New stats from Cisco on mobile data traffic

Cisco's VNI data traffic forecasts are quoted almost ubiquitously by vendors and operators in the industry.

They've just released some new mobile-specific figures and breakdowns - http://bit.ly/aFcWYa . I haven't had a chance to go through everything yet, but some initial highlights that jumped out at me:

• Top 10% of mobile users=60% of data and falling quickly, Top 1%=20%. Now similar to fixed. Panic over - and any vendors still citing scaremongering "data hog" slides at MWC will look rather silly. My view on the reason for this: a year ago, the semi-mythical 10% were almost all PC users with dongles. Now the smartphones have caught up a bit & evened things out.
• Offload is already diverting 31% of smartphone mobile data traffic to WiFi or femtos. Many of the other stats I've seen don't count or cover offload in their discussions of capacity use. Edit: on the other hand, the offload stats are still a bit vague. They don't distinguish between real offload substituting directly for macrocell data, vs. the incremental amount of traffic which would never have been used on the cellular network because of speed/price/coverage (ie elasticity effects).
• By 2015, M2M data will exceed tablet-based traffic, 295 vs 248 PetaBytes/month . That is hugely significant, because a lot of M2M stuff needs different business models, core architectures etc. Edit: my analysis is that because many of the use cases (eg CCTV) will be *upstream*, M2M will have an even more disproportionate impact on the cellular network.
• At end-2010, there were only 14m mobile-only Internet users, but this will grow to 788m in 2015. I've been saying for ages that the true importance of mobile-only was being overstated. I might have to have a closer look at their definitons & methods for forecasts, but nice to have 3rd-party corroboration of the current reality. Edit: looking at the stats, this seems to exclude non-Internet mobile data. It also appears to assume most users have at least *some* fixed web access, eg through schools or cafes.
• Laptops will still account for more than 50% of mobile data traffic in 2015
• Wireless home gateways generate mobile traffic than tablets - and still will in 2015. Edit: that's a fascinating insight, reflecting the importance & challenges of fixed wireless broadband beyond the "sexier" mobility-enabled devices like iPads. I'd be unsurprised if such users are far bigger generators of mobile video traffic, when connected to TVs and STBs.

Unfortunately, no mention in the study of signalling load - it's all about MB/GB of "tonnage".It's also not 100% clear if the figures are inclusive / exclusive of "optimisation" - ie whether it is data volumes actually flowing to & from the user device, or which comes in via the Gi Interface

More to follow when I get a chance