Bidding Experiments helps you to find top-performing bidding strategies across your account, which can be set up in a matter of clicks. Choosing the right bidding strategy can be crucial if you want to improve your account's performance.
It can be confusing to know which bidding strategy you should pick for the best results in any given scenario, and setting up campaign experiments in Google Ads natively can be tricky — our Bidding Experiments tool simplifies the process.
Many Google Ad users don’t have campaigns that have enough data volume to test properly. To fix that, Opteo allows you to run experiments that contain multiple campaigns, so you can aggregate the data and reach a statistically significant result.
You’ll find Bidding Experiments by choosing an account and heading to Toolkit at the top of the page:
How to create a new Bidding Experiment
You can either select 'Create Recommended Experiment' to create an experiment from an Opteo-suggested bidding strategy or scroll down to find a list of your campaigns alongside their Cost, Conversions, CPA, Conversion Value, ROAS and current bidding strategy. Here you can manually select which campaign(s) you would like to use as the base in your experiment.
In the next section, you can give your experiment a name and description.
Next, you'll see the campaigns included in this experiment and a dropdown option for the bidding strategy you want to experiment with. You can set a max CPC bid for any of the bidding strategies you choose to test in one click. Setting a max CPC bid will give you more control over Google's algorithm, so you can avoid sudden spikes in CPC.
You'll then decide the experiment's duration and the % share of the budget that you want to allocate to your experiment campaign. For example, if you’ve got a campaign that spends $100k a month and is responsible for a majority of your conversions, allocating 10% of the budget might be more appropriate than 50% of the budget. Allocating 50% of the budget to an experiment campaign would pose more risk for high spend, or high-converting campaigns.
Monitoring your Bidding Experiments in Opteo
After you have created your Experiment campaign, you can monitor the progress in real-time in your ‘Active Bidding Experiments’ tab. You can see per-campaign breakdowns, overall experiment group statistics, performance graphs and experiment parameters.
You can choose to end an Experiment early by selecting the campaign(s) you want to keep running, and clicking Apply Selected Experiments. Alternatively, you can stop everything entirely and press 'Cancel Experiment' at the top right.
Once an Experiment has been completed, Opteo will send you an alert and you can view the results.
Applying the Results of a Bidding Experiment in Opteo
At the end of an Experiment, to help guide your decision, Opteo gives you a percentage score that calculates whether you have enough data to make a statistically confident decision. This score takes cost, conversions, conversions value, CPA, search impression share and ROAS into account.
By default, the better-performing campaigns will be selected. However, you can override this and manually select the campaign(s) you want to run.
Once you are happy with your selection, click the green Complete Experiment to then apply the preferred bidding strategy to your campaign.
Technical notes:
In order for Opteo to generate recommended experiments or for you to manually create them, your campaigns must:
Be serving ads
Be enabled
Not be an experiments campaign(s)
Not use shared budgets
Not be using campaign-level conversion goals
Be either Search or Display
Be in the same advertising channel if multiple campaigns are being selected for an experiment (search for search, display for display)
Not be part of an existing experiment in Opteo
Not be in 'learning mode'