hello and welcome in this video session I’m going to show you how to create a/b
testing using Google Analytics experiments now if you do not have Google Analytics
installed I strongly encourage you start using
Google Analytics and create an account and set up your
tracking code and install it on your website now the reason that you want to analyze your visitors interactions is because it will give you insights as to how you
can make improvements to your website right and Google Analytics is simply the best
tracking platform on internet in Google makes it available for you
freely right so therefore if you are not on board as yet simply strongly consider it and install it and start using okay surely for you to be able to create
a/b testing you need to have Google Analytics
installed and you need to have an account and you have jave your tracking code
on your website so therefore this video session will assume that you
already have those in plane so let’s go ahead and create our AB tests so in your in your Analytics account simply
go in to reporting and simply find on the left-hand side here you’ve got
behavior option then underneath that you have another menu
option saying experiments simply press on that
okay once you do then it should give you the interface
for you to create your experiment simply press on this button here create experiment now as far as experimenting and testing
your website is concerned you can experiment and test everything that makes up your website therefore for you to improve your conversion
rates there is no point in imagining to make changes or thinking ohh should I do this yeah I let this I’m getting web traffic
through organic SEO or through AdWords so
therefore I made the change and I think this is the better version you
see when you want to improve your website improve your conversion
rates okay you need to have a baseline to to evaluate your results there is no point in just thinking about
things right so as far as a/b testing is concerned you can create many different versions of your web pages whether you
have an e-commerce site whether you have a basic static HTML
site whether it’s one page website okay tracking
and testing must be on your to do list it is very important this is how you improve your conversion rates okay by testing and seeing what works and what doesn’t and taking on board what works and you know disregarding what doesn’t will very soon allow you to see better results from your website you can change
your colors on your web pages you can change call to
actions buttons layout messages images you can test everything right so therefore for this example what I want to show you is the simplicity of you running AB tests so simply let’s say I want to test this
URL right I wanna tested against this version as you can see there’s
different call to action different image so that means I’ve already uploaded these files so I
have a web page that is specifically designed for this experiment I don’t want any others
to see this page as far as I don’t want people I don’t
want Google to index this page and you know I do organic SEO for it the purpose of this particular page is for the experiment itself okay as you can see it is different different
heading different images okay and I also have a
different variation which is same again different heading different images so now this is my original page as of now everyone sees this but I want to test to see okay well would this call to action work better than the original or would this call to action work better than the original that’s
what you’re doing that’s what you Test you test difference difference in everything difference on
your web pages your if you have an e-commerce site
different product images different call to actions button or whatever the
case may be what I’m saying to is it is only you
as an online business owner or website owner that you need to work out hmm if I do
this what would happen if I do this what
would happen hmm I believe if I change this hmmm I could perhaps get more conversions right so that means you sit down you think about hmm what can I test what should I test you can test
everything and what should I test how should I test it that
is totally up to you no one can say that to you you need to
work that out once you do then you simply have
different variations online okay whether you’re using static HTML files or whether you are using an content management systems okay it makes no difference if you are using WordPress for example
simply create new pages you know variations of what
ever you wanna test okay that simply that simple that’s how simple it is so we simply type the origin URL that we want to test
in our example it is this particular URL so simply
we place that and then we start the experiment now it’s saying okay you can name your
experiment to something changed headings and images let’s say that’s what I wanna name it that’s totally up to you make it something that you can remember okay now it’s saying objective for
this experiment what that means is let’s not complicate things what that means you can set up goals
goals could be something as simply as did they sign
up for my newsletter or did they contact me or did they by from me whatever the case be okay case may be goals are that you can test your
experiment against goals if you so please or you can test it let’s say if you have an eCommerce site in this option you can test the revenue or
transaction and so on if you have goals defined you can test
your experiment against it in this example we’re just gonna
go and say Bounce Rate I wanna see when people come do they bounce off or do they continue to other pages that’s what I wanna test because
for this kind of experiment you know call to actions and bit of image
change and so on that perhaps is logical for me that’s
what you do here you can test against something okay or visit durations or page views and so on so let’s just choose
bounce bounces okay it’s saying percentage of traffic
to experiment now if you’re making a huge change let’s say a layout you want to change
the entire layout or your web page or website right if that’s the case or you want to
change the menu structure put ’em to the top or left or right and so on then
you can say okay you know what I wanna run this
experiment but I want you’re saying to Google I want you to test fifty percent of my traffic instead of the 100 percent okay so let’s just play it safe what if you
wanna play it safe you choose different options here you
can only say 1 percent of traffic if you want okay so that’s totally up to you let’s
go with 50 percent you can have Google Analytics to email
you if you so please you need to just turn that on if you what get
advanced options we’ve got an advanced options leave things in
default but remember you have different options
here you can say okay I wanna test it for 1
week or three days and so on right set confidence threshold minimum this is the minimum minimum two weeks that sounds reasonable okay if you have a busy website then you
can perhaps you know play around with that confidence threshold leaving as 95
percent but you can go higher or down what that
means is okay for Google Analytics to work out you know for you to say okay you
know what I’m confident that particular variation worked better better than the other so there that’s
what you do you say I wanna be you know this confident or this confident but that ninety-five percent is pretty good anyway right so simply press next step on this page what we’re saying is OK the URL that I wanna test is this now I get to tell Google okay what is what is it that I’m testing against okay so once you you can change the name here or you can say okay I want a different variation so different
variations in our example was a different page so we simply put
that in there or you can continue adding different
variations to your pages okay it allows you add more
and more and more add more more more more okay so just remember you have that option you can test different different
different variations of your pages now we can say press Next now how do you want to set up your experiment
code we want to manually insert it send the code to webmaster if you have a web master you can send the code then he or
she will modify the source code of your
web pages right but surely for us to do our
experiments will manually insert the code which is very easy anyway where the code go is on your original
page this is original page that I’m testing right on that page the code that Google gives you goes just under the head element that’s where you put and you simply upload it now if you have content management
systems like WordPress and so on so therefore whatever the page that you’re
testing you will have a template for it and
simply make sure that this particular code is placed after the head element of your web pages right now you can keep this additional information for your
records but you don’t have to OK now once you do that then you can say next step and it’ll
go ahead and fetch everything for you let’s press next what it’s saying okay now ohh original page does not have the experiment
code okay your variation pages must have the the universal tracking code installed anyway right but as you can see here
it’s saying the code that we gave you cannot be found on the original page
that’s because I have not uploaded it now that I have uploaded it we wait till it fully uploads we can then simply re-validate and should
be okay to go now now it’s saying okay you can add notes to this experiment
now at this stage you may leave it in place
but if you’re running a lot of different experiments surely you can put notes here so that you remember what you’re doing
as far as you’re different experiment are concerned okay we can say okay you know I have changed this I have changed headings and images for 3 3 pages okay then you can then press Start and
it’ll start experimenting then you know depending on the amount of visitors that
you get to your website then you can see the reporting you can
see many many different things as far as you know which one performed better than
the other see in our example we were testing against Bounce Rates okay it will evaluate Bounce Rates using many complex
calculations because it’ll show different variations
accordingly okay so that means when someone comes to my website sometimes they will see this variation
sometimes they will see the original variation sometimes they will see this variation okay now we’re we’re testing only 50 percent of
visitors that means 50 percent will still go to the original where as the other 50 percent
Google work day out for you the other 50 percent will have this or that or this or that shown to them so therefore once you finish your
experiments you can then turn around and say okay your know as an example you can say this particular image and this call to
action worked better than the other two well
if that’s the case then you simply change your original page to include this information see that is the power of experiments and
a/b testing so basically it’s not like a/b testing it’s
like ABCD EFG testing really and once again you can test everything on your webpage regardless of the website that you have
okay simply create a page as in a different version of whatever you
want to test call to actions colors buttons or whatever and then start experimenting to see what works better than the other because
once you do that over very short amount of time you will
have the best possible web content on your website so therefore that will
eliminate imagining things or guessing things and as you see it is very basic to run experiments and a/b tests all you
do is create an account at Google analytics simply find option under behavior on the left-hand side here under
experiments and then just follow the steps as I’ve
shown you and then start your experiment let’s start
the experiment now it is launched it’s saying okay that’s running
now okay it’s started now and you know it’ll run for minimum of two weeks that is the minimum but it will depend on amount of web
traffic that you get okay and then you can see different reports and then analyze the outcome and then make changes accordingly I thank you very much for your visit and I
really hope that you will subscribe to my YouTube channel because I will be having more different
videos available for webmasters like yourself so that you can grow your online business thank you very much for your
subscription I will talk to you very soon thank you bye bye

Google Analytics Content Experiments AB Testing
Tagged on:                                                     

6 thoughts on “Google Analytics Content Experiments AB Testing

  • February 3, 2014 at 1:59 am

    You provide great information with easy to follow instructions and also the the reasons why you'd want to do these tasks, thanks!

  • September 19, 2014 at 7:38 am

    any explaination of what the experiment code does? I'm a developer and I'm wondering what it does so I can code around it.

  • October 13, 2014 at 4:00 pm

    Very happy with this tutorial, thanx!

  • January 13, 2016 at 7:23 pm

    Then the Google Analytics redirect de traffic of Original page to the variations? (until to the end of experiment.)

  • June 1, 2016 at 7:30 am

    And i…. thank you very much for your effort to explain The A/B Test with G.A

  • September 12, 2017 at 7:59 pm

    Amazing sentence .. its not AB testing, Its ABCDE testing..
    I understood your point that we have to add canonical tag to let google know that the content is not duplicate.
    Now the doubt is:
    What about variant 1 and 2. Do we need to add in robots.txt , so that google won't index it and show in SERPs.?

    Thanks for the amazing content @RankYa


Leave a Reply

Your email address will not be published. Required fields are marked *