I saw a mobile ad a while ago that I thought was pretty good.
The ad used the display url: TheirSite.com/Mobile.
I tweeted about this; and most people didn’t think it was a great display URL. Now, it could be that they are all on Andriod or iPhones. As these phones can render pages fairly well even if they are not designed for mobile devices, that might not be an intriguing display URL.
My poor BlackBerry sometimes has an awful time rendering sites that don’t have mobile components – so I thought the difference might lie in the type of phone someone has (the BlackBerry’s browser pales in comparison to just about any other smartphone – my iTouch and gaming consoles have better browsers).
This type of thinking usually leads me down the path of testing.
The total available searches on mobile devices is still very low compared to computers, just take a look at the differences in visitors by OS on this site:
The dropoff is pretty steep once you move past just a few operating systems.
However, you can’t just ignore more than 70,000 total mobile visits.
So the test was pretty simple three display URLs:
- TheSite.com/iPhone (only shown to iPhone users)
Note: the site being tested is different than the above site, but similar in traffic; however, I can’t show this particular site’s analytics per owner’s request.
Here’s the CTR by OS (I threw out some of the data because the number of clicks and impressions was too low to gain any conclusions; and the iPad is not a phone).
The iPhone users obviously like seeing iPhone in the display URL the most, and a more typical /ProductFolder ad second. That’s not too surprising.
The Andriod owners surprised me. They liked /Mobile as much as the /ProductFolder. I did not have enough data to segment out Android by operating system (so Android 1.6 and Android 2.2 are lumped together); and that could be the cause of the CTR difference.
The BlackBerry owners felt comforted by the /Mobile (and I don’t blame them).
The differences in CTR between the display URLs is significant (much more dramatic than some of my testing). Rarely do you see a 0.8% and a 2.2% CTR in the same ad group for search (this was not a well known brand, so TheSite.com often doesn’t work well).
This is just one test, and the data set wasn’t huge – so don’t think this is how you should design your ads.
This type of test should give you the framework for designing mobile display URL tests.
Mobile traffic is still very low (or non-existent) for some. However, mobile traffic increases every single month.
Understanding your visitor behavior now will help you anticipate how to reach them when they start searching for you on a mobile device.