Optimizely Roadshow 2014: Turning Data into Action (Part 2)

Dan Siro­ker, Co-Founder and CEO of Opti­mize­ly pre­sent­ed last week in Atlanta at the Opti­mize­ly Expe­ri­ence 2014 Road­show.  In his keynote, he post­ed a few ques­tions: “Would you want…

  • 764% more social fol­low­ers?
  • 23% more con­ver­sions?
  • 123% more con­tent engage­ment?
  • Or 150% more cus­tomers?”

Does a one-legged duck swim in cir­cles?   Of course we would!

Dan went on to show brand suc­cess sto­ries in using Opti­mize­ly for a vari­ety of tests to dri­ve those results.  He pre­sent­ed test­ing types in 4 buck­ets:

  1. Cap­i­tal­ize on your traf­fic
  2. Opti­mize every chan­nel
  3. Per­son­al­ize your expe­ri­ences
  4. Think like a sci­en­tist

Below is a recap of Dan’s exam­ples direct­ly from his pre­sen­ta­tion: here is a link

1. Cap­i­tal­ize on your traf­fic oppor­tu­ni­ty

Don’t look at a sin­gle chan­nel but look across chan­nels where cus­tomer activ­i­ty is grow­ing.

This was a fas­ci­nat­ing way to show the lack of bal­ance in indus­try focus… For every $92 spent on mar­ket­ing (get­ting some­one to your site), there is only $1 spent try­ing to con­vert them! (Source: IBM TeaLeaf, 2012).

Don’t throw more mon­ey at media if you aren’t look­ing at the expe­ri­ence and improv­ing the back­end!  A per­cent­age gain in con­ver­sion rate has a longer-term impact and will be more prof­itable at the end of the day.   Shown below, when PR or a pro­mo­tion goes into mar­ket it will have a short lived bump in traf­fic or con­ver­sions, but if you focus on con­ver­sion rate improve­ments you get longer term rewards.

Anoth­er exam­ple was Code.org’s cam­paign test­ing dif­fer­ent text and calls to action led to an extra 12MM addi­tion­al par­tic­i­pants to their “Hour of Code” project.

Code.org Optimization of Experience

1. Opti­mize every chan­nel

Under­stand­ing your chan­nel per­for­mance and behav­iors is impor­tant in test­ing.  Paid Search traf­fic will most like­ly act very dif­fer­ent­ly from email or social traf­fic.  There­fore you can uti­lize Opti­mize­ly to serve dif­fer­ent land­ing page ver­sions back on the refer­ral source.

Liftopia did this for Search engine traf­fic and sim­pli­fied the land­ing page to see a 20% increase in con­ver­sions.  Easy change, nice reward.

Know­ing your audi­ence is key!  The oppo­site effect can also take place when test­ing if you for­get to think about your audi­ence.  IGN want­ed to increase video views and test­ed mov­ing the video link in the nav­i­ga­tion to the far left (vs. far right).  Con­se­quent­ly the result was a 93% decrease in video con­ver­sions!  When dig­ging into the data by seg­ment the decrease came from return­ing vis­i­tors who sim­ply didn’t see the video link where they were used to look­ing.  Know Your Audi­ence!

1. Per­son­al­ize your expe­ri­ences

Using test­ing tools like Opti­mize­ly for per­son­al­iza­tion of expe­ri­ences is done by a small per­cent­age of those using these tools, but it is a great fea­ture to tap into.  Again, if you know your audi­ence and can tai­lor mes­sag­ing and head­lines to that audi­ence it can pay off.   Iron Moun­tain per­son­al­ized con­tent based on audi­ence seg­ment and deliv­ered a tar­get­ed mes­sage to spark a 123% increase in con­tent engage­ment.

1. Think like a sci­en­tist

Dan urged us to “Think like a Sci­en­tist” and be not only idea-dri­ven (curi­ous­ly cre­ative and ques­tions con­ven­tion­al wis­dom) but also data-dri­ven (mea­sure, ana­lyze, opti­mize and iter­ate!)

Trunk Club is a great exam­ple of think­ing like a sci­en­tist and going against con­ven­tion­al wis­dom of using the short­est pur­chase path flow.  As mar­keters and web design­ers we try to reduce the fric­tion in a con­ver­sion fun­nel and sim­ply the process – usu­al­ly by reduc­ing the num­ber of steps of pages in that check out process.  Trunk Club went the oppo­site way in expand­ing the num­ber of pages, yet in a sim­ple way.

The Result was a 133% more cus­tomers.  At the new con­ver­sion rate of adding cus­tomers the new check out process has the abil­i­ty to gen­er­ate $10–30MM in rev­enue the first year as Trunk Club cus­tomers.

Now that is worth test­ing for.

Hap­py test­ing, y’all.