Index (VNI) Study (s CSCO).
According to the study, the average broadband connection is now generating 14.9 GB of Internet traffic per month, up 31 percent from last year when it was 11.4 GB per month. And while a majority of this traffic is coming from online video –- streaming not P2P -– the trends show that we are using the Internet for more than just that. Give us more speed and we will use it all. And then we’ll want more of it.
Communication services such as Skype only increase the daily usage of the Internet. Add to the mix addictive sites like Facebook, Zynga and Groupon, and you can see that the Internet is becoming deeply embedded in our lives.
There is an interesting dynamic of the web –- the peak traffic -– that is equivalent of prime time on television. Peak-hour Internet traffic is 72 percent higher than Internet traffic during an average hour. In an average day, Internet “prime time” ranges from approximately 9 p.m. to 1 a.m. (for the local time zone) around the world.
Here are some of the key findings from the study.
- Peer-to-peer (P2P) file sharing is now 25 percent of global broadband traffic, down from 38 percent last year.
- Video — which includes streaming video, Flash, and Internet TV — represents 26 percent, compared to 25 percent for P2P.
- Over one-third of the top 50 sites by volume are video sites.
- Contrary to popular belief, none of the top 50 global web sites (by traffic volume) featured explicit adult content.
- Ten of the top 50 sites were associated with software updates and downloads (security and application enhancements).
- The top 1 percent of broadband connections is responsible for more than 20 percent of total Internet traffic.
- The top 10 percent of connections is responsible for over 60 percent of broadband Internet traffic, worldwide.
Earlier this morning I wrote about the rapid growth of VoIP –- 112 million broadband connections with VoIP attached to them. Today, voice and video communications traffic (such as voice over IP [VoIP] and voice and video over instant messaging) has reached 2 percent of all traffic, up from less than 1 percent last year. Here are some interesting voice & video communication-related data points:
- Skype accounts for 0.57 percent of total traffic.
- Other VoIP accounts for 0.64 percent of the traffic.
- In comparison, email accounts for 0.23 percent of the traffic.
Cisco is predicting that video calling will exceed 1 percent of consumer internet traffic by end of 2010. I bet Apple’s FaceTime is only going to help achieve that goal. In summary, I think all these numbers can be tied to my initial assertion: Broadband is the magical driver of all things on the Internet. Thanks to broadband, everything, including the web, changes.
Related research from GigaOM Pro (subscription required):
The statement that the “average broadband connection is now generating 14.9 GB of Internet traffic per month” is misleading given the additional stats that 10% of users generate 60% of traffic. A small number of users are downloading huge volumes of data, the rest are not. So broadband may be changing everything, but for the vast majority of users, they’re still doing basic things a lot, like browsing.
With the Internet peak times between 9 PM to 1 AM, would it be a reasonable analysis that more traffic is generated for leisure rather than enterprise/business use?
I have 2 questions:
1) Would Adobe Flash videos also not count as streaming video? Why are we showing both of them seperately. Is Streaming here confined to other formats like Windows Media and Quicktime?
2) How is Cisco able to differentiate between video downloads and any other download? I presume they are picking this data from logs of their core routers, but I am not sure how you can differentiate data at that level (except if you are doing Deep Packet Inspection- which is unrealistic to do on such a large scale I believe)
Nikhlesh
In the television world, most of the attention was around primetime mostly because that was the leisure time. We are seeing the repeat of that, though we do see some other patterns: people work longer hours and as a result are only able to do other activities such as calling friends/family late night.
on the Flash question: i don’t quite know but will look into this.
Cisco gets data from ISP partners and then creates a report based on it. I don’t that VNI is based on DPI. Carriers have had the ability to figure out what kind of traffic is flying on their networks for a long time.
I’m not at all surprised that 10% of users use 60% of bandwidth. You’ve got professionals and hardcore prosumers on one end of the curve, and casual users (often with less-capable equipment) on the other. The best solutions and services for one are completely useless to the other. One of the main problems that we’re just starting to get out from under is Big Content companies and the broadcast model. Besides being as obsolete as chain mail in combat, it encourages inefficient use of the network. With less centralised content creation and distribution, you not only have better use of the available bandwidth but a more involved society as well. Win for everybody…except the corps flogging obsolete business models.
Niklesh, I’d buy that assumption (that more data is used for leisure than work). If you look at what the high-traffic sites are, particularly for North American users, that makes perfect sense. Too many people’s workplaces are either Internet-hostile (blocking/filtering) or dependent on light-bandwidth tasks (minimal rich-media Web browsing, lots of email and text IM).
Jeff
Thanks for sharing your insights. I agree about the high-end of the users sucking up a lot of bandwidth, but one has to remember a small little fact the PCs taught us: Early adopters are the mainstream of tomorrow.
Just as the amount of data consumed by the top 10 percent goes up, so does the total data consumed overall. That is part of the continuing mainstreaming of the broadband.
Om, I didn’t mean to sound as though I disapproved of the high-end phenomenon; I’ve been one of those “high-end” users myself for well over a decade. Early adopters do define the mainstream. Always have, always will.
What I think is needed is for the US to finally get its act together and get serious about universal access to a regularly-upgraded high-speed (50+ Mbps) network. Have a regulated monopoly or oligopoly if that’s all we can sell to Congress, but virtually every country on the planet that’s serious about long-term economic growth, from Poland and Romania to Kenya to Thailand to Singapore is at least in the planning-and-budgeting phase of such a network. When people ask me why I always say “an upgradeable network,” I remind them that when I first started using data communications, 300 bps was “high-speed.”
It doesn’t matter how many billionaire Wall Street bankers we’ve got if the rest of the country is dead broke because we don’t do anything the rest of teh world finds useful.
The “early adopters” we need to be planning for aren’t the guys like us now; it’s the guys 30-50 years from now. Because absent a Manhattan-style Project with funding and resources to match, it’s going to take the US that long to dig out of the hole we’re in. And by the time we get there, we’ll need to plan for the early adopters 30-50 years after that.
IT’s a race. And it’s going to have at least as much impact on the future of our country as the space race or any of the arms races we’ve had. As a country, we need to take it that seriously. Too bad the educational and political system seems custom-designed now to make that as difficult as possible.
This all may well and good, but until we have Wi-Fi that doesn’t drop and video download speeds that get better vs staying the same for 10 years straight I don’t see things moving to the net that really should move to the net. The problem is as the pipe increases (speeds, data flow volumes) the data needing to flow keeps increasing at the same rate. I still have to pause for You Tube videos to load while connected with high speed cable broadband and a brand new laptop. =P
Howie
I think it is a “work in progress” reality when it comes to WiFi or for that matter any Internet-related technology. Look at where we are today and where we were at the turn of the century — it is night and day. But as you say, more needs to be done.
PS: I hate the WiFi-related issues as well and despite sitting next to the router, I am currently getting about 12 Mbps to the laptop. I plug-in Ethernet cable, well that is running at about 50 Mbps. So I feel your pain 🙂
Do you distinguish broadband from wireless internet connectivity? I ask because I have a personal problem paying twice for internet connectivity, I won’t get a smart phone until I can legitimately tether (i.e. it’s a supported service) and supplant my cable connection to the internet. Although I won’t do this if the 3/4G/LTE can’t support Netflix or Hulu streaming.
Do you guys envision a world where pretty much all connectivity migrates to the wireless spectrum? And if so, how long do you think this will take? Because like I said, I’m not going to pay $50/mo for cable internet and another $30/mo for mobile internet, but I’ll pay $60 or $70 for an equivalent mobile plan. Or maybe $100/month to have data on all my mobile lines (2) and let each mobile act as the gateway for that person’s laptop/desktop/etc. I know I sound cheap, sorry.
Broadband has definitely changed our lives. Today, we turn to the web without any hesitation for information on movie theater times, picking a restaurant or finding directions to a business or friend’s place. And increasingly in the last six months for on demand entertainment -thanks Netflix.
Thanks to the prosumerization we expect work environments to be equally friendly. This is distinctly different from the PC era when offices and businesses adopted PCs first. Professionals saw that they could be useful at home, and brought them home when the price permitted it. Early adopters of home PCs were probably hobbyists.
Broadband therefore started a revolution for the consumer in terms of applications and services, which I suspect businesses will adopt.
RN
The only thing faster broadband is driving now is commercial video delivery via the internet.
Then we are seriously phuct, because you can bet your bottom euro that the rest of the world is going to have better ideas than that, and somewhere, somebody’s going to get in the habit of executing. That “somebody” is going to have much higher-than-average, perhaps even dominant, economic advantage for their country. Why can’t that country be the US?
I would argue that a broader broadband in the US and developed markets is less of a change than the change we can expect in developing markets where broadband is new. The change that can be occur in a place like India can be dramatic as broadband takes hold.
It’s funny the assertion “Give us more speed and we will use it all. And then we’ll want more of it,” reminds me of the fact that building new highways increases rather than relieves traffic. I wonder if there’s a downside to trusting too much to high speed access. Yet it also seems impossible that not everyone has it yet.
This is very interesting. Do you have a similar study for wireless broadband traffic? Thanks