Saturday, January 7, 2012

2X data usage of Iphone 4S users vs Iphone 4 users - it's not Siri, it's likely not the 4S either

There's been a lot of coverage about the increased data usage on the IPhone 4S recently. All sources track back to the original report by Arieso, a network planning consultancy (completely unbiased I'm sure) 

I'm a big believer of the impending mobile data usage explosion scenario of the world, but I've got two issues  with the many reports out there now. 

1. It's not Siri!

The first is that a lot of the data usage increase has been pinpointed to the Siri feature. Although the original report doesn't do so, a number of quite reputable new agencies have gone on to place the blame entirely on the Siri feature of the 4S. E.g. Bloomberg has gone on to report it as "Apple Siri Feature Doubles IPhone data usage" . 

To get a sense of Siri data usage, I ran a number of queries and checked the data usage monitor to see how much each consumed. Varies quite a bit on download consumption - with much large downloads occurring, as you'd imagine, with web searches compared with just returned text transcriptions. Simple average puts it at ~20Kb download per query, see chart below. Artechnica has it at ~63Kb total data usage, but I assume that includes upload or had a much higher proportion of web-search queries. Combine this with a high-end estimate of 15 queries per day you get to ~9Mb per month, or 28Mb if you use 63Kb per query. 



That is tiny compared to a couple of benchmarks I dug up:
Using MacFormat's June 2011 online survey, average Iphone cellular data usage is 720Mb per month, so
a 9Mb increase is a tiny 1.3% blip.
Using O2 UK's June 2010 quote, overall average (not just Iphone) data
usage is ~200Mb per month. Even against this underestimate, a 9Mb increase only accounts for a 18% jump.

Net-net it would seem that Siri's answer is a clear "No" to "Are you a data hog that'll double my data usage?"

2. The Iphone4S enables, but does not cause data usage increase

The second bit that bugs me about the reports is that a number of news sites have simplified the research finding of "Iphone 4S users consume double the usage of Iphone4 users" to "Iphone 4S doubles data consumption" - which is quite a different from the first. The latter implies that the phone itself causes or requires 2x data usage, and this isn't true. Culprits include WSJ: "New iPhone doubles data consumption -study" and good old Asiaone : "Iphone 4S doubles data consumption".

The nuance is important because the real explanation of 2X data usage of the Iphone 4S vs the iPhone4 is likely to be caused by factor external to the handset itself, including:
a) The Iphone 4S, bought through an operator, typically comes with the largest data plans both in data allowances and speed.
b) As a new handset, the early-adopters are also the heavy data users, hence the process is self-selective towards data-hungry users
c) Some research methodology impact (hypotheses only, I don't have the raw research) - e.g. comparing recent 4S usage vs historical Iphone4 usage, not adjusting for the Xmas spike in usage.


So, some conclusions:
i) You shouldn't be holding back a Iphone4 to Iphone4S upgrade because it'll cause a spike in your mobile bill, you should be holding back because there's few incremental features it provides =)
ii) operators shouldn't specifically worry about the transition from previous Iphones to IPhone4S, they should be worried about the broader smartphone capabilities and data-hungry features they enable (wifi hotspots, cloud synchronisation, alternatives to voice and SMS)