9 {{ upvoteCount | shortNum }}
9 {{ upvoteCount | shortNum }}

Best practice for split test goal?

by Hopeless Bromantic
8 replies
I'm getting ready to test a call-to-action button on my site's homepage. This being my first test, I also need to figure out what page to choose as the test goal.

On the one hand, clicking the button sends the user to the first step of a free registration process, so it could make sense to use that page and find out which button option gets more registrations. OTOH, since this is a freemium site, one could argue that any "goal" would have to be whether the user ultimately purchases a subscription, and therefore the payment page would have to be the goal.

Since I couldn't find any articles to address a similar issue, I hope those of you who have experience with split testing can contribute your $0.02. Thanks!
#a/b #goal #practice #split #test
  • Profile picture of the author TheContentAuthority
    Hopeless,

    If you are talking about goals in analytics, you will have to make two separate paths in order to track two pages as goals. Otherwise you will only be able to do goals on one page.

    Is there a reason you can’t have both landing pages send the visitors through the same conversion funnel? You’re a/b test setup should record which page was more successful.

    Best,

    Shawn
    Signature
    Outsource to the experts...

    We customize your Blog, eBook, Press Release and Sale Copy content with your message.

    {{ DiscussionBoard.errors[8120931].message }}
  • Profile picture of the author Marc Rodill
    The simple answer to your question is to use both pages for your goals, at least if I understand your question correctly. You want to measure clicks to both the registration page and the page after registration. That way you know exactly what's going on.
    Signature
    Long Lost Warriors! The Secret Sales System! Act Now! Buy Now! Right Now!
    {{ DiscussionBoard.errors[8120944].message }}
    • Profile picture of the author Hopeless Bromantic
      Originally Posted by Marc Rodill View Post

      The simple answer to your question is to use both pages for your goals, at least if I understand your question correctly. You want to measure clicks to both the registration page and the page after registration. That way you know exactly what's going on.
      I didn't realize it was possible, but having multiple goals does sound like the most reasonable way to go. Thanks, guys.
      Signature

      Bros find strictly platonic dudes on seekBromance.com
      _______________________________________________
      "It's pretty simple. You work hard, you believe anything is possible, and you try to make the world better."

      {{ DiscussionBoard.errors[8121016].message }}
  • Profile picture of the author mbacak
    Honestly, this doesn't seem like a real split-testing to me.. this just seems optimization.

    Ok, I know everyone says making little changes is testing and yea technically it's true.

    However, to me split-testing slight a variation gets you slight change.

    Hence, why I call it optimization.

    I suggest testing radicals, such as a radical different look and
    feel so you get radically different variations in change.

    EXAMPLE of Radical: "UGLY as hell page" vs "Pretty page."

    Then do something like you are talking about doing.

    But, I don't want to hold you back from doing something cause
    you are a lot farther then most people because most people test and
    if they say they do they typically lie from my experience.

    Just some testing wisdom from someone that tests more then everyone I know.
    Signature
    The Listbuilding Club (join by clicking below)
    {{ DiscussionBoard.errors[8121472].message }}
  • Profile picture of the author John Romaine
    Originally Posted by Hopeless Bromantic View Post

    I'm getting ready to test a call-to-action button on my site's homepage. This being my first test, I also need to figure out what page to choose as the test goal.
    Hang on.

    Correct me if I'm mistaken, but if you're testing CTA buttons, then surely you should be running two buttons against each other to determine the most effective, yes?

    And NOT the destination page?

    Or have I misunderstood you?
    Signature

    BS free SEO services, training and advice - SEO Point

    {{ DiscussionBoard.errors[8121543].message }}
    • Profile picture of the author Hopeless Bromantic
      Originally Posted by John Romaine View Post

      Hang on.

      Correct me if I'm mistaken, but if you're testing CTA buttons, then surely you should be running two buttons against each other to determine the most effective, yes?

      And NOT the destination page?

      Or have I misunderstood you?
      The destination page is an obvious choice, I agree. I just wonder whether it's the one that gives you the most meaningful data. Maybe it is; I figured it wouldn't hurt to ask.

      Some people have said that, if your ultimate goal is to convert as many visitors as possible to buyers, then you should use the purchase page as a goal for everything. Sounds weird to me, which is why I wanted to find out if other warriors were doing anything like this.
      Signature

      Bros find strictly platonic dudes on seekBromance.com
      _______________________________________________
      "It's pretty simple. You work hard, you believe anything is possible, and you try to make the world better."

      {{ DiscussionBoard.errors[8123314].message }}
  • Profile picture of the author drewfioravanti
    Depends on what you want to find out. Do you want to find out which button gets more clicks or do you want to find out which button leads to more registrations?

    If you can answer that, you know which page to make the goal.
    {{ DiscussionBoard.errors[8123339].message }}
  • Profile picture of the author hustlinsmoke
    Dang thats very simple and you who always post negative things in threads can't do it. What a shame.
    {{ DiscussionBoard.errors[8570510].message }}

Trending Topics