We have updated our Privacy Policy and Privacy Options

Got It

Adobe Test&Target Superheroes: Testing Assumptions by Using Segmenting and Targeting

Share

They are common terms used by marketers everywhere.  They are powerful, full of action and when it’s time to ramp up efforts, you know you can always count on the Batman and Robin of marketing –  segment and target.

Batman and Robin

Mainstream marketers talk about “segmenting this traffic” or “targeting that group” – pretty basic marketing concepts based upon speaking to a specific audience.  Adobe Test&Target conversion experts use these terms a little differently.

For example, you may hear Brooks Bell conversion experts say, “We’ve analyzed the affiliate traffic segment and recommend creating a target to show those users a specific offer.”  As this example illustrates, Adobe Test&Target segments and targets are actually quite different from each other.  They are two features that are incredibly helpful on their own, and are even more powerful together.  Hence the Batman and Robin reference.

  • Segments have specific characteristics and are used to compare experiences, and are primarily used to provide marketers with data and insights.  To identify the traits of different segments, marketers should review the overall conversion data, and then compare it to each individual segment.

Here’s an example of what segmenting by traffic source, a very common segment, looks like in the Adobe Test&Target reporting. 

Capture

Cool Adobe Test&Target Reporting Feature #225:  When viewing a campaign report, you may see the word “interesting” called out next to one of your segments.  This lets you know that there’s something thought-provoking happening and that you should check it out!

Today’s example Adobe Test&Target conversion data reveals:

  • Overall Traffic has a 6% conversion rate.
  • Segment:  Branded SEM traffic has an 8% conversion rate.
  • Segment:  Affiliate Site Traffic has a 2% conversion rate.

Based upon this conversion data, here is an example of the marketer’s analysis:

  • Branded SEM traffic converts at a higher rate.  Why, and how can these principles be carried throughout all segments?
    • It may be because branded traffic specifically sought this experience, the landing page was what they expected and there was little friction.
  • Why does affiliate site traffic convert at a lower rate?
    • If marketers apply the same principles from above, they could assume that users did not have a clear picture of where they were heading and that there were major friction points once the user got to the landing page.

The marketer has a hunch that the content of the affiliate site link was not optimized (the link did not accurately describe the offer, displayed old branding, etc.), or that users were confused when dropped into another brand’s site (no co-branding, not a clear tie-in).

As Greg Ng, CXO at Brooks Bell says, “hunches never trump conversion data”, so these assumptions must be tested.  Marketers can use a/b split tests per segment, or get even more detailed by targeting.

  • Targets are used to include/exclude visitors in campaigns, experiences and mboxes when testing. Marketers can create targets that will determine what experience a user sees.

Let’s continue to use the example from above to illustrate the concept of targets.

The marketer has realized that the conversion rate of users coming in from an affiliate site is low, and they’d like to increase this conversion rate by providing affiliate traffic with a more specific experience.

To do this in Adobe Test&Target, the marketer should create a campaign that will test the assumption that creating a co-branded experience for users coming from the affiliate site will increase conversions.

  • The Control will be the current experience.  This will be the experience that users coming in from the Branded SEM traffic (determined by the segment) will see.
  • The Challenger will feature a co-branded experience created by an mbox (the topic for my next blog post – so stay tuned!).  This is the experience that only users coming in from the Affiliate Site (determined by the segment) will see.

By utilizing the Adobe Test&Target reports, marketers can continually monitor and analyze the conversion data to determine if their assumption was correct.

Here’s an example of an Adobe Test&Target summary report.

Rpt

In summary, marketers can create assumptions by analyzing conversion data per segment.  After assumptions are made, the decision comes down to a/b split testing or more detailed testing with targeting.  Once marketers start this process, they’ll discover more insights, more assumptions and more testing opportunities.

There’s just one lingering question…  If you’re wondering who is the Batman and who is the Robin in this scenario, I’ve got the answer.  Picture Robin hanging back in the shadows watching the Joker.  He’s taking notes and analyzing any of the Joker’s weaknesses (data).  Once he’s got some good information, he takes it to Batman.  Batman formulates a plot to exploit (test plan) those weaknesses and when the time is right, he goes in for it (launch the test!).

You guessed it.  Robin is segmenting and Batman is targeting.  Segmenting is the data and the brains behind the operation.  Targeting takes action on the data by testing.  Use this dynamic duo to see the action bursts of Pow, Wham, Smack – or in this case, Increase, Profit and Win!

My previous blog post, Using Adobe Test&Target, Increase Your Off-Peak Conversion Rate with Segmenting, focuses on how to set up time-based targets in Adobe Test&Target.  Check it out!