admobI’m a big fan of AdMob – they offer a great mobile advertising service and for quite a while have been publishing a monthly “Mobile Metrics Report” that provides a snapshot into the types of traffic seen on their ad network. Its nice to see a company willing to share information with the community.

I’ve also read quite a few blogs that have picked up bits and pieces of information from these monthly reports… and totally misrepresented the data to their readership. To that end – this is a lesson in statistics – and how to read AdMob’s Mobile Metrics Report. This months report can be found here – and focuses on the various ways to look at smartphone market share globally. Before you read it however – please try to remember a few things:

For those who don’t know AdMob – they are a mobile advertising network, which means they represent over 9,000 publishers on the mobile web and serve up the ads on those sites. They also represent a vast majority of the in-app iPhone and Android advertising banners you see (Over 3,000 apps represented). Every time someone lands on one of their sites or loads an ad in an app, they identify the handset and operating system and log it into their databases. This report is an analysis of that data.

Smartphone Handset Marketshare from the August AdMob Report

Smartphone Handset Marketshare from the August AdMob Report

There’s an inherent problem with his however, because App data and WAP data are being treated the same. While WAP sites are browse-able by any web enabled handset,  the vast majority of apps are on the iPhone OS platform, followed by Android. This means that when AdMob reports that the iPhone OS has a over 50% market share in the US, they are not reporting how many iPhones are in the marketplace, but how many views they are getting on their ads from iPhones. (AdMob does try to explain this in their disclaimers). Because AdMob is representing a huge chunk of impressions that can ONLY be seen by iPhones, this will always skew the data in the iPhone OS’ favor.

Of course – iPhones dominate the applications market, so this makes quite a bit of sense. Does that mean this is useless data? Not at all. The report still gives us a great snapshot into how people are using their phones.

“…the data includes our in-app traffic for both iPhone and Android platforms,  so yes that is certainly something you should consider when reviewing our numbers.  However, the huge growth in mobile usage of apps is, in many cases, replacing mobile browsing activity on devices like the iPhone and Android.. ” – Mike Fyall, AdMod Product Manager

Mobile applications are generally easier to use and have more functionality then the limited WAP environment, and even phones that can access traditional websites still have difficulty with sizing, and handling the vast array of browser plugins like Flash and Shockwave that many websites rely on. Mobile applications are optimized for the handset they are being used on, do not typically require zooming in and out to be able to read content, and all app bugs aside, always work.

up up and ad-way!

up up and ad-way!

Another interesting insight from AdMobs report is the steady increase in overall impressions in every region around the world. This bodes well for application developers looking to make revenue through free applications that are ad supported and relying on constant traffic to generate impressions. It also makes it so much more important for free apps to be utilizing effective marketing strategies to help them garner their share of that traffic in an increasingly competitive market environment.

  1. “in many cases, replacing mobile browsing activity on devices like the iPhone and Android..” I’m surprised to hear someone from AdMob saying that. Doesn’t the number of mobile publishers on AdMob’s network outnumber downloadable apps more than three to one?
    On a related note mobiThinking has just published a guide to mobile ad networks that might interest you. It includes profiles from nine leading mobile ad networks, including AdMob. The URL, if I’m allowed is:

    Comment by mobiThinking — October 4, 2009 @ 8:42 am

  2. Well, I think you have to look at traffic… not number of publishers. What percentage of the traffic on their network is coming from apps vs sites? I can only imagine that data would show apps are far more “sticky” and because of that get more traffic then individual sites do. Many people who bookmark sites online still dont do it on WAP sites, however having an app in your menu is like having a pretty icon based graphical bookmark reminding you to use it.

    Comment by watkinsmobile — October 4, 2009 @ 8:58 am

  3. I just finished reviewing the latest metrics report by AdMob and am looking for clarification as to the definition of “requests” (p2 Worldwide requests increased by 4%…).

    Many thanks…

    Comment by Daniel — October 25, 2009 @ 1:09 pm

  4. Hi Daniel,

    A “Request” would be a page requesting an ad from the server. Because they do not always have enough ads to fill the space, they call it “requests” not “ads served”

    Hope that helps!


    Comment by aaronwatkins — October 25, 2009 @ 2:29 pm

  5. […] How to Read AdMob’s Mobile Metrics Reports: “makes it so much more important for free apps to be utilizing effective marketing strategies to help them garner their share of that traffic in an increasingly competitive market environment.” […]

    Pingback by how-to-read-admobs-mobile-metrics-reports-the-appency-press | blog — November 12, 2009 @ 7:58 pm

RSS feed for comments on this post. TrackBack URL

Leave a comment