A/B Testing For Google Search Ads
In our last article, we looked at the options in Google Ads to allow you to target what potential customers are searching for with KeyWord research. In this article, we’ll take a look at why and how to test your ads on a regular basis to ensure that you’re maximising your spend. By checking that you have the best performing ads, you make sure that you don’t lose any potential customer conversions.
While there are not that many items to test with a Pay-To-Click (PPC) ad, you should still plan and execute these tests carefully. Let’s take a look at the process below.
What Should I Test?
When it comes to setting up A/B testing for your Google Search ad, you need to consider four items.
The headline
The main text body of the ad
The link
The keywords that trigger the ad to display
Headline
The headline is the text that is going to show up as a blue link to the search results. It’s important that the headline reflects what is being offered at the other end of the link. It should be short and to the point and include the keyword being searched for if possible. These keywords will then appear in bold on Google so it helps your ad pop a little more on the screen.
Text Body
This is the more expansive text that gives your potential customers more of an idea as to what you are offering.
Link
You should give careful consideration as to where in your website you will take users when they click on the ad. Where they end up can affect the conversion rate as much as the ad itself. In general, the link should not take users to your home page. Rather, there should either be a specific landing page for the ad or you might take them to a product page related to their search.
KeyWords
We covered keyword research in some detail in our last article. However, a/b testing is a way to find out which keywords from your list are most effective. Pay attention to which keywords work well with which ad variations and ensure that you focus your spend in that area. Some ads may perform better with specific search terms.
Once you have identified the items that you want to test, you’ll need to generate values for each (for example, if you decide to test the headline, you will need at least two different headlines).
If you decide to test multiple items then you will need to run more tests. For example, if you test two versions of the headline and two versions of the link, you will need to run four different tests. The more items and versions you change, the more tests you will need to run.
What Are You Testing For?
A/B testing for Google Search ads can be done for a number of reasons. It’s important to be clear as to the main objective of the campaign from the start. For example, are you looking to add more contacts to your email list? Are you trying to generate more leads or simply trying to drive sales of a specific product? Without clarifying the purpose of the campaign it becomes very difficult to analyse the results of the testing.
Also, be clear as to the difference between click throughs and conversion rates. While click throughs (simply clicking on the link) may be good enough to drive awareness of your business, conversion (purchase, sign up etc.) is probably a better measure to track.
You may find that some keywords have good click through rates but poor conversion. It could be that people searching for those terms aren’t ready to purchase (or sign up) just yet and could just be looking for more information and your product or service.
How to Measure Success
As mentioned above, tracking the click through rate and the conversions is going to be an important measure of the testing process. Focus on those combinations of headline, text, link and keywords that maximise both the click through rate and the conversions.
It’s best to let the test ads run for between a few days and a couple of weeks depending on the amount of traffic that is being generated. Take note of where your ad appears during that time as well. If your bid is too low, you may find that your ad is showing up too far down the list to be effective.
Look for test results that produce a clear winner, making slight changes to each item until you find the very best combination. Remember that there needs to be consistency between your ad and the page on which the searcher lands on when they click your link. If you’re having issues with conversion rate, this may well be the issue.
Tips For Better A/B Testing on Google Search Ads
Here are our guidelines to achieve the best results when you conduct your A/B testing on Google Search ads.
Test the two versions of your ad at the same time so that you can be certain of the cause of any variations in the results.
Test one thing at a time so you can clearly identify the item that is driving the changed data
Start your tests early in the campaign and run those tests frequently.
Focus on the data, not on your personal feelings about the ad.
Run the tests long enough to drive meaningful results. A few thousand impressions is required to get accurate data. Remember, you are only paying for click throughs, not page impressions (ad displays).
Summary
Testing your ad campaign at the start is critical to maximising its effectiveness. It’s highly unlikely that the first version of an ad will be the best one. By testing slightly different versions of your ad you can determine which combinations best drive click throughs and conversions.
A well designed A/B testing campaign takes some time to set up, manage and review but it’s vital if you want to maximise your ad spend. Not all businesses have the time and resources to manage all the moving parts of a Google Search Ad campaign. However, they can be very effective when it comes to driving awareness, generating new leads and increasing revenue.
Kennedy Wood Marketing can help you design, develop and implement a campaign for your business. This includes the design of a thorough A/B test plan to fine tune your ads and drive your conversion rates.
If you are interested in running a Google Search Ad campaign for your business but you're not sure how to get started, please reach out to us here.