You NEED To Be Split Testing - And You're Only 3 Simple Steps Away From Doing It

13 replies
Howdy Warriors,

How many of my fellow warriors are actively split testing everything that they do? And if you aren't yet...why not?!

If you don't know how, hopefully this post will help you out. It really is super easy to start split testing, so even if you are only get 10 visitors a day to your site, the sooner you start the faster you can start improving your site on a scientific basis, and not just by gut feeling.

So here ya go, 3 simple steps to start split testing your site. Let's take an example squeeze page, since that's probably the first thing you should really focus sending traffic to if you're trying to build a list (which is the best idea IMO)

Step #1: Spend some good, quality time looking at the sales copy of your page. Even if you are just trying to get an email address, you have to use very persuasive copy to 'sell' your free product.

So put a lot of time into the headline and bullet points and try to see it from your prospect's perspective. After you have it looking the best it can, simply copy the page so there's 2 of the exact same squeeze page.

Then change ONE thing on the copied page (headline is a good place to start) and just name it something similar. So now you have www.yourdomain.com/squeezepage and www.yourdomain.com/squeezepage2

Step #2: Go to Google's Website Optimizer and set up your account if you don't have one yet. Follow the simple instructions they lay out for you to set up your split test.

This will include:
  1. creating a new split test
  2. inputting your 2 squeeze page variations as well as your conversion page (which is the page your visitors will see next, usually a 'thank you' page or OTO page)
  3. installing the javascript code on your pages. They give good examples to show you how to do this. Or if you're like me and love Wordpress, then you can download the "Google Website Optimizer for Worpress plugin, which makes it awesomely easy.
  4. And pushing the done button!
Step #3: Now all you have to do is drive lots of traffic to your site. Ideally you should wait until you have 100 conversions per page to know which one is the winner. So in this case you would need 200 conversions.

So if your squeeze page is converting at around 40%, then you only need 500 visitors to your site and you'll have a successful test. After you get all the conversions you need log in to your google account to determine which page is the winner.

Then trash the other page, copy the winning page, create another variation and do it again!

That's really all there is to it. Sure you can start testing and tracking more and more as you get deeper into it, but this is a great way to get your feet wet and start improving any landing page you're using.

I hope this post helped some of you out there that have maybe been a little intimidated to start testing.

If we could get some more experienced marketers to chime in with some more testing and/or tracking tips, I (and others I'm sure) would GREATLY appreciate it.

Have a great day warriors!
Tyson
#split
  • Profile picture of the author Lisa Gergets
    (Psssst...you could expand this and do an entire instructional WSO on how to split test!)

    And no, I don't split test. {smacks self with newspaper} I really REALLY need to but it seems kind of overwhelming...I barely have time to test, let alone split test.
    Signature
    Sign up to be notified when Success on Demand goes live, and receive a FREE mindmap that you can follow to create and launch your OWN IM PRODUCTS!
    {{ DiscussionBoard.errors[1929093].message }}
  • Profile picture of the author acrasial
    I like your threads! LOL! I haven't really split test much up till now, honestly...because it works, and has been tested by others already.


    But I do always wonder if it can be done better, and once I got another warrior to make a completely different opt in page for me....but never did use it, because a different warrior suggested that what I had works already (since I know it works and they know it works), and gave that as a reason to shoot down my idea.
    {{ DiscussionBoard.errors[1929098].message }}
    • Profile picture of the author Tyson Faulkner
      Hey thanks

      My biggest problem with creating landing pages is that I always have a lot of different ideas and ways to word my headlines and bullet points. So instead of guessing, I just set up a quick split test and forget about it for awhile until I have enough conversions to show which ones my visitors react to better.

      This way I don't have to say "What if..." to my copy. Plus it never seems to be that great right off the bat, so this gives me something to look forward to and know that it will improve over time.

      You guys should really try it! It literally takes 5 minutes and can really improve ROI and how much you get out of the traffic you have.
      {{ DiscussionBoard.errors[1929152].message }}
      • Profile picture of the author innocent07
        Banned
        I think it is on everybodys list of 'things to do', but some people still havn't done any testing.
        {{ DiscussionBoard.errors[1929160].message }}
  • Profile picture of the author Quentin
    I guess if your business needs split testing because you are using landing pages etc then it is a good idea.

    I personally prefer to get my customers opinion and watch my google stats to see where people are going and what they are reading. This then gives me a good idea of what they want.

    Because I mainly draw customers with content and don't really use landing pages it is more difficult to split test.

    I do a sort of split testing by competing with myself.

    My products are fairly general so I tend to build opposing sites which I guess is a type of split testing.

    For example for my streaming audio product I niche it down so I have streaming audio for schools, streaming audio for churches, streaming audio for poets etc.

    Then I play them all off against each other. This means that instead of getting one lot of traffic for just one site I get multiple traffic for many sites.

    Quentin
    {{ DiscussionBoard.errors[1929184].message }}
  • Profile picture of the author jazbo
    Nope, did split testing, got bored, gave up. I know, I know, but I would rather stick pins in my legs.
    Signature
    CONTENT WRITER. Reliable, UK-Based, 6 Years Experience - ANY NICHE
    Click Here For Writing Samples & Online Ordering
    {{ DiscussionBoard.errors[1929340].message }}
    • Profile picture of the author krcorser
      "I know, I know, but I would rather stick pins in my legs."

      hahahahaha...
      {{ DiscussionBoard.errors[1936194].message }}
      • Profile picture of the author ristvin
        I'm an avid split tester... I've used a number of split testing software over the past years... It's amazing what differences in conversion rate one can achieve by split testing...

        I've done simple A/B split testing to multivariate testing... Only tried Google once when it first came out with the Site Optimizer... Went back to the software that I had already acquired as they did a better job...

        Have used Paul Hanncock's A/B Power Split Tester... Started with that script... up to the Split Test Accelerator which is a multivariate, Taguchi type test script... I recommend Tim Robinson's Conversion Chicken to clients these days as it's inexpensive and a great multivariate tester....
        {{ DiscussionBoard.errors[1937902].message }}
        • Profile picture of the author Tyson Faulkner
          What do you find that those software programs offer that Google's tool doesn't?
          {{ DiscussionBoard.errors[1937942].message }}
        • Profile picture of the author povchef
          When I started out the idea of split testing just seemed like a lot more work I didn't have time for and it cost me time and money, honestly the money isn't the real kicker it the amount of time I 'WASTED' not split testing.

          Now I test adds both free and ppc. website headers/sub-headers, emails, sales type... I do find that my list are a great way of testing a lot of little word changes.

          Yes it takes a little longer to set up two adds and find the ones that work best or the header that gets the most conversions but its as important as having a product to sell in the first place.
          {{ DiscussionBoard.errors[1937991].message }}
          • Profile picture of the author andrewwilkie
            Well,

            I'm doing my 1st split test to a sales page with GWO . If I can figure it out, so can you!

            I've made some very simple changes.

            Added a extra buy button.
            Changed the scrollbar colour.
            Added a graphic.
            Made the guarantee stand out with a border.

            Yes, all together on one page. Even with a very small amount of traffic. My conversations are up. This test has only been going 5 days and I've not reached even 100 visitors.

            Yet, I made no changes or tests with the copy itself yet.

            I'm guessing the secret is create your own "Page-Ops Manual, so that each new page has a better starting position and when somebody says try X. Do it. Even if you don't think it sounds a very good idea. At least you'll know.

            If it's a hassle, get somebody else to do it for you.
            {{ DiscussionBoard.errors[1963425].message }}
            • Profile picture of the author GypsyRay
              Thanks for starting this thread! This was my exact questions. I'm going to try Google Optimizer first and then Conversion Chicken to see which one I like better.
              {{ DiscussionBoard.errors[2430596].message }}
  • Profile picture of the author dinahbee
    Banned
    [DELETED]
    {{ DiscussionBoard.errors[1929387].message }}
    • Profile picture of the author Tyson Faulkner
      Hey Quentin,

      Do you have some sort of landing page that you send that traffic to though? Are you building a list of prospects or anything? Are you using a sales page to sell your products?

      IMO it would be wise to split test anything you can to see how your customers and visitors react to different content layout, headlines, adding audio/video, etc.

      If you're using Wordpress I saw a split testing plugin the other day that allows you to split test two posts to see how they convert. Maybe try a different call to action or different intro.

      Just some ideas, but testing can be huge in your business if you are consistent with it.
      {{ DiscussionBoard.errors[1931777].message }}

Trending Topics