Do you think the results in Google Ads have stagnated? Or isn't it going exactly as you had hoped? The platform is constantly evolving, and what worked last year may not work today. At the same time, Google provides you with ever more automatic tools, and it may be tempting to get back and let the system take control.
But it's often when we test something new, something untrained and something that might not work out, that we learn the most. That's when we find the solutions that actually raise the results.
So experimentation is a natural part of our everyday life when we work with Google Ads. In this article, I share how we work with experiments in practice: what we test, how we document it and how we use the results to improve over time.
Why experiment?
Google Ads is algorithm controlled. To a large extent, it is about giving the right signals to the accounts and campaigns, but how do we know what signals actually work in the accounts, for the campaigns and products in the various markets?
Experimentation gives you:
- More control: You understand better what drives the results
- More security: You know what's worth scaling and what should be cut
- More Speed: You avoid guessing and getting actual answers
Examples of experiments before high season
In advance of Black Friday, we put more emphasis on experimentation. Already in October we started a number of tests across accounts, campaigns and products. The aim was to be prepared when demand and budgets increased and when competition adapted.
Here are some of the experiments we did:
Promotional structure: We adjusted how the campaigns were organised and grouped to ensure visibility across offers and ordinary prices. It also provided better transparency and easier scaling in those areas with the highest potential.
AI MAX: We opened search campaigns with wider matching to capture relevant searches we would otherwise not cover, and let Google match the user with the most relevant landing page.
Asset groups in PMAX: We tested “feed only” against asset groups with their own images and messages to see which combination delivered best.
Feed adjustment: The products were marked to be taken out in their own campaigns according to a new campaign structure.
Audible signals: We compared Performance Max with and without separate target group signals to assess the impact on conversion value and CPA.
New customers: We tested new functionality in Google Ads to only reach and convert new potential customers.
How to document experiments
Without documentation, the list is easily lost. What did we test, why and what was the result?
We document all tests easily, but clearly. For example, this may be a useful template to use:
The important thing is that it's understandable, even three months later, for yourself or a colleague. At the same time, it's a good way to note what experiments we want to test in the future. When your experiments are documented and shared, you also build a culture that rewards curiosity and insight.
Finally
Make testing a habit on Google Ads. Often, small, regular tests provide a better basis for decision-making, safer scaling and more effective campaigns. And one last recommendation: make sure you always have an experiment going.
And write it down, share it with the team and use it to get a little bit better in Google Ads.