Test & Learn: How To Improve Performance With Google Ads Experiment Tools

Ever had a client be disappointed by the results of a test? I definitely have, and this is always my response: there is no such thing as a bad or failed test.

You tested a new CTA in your ad copy? CTR dropped? You now know your original CTA was more effective at getting people to click.

You tested a new landing page with condensed content? CVR dropped? You now know that the additional information on your original page was valuable to your customers and was helping drive conversions.

Every test is extremely valuable because the purpose of a test is not only to show us what works, but also what does not, and this is often overlooked.

A test that produces a negative result is not a failure; it has still given you information you didn’t have before. It has helped you rule something out, challenged an assumption, or given you a clearer understanding of what your customers respond to.

The only real failure is running a test, getting a result, and learning nothing from it.

Hence, in this blog, we are going to look at the different features available within Google Ads that help you test and learn, which feature is best suited to each type of test and why running these tests regularly is crucial.

Custom Experiments For Search & Display

The Custom Experiments tool in Google Ads allows you to test changes to an existing campaign in a controlled way, without having to immediately make any changes to the original / existing campaign.

It can be used to test elements such as Smart Bidding strategies, keyword match types, landing pages, audiences and ad groups across Search & Display campaigns. 

Once you have selected a base campaign, you can create a separate trial campaign with the settings you would like to A/B test.

This allows the performance of the original campaign and the test campaign to be compared before deciding whether the changes should be applied to the base campaign or whether the experiment should continue as an independent campaign.

One of the main benefits of using this feature is that it allows you to test a specific change while maintaining a clear comparison against the existing campaign; you can test the new strategy through an experiment rather than making the change across the entire campaign immediately.

This approach helps you make informed optimisation decisions while reducing the risk of implementing changes that could negatively affect campaign performance.

Ad Variations

The Ad Variations tool is specifically designed for testing changes to ad copy across Search campaigns. The tool allows you to create a modified version of an existing ad plus compare its performance against the original, and then, if the results are positive, apply the changes to the campaign.

This makes the feature particularly useful for testing responsive search ads or one specific copy change across multiple campaigns.

For ad copy testing, the ad variations tool is best suited as it allows you to focus specifically on how a change in messaging/copy affects key performance metrics such as the CTR.

You could test a new headline, description or call to action across multiple campaigns and compare the results against the original ads.

While Custom Experiments can also be used to test copy changes within campaigns, they are broader in scope and are better for testing campaign-level changes such as bidding strategies, audiences or match types.

For a straightforward ad copy test, the ad variations tool would provide you with a more focused and efficient approach.

Another advantage is that the tool makes it easier for you to test one specific change across multiple campaigns.

This is really useful in instances when you might want to understand whether a new messaging approach works consistently across the wider account, but you would like to test it on all your campaigns rather than just within one campaign.

Regularly conducting ad copy testing is crucial because ad messaging can potentially become less effective over time as competitors, customer expectations and search behaviours change.

It helps you find incremental improvements and avoid relying on the same creative indefinitely. Like changing a shop window display, regularly refreshing your ad messaging helps you test what makes people stop, look and take action.

Performance Max Uplift

The Performance Max Uplift experiment feature is designed to help you understand the incremental impact of upgrading to Performance Max.

Rather than simply looking at the conversions attributed to Performance Max, the feature helps establish whether the campaign is generating additional results that would not have happened otherwise.

The experiment does this by comparing a group exposed to Performance Max with a control group that is not. If the Performance Max group generates more conversions or conversion value at a better CPA or ROAS than the control group, the difference provides an indication of incremental uplift.

When the aim is to understand the true incremental value of Performance Max, this tool is much better suited in comparison to Custom Experiments.

Custom Experiments are better for testing specific campaign changes, such as bidding strategies, audiences or match types, whereas Uplift experiments focus on whether adding Performance Max actually drives additional conversions or conversion value beyond existing campaigns.

This distinction is crucial because some conversions attributed to Performance Max may have been captured by other campaigns anyway. Regularly running these tests helps you to understand whether Performance Max is genuinely adding value as competition and customer behaviour change.

Conclusion

Each of the tools explored above is designed for a different type of test; from campaign changes and ad copy to measuring the incremental value of Performance Max.

Regularly running the right experiments allows you to identify what is genuinely driving better results, learn from them and adapt as performance changes.

By making testing a consistent part of campaign management, you can continue to improve performance and, most importantly, make optimisations a little less of a guessing game. 

Test and learn. Remember that means learning from every outcome, not just the wins.

If you require assistance with your Google Ads strategies, we are here to help. Contact us today to get started.

Author - Sana Kaladia

Sana holds a degree in Economics and has two years of experience in PPC. She currently works as a Senior PPC Account Executive.