Just about a year ago I wrote a post I called "Waiting for mobile" that somehow never made it online. The genesis of that post was a graphic I'd seen from allaboutsymbinan.com that was built from comScore data. The data seem to say that while smartphone use was rising rapidly, most of their owners were not using them to access the Web.
Now comScore has released new numbers for 2011. There are about 300 million mobile subscribers in the US and comScore says that around 75 million of those are now using smartphones. That equates to about one quarter of the total number of mobile phones. The same comScore report provides some use statistics for various online behaviors such as web browsing (39%), downloading apps (38%) and accessing blogs and social network sites (28%). So it seems that a majority of smartphone users still are not going online or, if they are, it's rare.
You can quarrel with the numbers if you like; I don't know enough about the methodology behind them to say how accurate they are. And you can do the math in various ways but it seems to me that it's tough to argue that more than about 15% of the US population is sufficiently active online with their smartphones to be potential survey respondents. The one exception to all of this is text messaging, but most of the mobile enthusiasts I talk to take a dim view of the potential of SMS for serious research.
So I think that mobile continues to be a niche methodology, an intriguing one but a niche nonetheless. I have seen some really interesting uses of mobile around major events and in ethnographic studies, but I think it has some serious weaknesses for gen pop studies and anything other than some very unique target groups. Recognizing those weaknesses, staying focused on how to execute well and then fitting mobile into the broader spectrum of methodologies is where we need to put our energies. It could well be the next big thing, but not yet.