Increase Performance with A / B Tests in Google Ads

In order to advertise successfully in the long term, advertisers often ask themselves which levers need to be pulled to increase performance. Since there isn’t always the obvious right or wrong path, campaign testing is an excellent option that every advertiser should know about. Many factors affect the profitability of a Google Ads campaign. The choice of bid strategy or landing page can determine the success or failure of a campaign. Google Ads offers advertisers a test environment in which possible scenarios can be tested for a period of time and, if the results are satisfactory, they can be transferred to the original campaign. The scalability of the results obtained is important.

What are A / B tests?

With A / B tests or split tests, advertisers can test two variants of a campaign in parallel without having to make changes to the original campaign. Testing changes one after the other instead of in parallel is often not useful due to the lack of comparability given different starting conditions such as fluctuations in seasonality, competition or search volume.

The following main criteria can be tested:

  1. Landing Pages
  2. Bid strategies
  3. Ad texts

In a test, variant A corresponds to the original campaign and variant B to the campaign with the criterion to be tested. In order to achieve relevant results, both campaigns should each receive 50% of the traffic . 

Implementation of a campaign test

The first step in setting up an A / B test in Google Ads is to create a draft that can either be applied to the original campaign or activated as a test. When creating a draft, a copy of the original campaign is created, a mirror of all content such as ad groups, ads, keywords, bid strategies. Advertisers can find this option under the menu item “Drafts and Tests” in the Google Ads interface.

Drafts are only available for Search and Display campaigns. A setup of drafts has not yet been approved for video or shopping campaigns. When a test is run as a draft, it tests the changes made to the new campaign and compares its performance to the original campaign over time without affecting performance. 

When a draft is ready, advertisers have the option to create a test instead of applying the changes directly to the original campaign.

The user has two options for dividing the test:

  1. Split on the basis of cookies : With cookie-based test groups, users usually only see one version of the campaign, regardless of the frequency of the search query. The data obtained are as accurate as possible, as no other factors falsify the results.
  2. Distribution based on search queries : In the case of test groups based on search queries, either the ads from the trial version or the original campaign are shown for each search. This is done at random and enables both versions of the ad to be played to the same user if the same search query is made multiple times.

A test can be ended at any time. The various key figures are evaluated in the campaign overview. If the results achieved caused an increase in performance, the transfer to the original campaign follows.

What options do A / B tests offer?

A / B testing can be used to test a variety of adjustments. The main ones that every advertiser should know are:

1. Campaign test landing page

If advertisers have two potential landing pages for a product or service, A / B testing of those two landing page URLs can help determine which one is producing the desired results. For example, a product has an order page in the shop and a login page for free trial access. The user may be directly interested in purchasing the product. Alternatively, advertisers want the user to reach the login page for the test access and use it to convert after the test period has expired. An A / B test can show which of these landing pages encourage users to convert and increase the profitability of the ads. Small changes on the target page, such as the positioning of an interferer, often lead to changes in performance,

2. Campaign test bid strategies

A split test can be used to check the extent to which a different bid strategy affects the campaign performance. Advertisers adjust the bid strategy in the test campaign settings. All automatic and manual bid strategies are available for the test environment . For example, an advertiser would like to check whether lowering the bids still leads to sufficient traffic or whether a shop geared to generating sales with the strategy “target ROAS” or “maximize conversion value” achieves a significant increase in conversion values.

3. Campaign test advertising texts

Ad texts are crucial as the first point of contact with the user. The more appealing and appropriate the ad appears to be for the target group, the more clicks and conversions advertisers get. This shows the importance of doing continuous testing. For example, the following can be tested:

  • Address: address of the user
  • Wording: inexpensive vs. Cheap
  • Call-to-action: buy now vs. Order here

By determining the best performing ad content, click and conversion rates can be increased.

You might be interested in the following topics:

The Relaunch Channel Check: What Can Go Wrong with SEO, SEA & Co.?

SEA beyond the Basics (part 1): Score Points with Weather Displays in Sun and Rain

What is SEO: The Basics Guide to Search Engine Optimization [2021]


With Google Ads, A / B tests or split tests make it easy to determine how users react to changes without applying them to all traffic and incurring high costs . It is essential for the tests that advertisers only adjust one criterion (no combination of a new landing page, ad texts or bid strategies) in order to determine which change leads to a positive or negative development in performance. A / B tests are helpful in order to better understand users, generate reliable data and increase performance in the long term .

Leave a Reply