I recently researched the impact of mobile-optimized expe­ri­ences on mobile vis­i­tor engage­ment. To com­pare mobile usage between the “mobile opti­mized” and “stan­dard” ver­sions of a web prop­erty, I cal­cu­lated key engage­ments per visit on both the stan­dard and mobile web­sites for 15+ pop­u­lar devices includ­ing the Apple iPhone, Motorola Droid, HTC Nexus One, Palm Pre, Black­Berry Storm 2, and Nokia N97. I also ana­lyzed impact on the iPad which I’ve detailed at the bot­tom of this post.

Tip: For your own analy­sis you should use a engage­ment or con­ver­sion met­ric that dri­ves rev­enue for your business—ad views or page views for media com­pa­nies, rev­enue for retail­ers, etc. You should also ensure you are accu­rately iden­ti­fy­ing mobile vis­its and vis­i­tors. CNAME/First Party Cookie imple­men­ta­tions are a must here—for more info, read Top 5 Mobile Imple­men­ta­tion Gotchas.

Pro­vide Mobile-Optimized Expe­ri­ences for Mobile Users

The results of my analy­sis shouldn’t sur­prise you—mobile-optimized expe­ri­ences pro­duced an aver­age 75% higher rate of engage­ments per visit for mobile users. Such sig­nif­i­cant improve­ments in engage­ment for mobile vis­i­tors help jus­tify invest­ment in mobile-optimized chan­nels but it’s only the first step.

Fix Trou­ble Spots in Mobile Channels

Now you’re feel­ing good about the gains from pro­vid­ing a mobile-optimized expe­ri­ence, it’s time to make addi­tional improve­ments. Com­par­ing total engage­ments per visit for both your stan­dard site and your mobile site (or app) will help you dis­cover break points in the mobile expe­ri­ence. For the data set I ana­lyzed, total engage­ments per visit were 20% lower for the mobile site than for the stan­dard site. The gap rep­re­sents a sig­nif­i­cant oppor­tu­nity to improve the user expe­ri­ence on the mobile site and cre­ate real gains in engage­ment and con­ver­sion. By drilling one level deeper to spe­cific steps in your key engage­ment or con­ver­sion process, you can dis­cover user drop off points.

funnel-mobile-site-comparison

In the mock-up data above, nearly the same per­cent­age of total site vis­its result in a “lead ini­ti­a­tion” for the stan­dard and mobile sites but fall­out between ini­ti­a­tion and com­ple­tion is much greater for the mobile site. In this exam­ple, fine tun­ing the lead ini­ti­a­tion page on the mobile opti­mized site so it con­verts as well as the same page on the stan­dard site could increase mobile leads com­pleted by 3X!

Tar­get Expe­ri­ences to Spe­cific Devices Where Appropriate

By cre­at­ing mobile opti­mized chan­nels for your cus­tomers and then fine tun­ing those expe­ri­ences, you can pro­duce very sub­stan­tial improve­ments in engage­ment in con­ver­sion; but what hap­pens when a new cat­e­gory of device is thrown into the mix? A lot of mar­keters are con­sid­er­ing how to han­dle new tablet devices like the Apple iPad or the HP Slate. Mar­keters are ask­ing ques­tions like, “should iPad vis­i­tors receive the stan­dard site, the mobile-optimized site, or a third fla­vor cus­tomized for tablets?”

The answer to this ques­tion and oth­ers will vary depend­ing on busi­ness model, indus­try, and many other fac­tors; how­ever, these types of busi­ness deci­sions should be dri­ven by ana­lyt­ics. For fun, I included vis­its from iPad users in the dataset I ana­lyzed to deter­mine the uplift pro­duced from mobile-optimized expe­ri­ences. The results of my analy­sis may sur­prise you—mobile-optimized expe­ri­ences pro­duced an aver­age 15% lower rate of engage­ments per visit for iPad users. For this par­tic­u­lar set of data, the cus­tomer would be bet­ter not redi­rect­ing iPad users to the mobile ver­sion of the site.

While the dataset pro­duced by vis­i­tors using an iPad was small and ulti­mately may not rep­re­sent the pop­u­la­tion after launch, engage­ment and con­ver­sion met­rics for the iPad and other tablet devices should be closely mon­i­tored to ensure users are receiv­ing an opti­mal expe­ri­ence. By tar­get­ing spe­cific expe­ri­ences to spe­cific devices like the iPad, you may be able to attain an addi­tional increase to key user actions.

Here’s some addi­tional resources you can lever­age to ana­lyze the impact of the iPad on your online initiatives.

Sam­ple iPad User Agent—use­ful for updat­ing mobile site redi­rec­tion rules:

Mozilla/5.0 (iPad; U; CPU iPhone OS 3_2 like Mac OS X; en-us) AppleWebKit/531.21.10 (KHTML, like Gecko) Version/4.0.4 Mobile/7B317 Safari/531.21.10

iPad Ana­lyt­ics Sup­port—Omni­ture pro­vides both mea­sure­ment and report­ing sup­port for the iPad: iPad App Mea­sure­ment & iPad Mobile Device Report­ing (press release)

iPad-device-identification

iPad Site­Cat­a­lyst App—updated ver­sion of the app opti­mized for the iPad:Site­Cat­a­lyst App for iPhone, iPod Touch, and iPad (Ver­sion 1.2)

Home­work assign­ment—browse to the mobile reports in your Site­Cat­a­lyst report­ing envi­ron­ment and search for “iPad” to see if any­one with a “pre-release” device has vis­ited your web­sites or apps.

  • Dave

    Ed– this is very inter­est­ing how­ever it would be help­ful to bet­ter under­stand your method­ol­ogy. First you state that mobile-optimized expe­ri­ences aver­age 75% higher rate of engage­ments per visit for mobile users. But the next para­graph states “For the data set I ana­lyzed, total engage­ments per visit were 20% lower for the mobile site”.

    • http://blogs.omniture.com/author/ehewett Ed Hewett

      Dave-to clar­ify, mobile users are more engaged when they receive mobile-optimized expe­ri­ences than when they receive a stan­dard desk­top expe­ri­ence. How­ever, mobile users still have lower engage­ment rates when com­pared to desk­top users. This demon­strates two things:
      (1) pro­vid­ing mobile opti­mized expe­ri­ences is good
      (2) there’s still lots of room to improve mobile opti­mized expe­ri­ences (of course ana­lyt­ics play an impor­tant role in iden­ti­fy­ing trou­ble spots, etc.)

  • Dave

    Ed–thanks for this as I’ve been using this 75% stat in recent pre­sen­ta­tions to clients. How­ever the last 2 times I’ve shared it I’ve been asked what it actu­ally means and how you were defin­ing engage­ment. Any chance you could con­tact me offline?

    • http://blogs.omniture.com/author/ehewett Ed Hewett

      Dave–i looked at mul­ti­ple prop­er­ties so the mea­sure of engage­ment was dif­fer­ent depend­ing on the site. For retail, the mea­sure was rev­enue per visit, for media, the mea­sure was page views per visit, etc. You may also be inter­ested in this follow-on post: iPad Users Twice as Engaged on Desk­top Web­sites.