Over the last decade, the Cisco Mobile Visual Networking Index (VNI), our rolling five year forecast of Internet trends has delivered some amazing and eye-popping predictions. Today’s VNI Mobile Forecast update, for example, shows that the deployment and adoption of 4G is accelerating even faster than we predicted just a year ago.
In the United States, just like the rest of the world, mobile data consumption continues to climb, driven by insatiable demand for video, video and more video, coupled with a surge in mobile users and devices, and an expansion of mobile networks to serve the Internet of Things – connected cars, homes, health care and more.
For policymakers, the flood of data traffic has created challenges with radio spectrum, forcing nations to find more efficient ways to allocate this scarce resource. And for the industry, it has created challenges forecasting what’s next. So far, we’ve done very well: in each of the last five years, initial projections have been within 10% of the reported 2015 data estimates.
Cisco is committed to being as accurate as possible on the VNI, as well as being transparent. That’s why we go back every year and compare our projections with subsequent mobile data estimates. Based on this year’s comparison, we reset the 2014 U.S. baseline for total mobile data consumption and traffic growth after incorporating data traffic estimates from CTIA, the Wireless Association. CTIA is the most comprehensive source for historic data in the U.S., since they collect data from wireless telecom operators through a voluntary annual survey.
The CTIA data shows fluctuating, positive, growth in U.S. mobile traffic from 2012 to 2014 with larger peaks and lesser peaks, including traffic growth of 26% in 2014. Based on CTIA’s survey data, as well as a close review of traffic samples and direct data sources that are part of our methodology, we reset the 2014 monthly baseline down — from 531 petabytes to 322 petabytes. This revised number may in fact understate the true traffic growth since some mobile operators suggested on their earnings calls that their traffic grew in excess of 50% between 2013 and 2014. But the CTIA survey number is one of the most comprehensive estimates of historical growth, and our 2014 baseline volume adjustment now aligns with CTIA’s figures.
Over the course of 10 years publishing the mobile VNI, we’ve reported out key trends about Internet traffic, connected devices, applications, and shifts in demand. These data are used by network operators and other industry players, including Cisco customers, partners and competitors, as well as government and industry decision-makers. That’s why it’s so important that Cisco be transparent about our methodology and validation efforts.
The adjustment reflected in today’s report does not change the forecast growth curve much at all. The 2015 growth (year-on-year) is forecast at 56%, and the 2015-2020 compound annual growth rate is forecast at 42%. The fundamental trend lines hold. There’s more devices and more video consumed over faster networks, which will continue to transform the world around us.
Robert
Thanks for the detail. Of course there are occasional mistakes, but the VNI remains the best study of its kind in the world.
I’m particularly glad VNI provides enough data I can look at different questions. The prediction that U.S. growth will fall to ~37% confirms my conclusion that the spectrum “crisis” is wildly exaggerated in most places. It’s easy to see that’s true; wireless speeds could not be going up as fast as you measure and project if spectrum shortages were really slowing things down.
In fact, the problem most telcos have today is finding enough customers to use the capacity of their existing networks. The problem is likely to get worse because the technology is enabling capacity building at a ferocious rate. Carrier aggregation is allowing use of currently fallow spectrum, such as the 2300 band AT&T owns. AT&T currently has enough unused spectrum to triple capacity. MIMO/beamforming can at least double that again. There’s another 20% possible by sharing spectrum, which Sprint and T-Mobile are doing indirectly for Google’s phone service.
Add the capacity of the 13M cable, AT&T and Verizon WiFi hotspots, which can double effective capacity where it’s most needed, dense areas. Dish owns enough unused spectrum to build 3 AT&T-sized networks. There’s massive capacity coming available in most of the U.S. at 3.5 GHz, an important move by the FCC.
Without a major CapEx increase, it’s easy to see a path to increasing capacity 10-25 times even before 5G and M-MIMO.
The engineers are ready to deliver growth well above the VNI traffic estimates. Delivering what’s possible is dependent strong, aggressive policy direction, including an FCC that requires efficient use.
The data in the VNI also puts the lie to the contention that more spectrum/wireless capacity will have a large economic impact. As you note, most of the increased demand is for video and music. That’s a good thing and can enrich our lives, but does anyone really believe that watching more football and YouTube when you’re not at home yields a major income boost?
Thanks, Arielle and team, for some excellent work.
Dave,
Thanks for your kind words about the VNI forecast and team and your thoughtful comments. I agree with your point that innovation is rising to the occasion and that technology is getting better at wringing more efficiency out of spectrum. The problem is that, especially with the explosive growth in video consumption, including HD and increasingly 4K video, these improvements by themselves cannot keep pace with the demand. It’s an arms race between increasing demand and new techniques to be more efficient. In fact, we are seeing the mobile operators investing in these new technologies, something they would not be doing if they thought demand was flat or growing only slowly and they could just use what they have. Over the 2015-2020 forecast period, we see US mobile data traffic growing at a very healthy 42% CAGR and busy-hour traffic growing even faster at 45% CAGR which makes the additional point that you can’t just build or manage for “average” traffic. Finally, we are seeing that operators are increasingly curbing demand with data caps that are not keeping pace with consumer behavior, which is why WiFi offload is already 62% of traffic from mobile (SIM enabled) devices and forecast to grow to 69% in 2020.
As in the past, I expect we will be discussing the interpretations of our findings in multiple venues.
Best,
Pepper