How to Use the AdWords Split Test Feature for Your Ecommerce Company

By Sergio Aicardi on August 26, 2015

One of the features available through AdWords is its split-testing feature, but unfortunately many companies are either ignoring this feature or not using it to its full potential. There are essentially two approaches that you can use with this tool to split test your ads (and therefore know how to know when you’ve found an ad that is performing well), so it’s a great option to master in the New Year. Using these methods can help Google AdWords can increase your click-through rate, conversions, and quality score, so the sooner you can get started the better.

Getting Started

Before getting into what you should test and why it matters, it’s important to know where you can find the feature. For most of you this is something you’ve been working with for years, but below is a quick refresher about how to find your ads in AdWords and how you can then find the split test feature:

  1. The first step is of course to log into your Google AdWords account and click on “Campaigns.”
  2. Next you want to choose an ad group under one of your ecommerce campaigns.
  3. From here you will be able to see any existing ads that you have already setup (here you will need to have at least two adds set up so you can split-test them against one another in the future).
  4. Google will show you a preview depending on where your ad is positioned.

I would also highly recommend checking out the following video by QuickSprout that breaks all of the following steps down as a demo if you need extra help: http://www.quicksprout.com/university/how-to-split-test-ads-in-google-adwords/

Once you’re in the right place you’re ready to run a split test. There are four things that you can test within an ad: Your title, your headline, the first and the second line, and the display URL.

So, what can’t you test?

It is advised that the only thing that you don’t want to change when you’re split-testing two ad variations is the landing page. This does not make for a good comparison because if you have different ads going to different pages, it’s very difficult to figure out conversions, which is always a primary goal when analyzing two ad campaigns. While it doesn’t make a difference on click-through rate, it does make a huge difference for conversions.

Two Different Ways to use the AdWords Split Test Feature

Method 1: Split-test Two Different Ads

Ultimately there are two approaches that you can take to split testing. The first approach is to split-test two different ads. Google will show each ad for half of the time and then you are able to see which one has a better click-through rate as well as which has a lower cost-per-click.

When you pick two ads to split test, you want to make sure that they are very different. This helps because when you create a new ad, Google is set on default to give you one of the ads that you’ve already created. From here what you want to do is change this as much as possible rather than taking one small variable and then testing it against one another. That method would take forever to figure out which one is actually the better ad. Some examples of things you can make drastically different include:

  • A completely different headline.
  • A completely different description line one.
  • A completely different description line two.
  • And a different display URL.

When your test is set, click on “Save Ad” and you will ultimately have two very different ads to compare. Below is a screenshot example that shows the difference between two display URLs. As you can see, it’s pretty self-explanatory—split testing is really all about the results.

example of ads 2

Remember, you should have somewhat of an idea where you want your PPC efforts to go before you actually perfect your ads. It’s important you know what pages and what areas of your company you want to use for paid search first before you start testing. For more information about PPC campaigns and where to begin, visit here.

Method 2: Split-test Several Ads Against One Another

The other method you can use is to split-test several ads against one another and put in a different variable for each one. When you click on “New Ad,” you change one thing rather than making two ads drastically different. This will allow you to look at an ad in detail for each of its features. Some ideas of variables to change are the same as mentioned above: A completely different headline, a completely different description line one, a completely different description line two, and a different display URL. Similar to the drastic changes, you can manipulate these variables across several ads (rather than 2).

For example, with a different headline you can keep ads primarily the same, but then change the headlines across each of them. One of the benefits is that you can just alter the headline on each if that is the only thing you want to compare.

On the other hand, you can also keep making these slight variations until you have a different headline, first line, second line and display URL combination for all of your ads, and then from here you will want to run them. You will be able to see based on the click-through rate and the cost-per-conversion and the average cost-per-click, which aspects of the ad are performing best. This allows you to see what you may want to change or use for future ads. What’s great is that you can use this information to help both your online and offline advertisement campaigns.

Below is a great example that shows two different variations of an ad with a few changes and then the results that go with it:

examples of ads

For a list of even more examples, I highly recommend this infographic from Search Engine Land. It offers a great list of different ways you can split test to create the most successful ad.

A Note on Statistical Significance

One thing to keep in mind is that just because one ad appears to be doing better does not make it so! You need to be sure that it is statistically significant before you go banking on one feature performing better than others. This is just as important in the process of using the split test feature.

Stay Organized

Organization and log keeping is essential in this process. You will definitely want to start keeping track of the results and the changes that you make. If you’re doing several split-tests, once you’ve gotten to a really important variable that you are testing and have eliminated some tests that were not important, you want to start making a note of important results for future reference. I recommend visiting this article to learn more in-depth about excel and how to keep things organized (SEO as well as PPC campaigns).

You can also go to “View Change History” and this will show you any changes that you made to that ad group. This will show you information about changes that you’ve made to keywords. This is particularly great if you’ve been doing split testing for a while and you want to go back and see what elements of ads have worked for you and which haven’t before you begin a new ad campaign.

The Takeaway

These two methods are really all you need to know to successfully use the split testing feature. The first being that you can split-test two very different ads against one another, or the second, that you can split-test several ads against one another, each one slightly different. I recommend trying both of these methods before launching any major ad campaign. You also need to make sure that you interpret the data before making any major decisions, and make sure that the differences are statistically significant. Running split tests can have a very positive impact on your marketing strategies.

Do you have experience using Google ad campaigns? Using split testing? Please let us know your experience in the comments section below.

 

 



About The Author

Sergio Aicardi
Sergio Aicardi /

Search engine marketing is what I am passionate about and enjoy doing. I feed off of the knowledge and the fact that SEO is always and will always be changing. Remember, when it comes to SEO, always think like a search engine and you will succeed.