Multivariate Emai Marketing & Landing Page Testing Questions!

4 replies
Team,

I have a question.

We have 2 email lift letters: a long form and a short form.

We are testing only one subject line.

We have 2 landing pages to test.

We have 3 headlines within the body copy of the long form lift letters but I think we will only settle on one to make the test easier.

Can I trouble the WF a/b split testing/multivariate testing experts to give me some feedback on what would be the right way to get the highest and best practices for conversion with these email tests?

We have 3 separate lists that we are testing. One we have 6 sends against, one we have 3 sends, and another 5 sends. All three are different email list providers.

Thanks,
PRW
#ab testing #emai #email marketing #landing #landing page #marketing #multivariate #page #questions
  • Profile picture of the author twistedpixel
    I would do a straight split test on the email and use multivariate on my lander to accelerate and simplify your campaign.

    The multivariate will allow you to send all your traffic to one page and test various elements on page at the same time.

    Split testing your landers can be a slow process due to the limitation of testing only one element at a time.

    You can use the free Google Website Optimizer for the multivariate stuff and with a decent amount of traffic (2k+) you will see clear winners emerge.

    Good luck!

    Just curious...are your landing pages static html or hosted on wordpress? I'm wondering as I have a new WP plugin for mutivariate that needs testing (easy to use) and wouldn't mind hooking you up if that's the case.;-)
    {{ DiscussionBoard.errors[4640901].message }}
  • Profile picture of the author David Keith
    Originally Posted by paulrwhitacre View Post

    Team,

    We have 3 separate lists that we are testing. One we have 6 sends against, one we have 3 sends, and another 5 sends. All three are different email list providers.

    Thanks,
    PRW
    when testing anything, the key is to keep some constants. you cant accurately test the effectiveness of headlines or email copy when you are sending to different lists on different providers.

    by providers do you mean autoresponder providers or email list providers such as buying / renting lists. if its the later, your results will be much more dependent on the quality of the people they provide than anything (which varies a lot by provider in most cases).

    also, you mention 3 separate lists, were these subscribers collected using the same methods. ie...free gifts or paid subscribers. and have you tested how each of these 3 lists respond to the exact same offer.

    if they normally perform the same with open rates and click throughs then you may be able to get accurate data. but if you have not tested how similar these 3 lists perform, you would have to go about testing every variable on every list.
    {{ DiscussionBoard.errors[4640944].message }}
    • Profile picture of the author paulrwhitacre
      Twisted,

      We have a wordpress site but we are simply using HTML landing pages.

      Can you explain what you meant by a 'straight split test' on the email?

      What gets split?

      Owslaw,

      They are rented lists.

      The one list we have 6 sends to, I think I will send the same subject line, the same body copy, but 3 times send to the 4 offer and 3 times send to the 2 offer.

      That or I could send all those to the 4 offer and change up short form for long form. Then the 4 offer could remain a constant. It's tests of 12000 people per send. I don't know if our open rate/click through rate will be high enough.

      Any other thoughts?
      {{ DiscussionBoard.errors[4646093].message }}
      • Profile picture of the author twistedpixel
        I meant a basic A/B test. I'd probably test copy (subject/headline/call to action, etc.) versus number of words.

        I hope you share your chosen methodology and results as I haven't done much email testing and would love to learn what worked for you.
        {{ DiscussionBoard.errors[4647442].message }}

Trending Topics