Ad testing metrics
Learn how to use ad testing metrics to make test decisions
Last updated
Learn how to use ad testing metrics to make test decisions
Last updated
Ad testing metrics define how valid test results are produced, together with . Adalysis can test six metrics when generating ad test results, independently of each other.
You can disable metrics that don't apply to your account or campaign. That means they won't show on test results, so you can focus on what's most relevant to you. For example, accounts that don't set a conversion value won't need test results for conversion value/cost and conversion value/impressions.
You can find your metric settings under Ad testing.
When appropriate, we use the same metrics you see in your Google Ads account to make the data easier to review.
When Adalysis finds a statistically significant test result, we'll display the confidence level as either 90%, 95%, or 99%. Confidence values of less than 90% are displayed as --. For example, these test results have statistically significant CTR results with a confidence of 99%.
CTR
Click-through rate
The ratio of clicks and impressions
Conv. rate
Conversion rate
How often a website visitor completes a desired action, like making a purchase, out of the total number of visitors
Conv. / impr.
Conversions / impressions
The number of ad impressions that result in conversions
Cost / conv.
Cost / conversions
The cost per conversion, or CPA (cost per action)
Conv. value / cost
Conversion value / cost
The conversion value over cost, or ROAS (return on ad spend)
Conv. value / impr.
Conversion value / impressions
Conversion value per impression measures the average revenue generated each time your ad is displayed.
Higher confidence means the result's long-term validity is more reliable. However, higher confidence levels generally require more data. Depending on your test strategy and account volume, you'll need to decide what confidence level you want to reach before pausing losing ads. You can also use to control the volume of data needed.