A/B Testing: The Scientific Method Of Marketing

how_ab_works_diagram_695

A/B Testing is the scientific method of marketing. It is used to test different aspects of marketing techniques your business uses to find out which one is more successful than the other. Lets use the diagram from above as an example. Lets say your business is trying to figure out to gain more traffic to your website through subject lines in your email. You have a Group A (Control Group) and Group B (Test Group).  Group A is going to be your control group, so if you have a subject line that you  normally use or a constant format for your subject line then you would still use it here. For example I am responsible for sending out the emails for my marketing club. Typically I use the subject line “SMA Meeting (Date)” This is what I would use for Group A. Now Group B is your testing group. This is where you use your experimental subject line. For example to draw more members to come to the next weeks meeting I might use  the subject line: “SMA Networking Social”.

After you have your Group A and Group B it is time to pick who should be sent which. This should be completely random. You don’t want to be bias towards who sees what because that could cause your data to be skewed. Some suggestions would be just randomly choosing half and half or creating an excel sheet and choosing a random number and then counting the list till you get to that number. Once you get to that number take that email and start all over again.

Once you have your groups set make sure to set a conversion goal. For example with the email subject line you could use a conversion of how many people open the email or if that is showing little data, how many people go to your webpage through the email link. One thing to keep is mind is that you need to make sure you are testing both of these subject lines at the same time so that  everyone is getting the same information in the email and to minimize the outside variables affecting the users decision to open/not open the email.

When it comes to email a great site to use is MailChimp. Follow this link and it will give you the steps to creating a A/B campaign through their site : http://kb.mailchimp.com/article/how-can-i-run-an-a-b-split-test-campaign-on-my-content/. MailChimp can track how many people open your emails or follow the link on the email and other exciting things! If you are deciding to test variables on your webpage you can easily create a code to embed into your website. Visual Optimizer can take you through the steps  if you follow this link: https://visualwebsiteoptimizer.com/ab-testing/.

So you have set up your conversion rate and now it’s time to get testing! Like I stated earlier, test both groups at the same time.  Make sure you don’t stop your test too early. If you end a test too early than you will not have as much information as possible and you may not be able to make valid conclusions. On that note make sure you don’t end too late because this could begin to make you lose conversion rates all together. I wish I could tell you what the perfect length is for your A/B testing but it really all depends on what your testing and what your goals are.  Luckily Visual Optimization has a handy-dandy calculator that  you can use to figure out a suggestion of how long you should keep your testing going. The calculator is located through this link: http://visualwebsiteoptimizer.com/ab-split-test-duration/.

Now once you have tested to the fullest you get to look at the data. To me this is the funnest part. Being able to analyze data and see if your hypothesis was right. Sometimes the results can be way different from your hypothesis but you do not know best. 

1

Yes, really. Numbers tell the truth as long as you did your testing correctly. Granted there will be some skewed data and sometimes your A/B test is going to fail, but do not go with your gut feeling. Also if you do not like the numbers that are showing half way through your test DO NOT MAKE CHANGES. That will cause more variations in your data and you would have to throw all of your data away. So once you have analyzed your data the one that reaches your conversion goal the most will be the one you want to go with.  For example if one subject lines gives you 25 clicks and the other gives you 5 clicks, then you are going to want to go with the 25 click subject line.

That is A/B testing in a nutshell. Make sure you keep in mind that A/B testing does not work every time and you should not feel discouraged if it shows no results. You should keep doing tests over and over because you can only improve your marketing techniques. I think it is kind of obvious why A/B testing is so important  for your business. By partaking in this method you are going to improve your websites and your marketing because the users are going to be pointing out what needs improvement without even knowing they are doing it. This is only going to bring greater traffic and leads to your site.  This also makes sure that your business is focused on constant improvement. If you would like to read more about A/B testing and see some successful cases follow this link: http://visualwebsiteoptimizer.com/case-studies.php

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s