Is the mobile activity you are counting really mobile?
By Mike Moran. Filed in Internet Marketing, Monthly Newsletter |Tags: iPad, IPhone, MiFi, Mobile device, Mobile phone, Wi-Fi
Image via Wikipedia
We’re all spending a lot more time thinking about mobile marketing these days, but the statistics that we use to track it aren’t keeping up. It was so easy not so long ago, when mobile phones were the only way to be online while on the go. But so much has changed in the last couple of years that we are now counting many activities as being mobile when they are not, and we are not counting many activities as mobile when they really are. Confused? I’ll be happy to explain.
Our feeble metrics systems need some way of categorizing mobile usage, and what made a lot of sense (the key word here is “made”) is to look at the device that is being used. So, if someone hits a Web site using an iPhone, that is counted as a mobile use of the site. When a computer is used, then it’s not mobile (stationary?).
But that was then and this is now.
Much of our computer usage is mobile, but it doesn’t count as mobile. How many of you have netbooks or laptops with embedded cell network connections? Or how many of you walk around with a portable cell phone Wi-Fi router, like Verizon’s Mi-Fi? Or tether computer access to your cell phone? How many of you use Wi-Fi on planes (or even trains, buses, or cars)? All of that usage is literally mobile, but our metrics systems don’t count it that way, because you were using a computer, which is assumed to be not mobile (fixed?).
And we have the opposite problem, too. If you are using your cell phone to access the Internet at home, is that really mobile? Just because you didn’t want to walk into the next room to look at your computer?
But my favorite dilemma concerns the iPad. Metrics systems typically count all iPad usage as mobile, even though we know that most people don’t even have access to the cell phone networks using their iPad. They are sitting on their living room couches hooked up to their home Wi-Fi, but that counts as mobile.
This struck me when I was told recently that 20% of all visits to a client’s Web site were from mobile devices. This struck me as impossible, because the client offers help for a complex and intimate problem that I didn’t think people would be checking out on their cell phones in public. Now, undoubtedly I was underestimating the number of people doing just that, but I still don’t believe it is 20%. It’s more likely that iPad usage is driving the number up and that people are using their cell phones in their office to avoid Big Brother checking their Internet usage and using them at home so that no one else can see what they are doing. But none of that is what I think about when people say “mobile usage” although now I see that it should be. I don’t know what the answer is here, but we should be thinking about what our assumptions are for what these mobile users want–maybe it makes more sense to divide devices into big screen and small screen than mobile and non-mobile (immobile?).
So, the next time your Web metrics analyst tells you that 20% of your traffic comes from mobile devices, ask him how he knows. And then decide if that number means what it used to mean. I think it doesn’t.
Tuesday, November 2nd 2010 at 12:53 pm |
Very interesting post here, mobile is still so new and deciding what metrics matter is still new as well. I must say that when I think “mobile usage” I too don’t factor in all these items, and maybe we should be.
Tuesday, November 2nd 2010 at 10:29 pm |
The validity of the stats depends on how you use them, I suppose. A user connecting to my site from an iPhone, in his home, through his WiFi router, is still a connection from a mobile device. The keyword being “device”. If my interest is primarily about the device’s capabilities, then I would say the stats are probably still valid. If my interest is to find out how many users are accessing my site outside of home or work, then the stats may be misleading.
Tuesday, November 2nd 2010 at 10:36 pm |
I understand what you are saying, Steve, but I am not sure why that would be important. An iPad can display a Web page just as a computer screen can, so I don’t know what capabilities an iPad has that would be pertinent to a marketer other than whether someone is out and about or sitting on the couch, which we can’t know. My point is that people are using “mobile” for decisions that the data does not support. You are right that if they know what the data means (it’s a mobile DEVICE) that they are OK, but that is why I wrote the post.
Wednesday, November 3rd 2010 at 2:23 pm |
Ah, you said a lot to say nothing. But I think I got the point. There’s only so much you can really rely on when it comes down to our metrics.
I wrote a piece on something similar a while ago, “Facts are Meaningless”.
“Facts are meaningless. You could use facts to prove anything that’s even remotely true!” – Homer Simpson
Saturday, November 6th 2010 at 2:25 pm |
This is an interesting topic as I’ve been conducting quite a bit of research into the mobile marketing arena, and it never occured to me that many of the measurement metrics assume that a smartphone or cell phone are involved. I know I use my laptop with a Verizon card when I’m on the road. I now understand that the analytics I am using can include many of the newer non-mobile devices.
Friday, November 12th 2010 at 4:35 pm |
Good topic… For me it is all related to the Capabilities of the device the user is using. Which may be the combination of physical device capabilities, OS and browser.
If I have built my entire site using flash then it is useless to someone who tries to access it from a device which does not support flash.
We have all run into issues with sites built for IE, which do not work correctly with Firefox or other browsers.
So for me it is more about users capabilities to view and interact with my sites content.
Saturday, November 13th 2010 at 12:33 am |
So, it looks like we need to move away from the notion that some devices are inherently mobile and some inherently are not, and move to the notion that Web apps need to support client devices based on a device’s capabilities (including the bandwidth/latency of its network connection), instead. For instance, if the client device has location services and the app can do something useful with location services, then do that regardless of other attributes of the device or its connection.