What does it mean to split-test a Sales Page?

Discussion in 'ClickBank' started by homesure, May 15, 2010.

  1. #1
    I was asked this question some few days back by one of my PHP students and I said to him "It means splitting traffic into equal parts and sharing to different variations of your sales pages that promote the same product to see which one brings in more sales."

    I think my definition was not wrong. If you are promoting some product, there are TWO things you may do to greatly increase sales:

    1. Pre-sell
    2. Split-test your sales page

    I have seen some guys doing random splitting. You don't get accurate test result here because 1 page ends up getting more traffic than the other. Evenly sharing traffic is ideal as all the tested pages get equal share of the traffic.

    What should you consider when split-testing?
    • Headline message
    • Headline color
    • Videos
    • Bonuses
    • Call to action button
    • Long or Short sales copy
    • etc
     
    homesure, May 15, 2010 IP
  2. NCMedia

    NCMedia Well-Known Member

    Messages:
    1,827
    Likes Received:
    98
    Best Answers:
    0
    Trophy Points:
    115
    #2
    Test #1:

    Your first sales page with your graphics and copy being long winded and aggressive.

    Your seconds page is shorter, with less copy, and it's a bit more personal one to one softsell with less aggressive calls to action

    Your third page is a flash promo with audio bits throughout the page and it goes through 2-3 pages instead of all one long one (a continue button on each page...).

    ---

    Now you run all three of those pages either on a rotator or just target each page within PPC/other campaigns. You throw 1000 uniques at each page and you'll see that one will perform best, and one will be crap.

    ------

    Delete the crap one, and the one that came in second place you make edits to and try to improve, maybe influence a bit of the winners formula but not too much.

    Throw another x,xxx uniques at each, you'll pull a winner again.

    Repeat, the landers may start looking closer and closer to being the same, and you're just tweaking bits of copy, graphics, above the fold elements etc.

    DONE. Now you know that this page converts the best and you did a simple split test to prove it.


    You can now do the same for your price points, your upsells, you actuall PPC ads/keys/groups etc. Split testing is done on everything in life too - you try 3 toothpastes you stick to one, you try 20 cereals, you like and buy 3... Figure out what they customers like and why, but also figure out what your prospects DON'T like, and improve it.



    EDIT: I'm an idiot, I actually just really tried to help you by answering your question, then read through and you're just trollin' for sig clicks to your $5 split test guide.

    /fail. *You also should outline whether it's for aff's or vendors as obviously objectives are not the same.
     
    NCMedia, May 15, 2010 IP
  3. homesure

    homesure Well-Known Member

    Messages:
    114
    Likes Received:
    2
    Best Answers:
    0
    Trophy Points:
    133
    Digital Goods:
    1
    #3
    Great great great!
     
    homesure, May 15, 2010 IP
  4. jacky8

    jacky8 Active Member

    Messages:
    1,416
    Likes Received:
    23
    Best Answers:
    0
    Trophy Points:
    80
    #4
    Split testing is a very long, hard and expensive process.

    Suppose if you have only 5 things to test and you want to test them the right way, you need to do all of them one by one. It costs a lot and your cost increases with the number of variables you want to test. It may also happen that a thing that you eliminated may work better with something down the line and you will never know about that.

    Moral of the story: Split testing is expensive and it can take at least $250 - $500 to split test even a single variable. It is a minor part of getting your product to convert well. What really matters is your copy / layout / offer quality and not the split testing budget / skills :D
     
    jacky8, May 15, 2010 IP
  5. NCMedia

    NCMedia Well-Known Member

    Messages:
    1,827
    Likes Received:
    98
    Best Answers:
    0
    Trophy Points:
    115
    #5
    ^ (OP) Please do not use that in your guide/software/etc. you are selling.

    And: Why don't you just make a thread promoting your thing properly instead of trolling for sig clicks with all your threads?
     
    Last edited: May 15, 2010
    NCMedia, May 15, 2010 IP
  6. johnjohn24

    johnjohn24 Peon

    Messages:
    129
    Likes Received:
    0
    Best Answers:
    0
    Trophy Points:
    0
    #6
    Forum Trolling :mad:
     
    johnjohn24, May 15, 2010 IP
  7. SEOwriterplus

    SEOwriterplus Active Member

    Messages:
    100
    Likes Received:
    4
    Best Answers:
    0
    Trophy Points:
    78
    #7
    Well... if you don't get the same numbers of views, responses, or whatever you'd like to measure,
    you could also do some calculations (I can't think of the word for the exact process now), but basically
    if you get 100 views for one page and 200 on the other, you could double the numbers of page
    (conversions, clicks, whatever) and still compare which one did better.

    Of course the higher the numbers, the more significant and useful the data.
     
    SEOwriterplus, May 16, 2010 IP
  8. MarkAndrews IMCopywriting

    MarkAndrews IMCopywriting Peon

    Messages:
    112
    Likes Received:
    2
    Best Answers:
    0
    Trophy Points:
    0
    #8
    For copywriting split testing, I endorse Mike Humphreys (Superb Copywriting) product...

    Check it out right here: http://www.easymultitracking.com

    Non affiliate link.

    He's also a member here at DP and at the Warrior Forum and is a very well respected copywriter.

    Best,


    Mark Andrews...
     
  9. Orangespecka

    Orangespecka Guest

    Messages:
    26
    Likes Received:
    1
    Best Answers:
    0
    Trophy Points:
    0
    #9
    gonna check it out - tks
     
    Orangespecka, May 17, 2010 IP
  10. MarkAndrews IMCopywriting

    MarkAndrews IMCopywriting Peon

    Messages:
    112
    Likes Received:
    2
    Best Answers:
    0
    Trophy Points:
    0
    #10
    If you want a free split testing tool for A/B multivariate testing you could always check out Google Website Optimizer.

    It's pretty simple to use and with this Google Website Optimizer tool there are of course very easy tutorials on hand, to take you step-by-step through the entire process.

    http://www.google.com/websiteoptimizer/b/index.html

    I hope this further information helps a few of you out looking to split test your copywriting results.

    Best,


    Mark Andrews...
     
  11. mcapodici

    mcapodici Well-Known Member

    Messages:
    228
    Likes Received:
    2
    Best Answers:
    0
    Trophy Points:
    120
    #11
    Someone said split testing is hard and expensive. It's neither!

    As the previous post mentions use Google Web Optimizer - it is free, simple to use and very efficient.

    I have been using it to split test sales pages. I have set my "Goal" to be a visit to the order form rather than a sale (because I am not getting enough sales to get meaningful results in a short timeframe).

    You can set your goal to be an order form impression, sign up for email, a sale, visit to product spec page whatever you like.

    When your goal is reached it is called a "conversion".

    100 conversions are needed to make a good assessment of whether page A or page B is better.

    My first split test improved performance by about 20%.
    My current split test has just started but hopefully will again improve performance.

    I advise to split test different headlines, alternatives for the first paragraph, different graphics, colours, buy buttons etc. The "big stuff". Don't split test a change in the 30th sentence for heavens sake! Split test one little thing at at time.

    I am to keep split testing until my affiliates earnings per hop > $1 then I can boast about it on here :cool: to attract more affiliates.

    In tandem with split testing it is good to know what a good sale letter should look like - so check out the books by Maria Veloso on Amazon.

    Don't be afraid of split testing. I was for ages and now I've started its the best thing I've done and I feel like I am really in business not just guessing anymore.
     
    mcapodici, May 18, 2010 IP
  12. Reedge.com MoreConversion

    Reedge.com MoreConversion Peon

    Messages:
    25
    Likes Received:
    0
    Best Answers:
    0
    Trophy Points:
    0
    #12
    Indeed its not hard to split test. There nice tools out there that can assist even better then GWO.
     
  13. laxman363

    laxman363 Active Member

    Messages:
    2,173
    Likes Received:
    81
    Best Answers:
    0
    Trophy Points:
    90
    #13
    split test should be considered only if you have enough money.if you are on atight budget then go for a sales copy that you think will do the best earn money and then do some split testing
     
    laxman363, Aug 14, 2010 IP
  14. Reedge.com MoreConversion

    Reedge.com MoreConversion Peon

    Messages:
    25
    Likes Received:
    0
    Best Answers:
    0
    Trophy Points:
    0
    #14
    Ehh testing tools are free so lets combine it. Get salescopy (paid) and test it on free testing tools like google web optimzer and http://www.reedge.com
     
  15. Personyze.com

    Personyze.com Peon

    Messages:
    3
    Likes Received:
    0
    Best Answers:
    0
    Trophy Points:
    0
    #15
    When doing split testing you might want to limit your test only to specific visitors and not 100% of them or the first 1000.

    To do so, you need a tool that offers smart segmentation. A tool that will allow you to define tests only for a very specific group of visitors.

    With tools such as Personyze.com you can define a test that will only run on specific group of visitors such as "From New-York" AND "not visited the site before" AND "using an iPad" AND "have more than 100 Facebook friends" AND "came from Google with specific keyword" etc...

    This segmentation capability will set your test free, allowing you to be very flexible in which visitors you would like to include/exclude from your test .

    HTH.
     
    Personyze.com, Jun 5, 2011 IP