Test & compare thumbnails

You can compare the performance of up to three video thumbnails to gain insights that help you with your content strategy. Thumbnails must follow our Community Guidelines or your access to the feature may be removed.

NEW: Test & Compare Thumbnails

Eligibility requirements

Test and compare your thumbnails

  1. Sign in to YouTube Studio.
  2. Select the video you want to test. 
    • Test a new video: From the upper right corner, select CREATE  and then Upload videos. Then upload the video. 
    • Test an existing video: From the left menu, select Content. Click the video you’d like to edit.
  3. Under “Thumbnail,” click Test & compare.
  4. Upload up to 3 thumbnails to test.
  5. Click DONE. The test will start once you publish your video.

Note: You may stop the test and manually select one of the thumbnails at any point by using "Stop & Set." 

Test results 

The winning results may take a few days or up to 2 weeks to be finalized. When the results are ready, they will be available in YouTube Studio. You may then choose to use the winning thumbnail or rerun a new test. 

The following factors will affect the time it takes to receive the results: 

  • Thumbnail diversity: The more they differ, the likelier your test will finish faster.
  • Impressions: The more thumbnail impressions that your video receives, the likelier your test will finish faster.

Frequently asked questions

How does it work?

At the conclusion of the experiment, your thumbnail will have one of following results based on watch time share: 

  • Winner: This thumbnail clearly outperformed the other thumbnails based on watch time share, and we’re sure that these results are statistically significant based on data from viewers.
  • Preferred: This thumbnail likely outperformed other thumbnails based on watch time share. The performance improvement shown by the preferred thumbnail was not enough for us to confidently declare it a winner. The preferred thumbnail may be more engaging to viewers, but this cannot be determined with a high level of certainty.
  • None: All the thumbnails performed similarly, which resulted in no strong statistical difference in engagement. In this case, the first thumbnail you upload will be the default thumbnail for your video. Alternatively, you can always manually change to the thumbnail of your choice.

Control group

We may maintain a small percentage of traffic as a control group of viewers that are excluded from the experiment. The control group will only see the default thumbnail. The video performance from the control group is excluded from the experiment calculations.

Why am I not receiving a “Winner” test result?
It's normal not to see a “Winner” test result. When you are running an experiment, it’s common to receive other test results like “Preferred” or “None.” To review your test report, use the Reach tab in YouTube Analytics.
There may be a few reasons why your video didn’t have a clear “Winner”:
  • Minimal difference in thumbnails: It may be that the difference in the thumbnails chosen for the experiment did not have a measurable impact on video performance.
  • Not enough impressions: It may be that your videos are not generating a sufficient amount of impressions. If your video receives a higher number of views, then the more likely a “Winner” will be declared.

Why is the watch time share used to determine the winning thumbnail?

Great thumbnails serve an important purpose beyond getting viewers to click. They help a viewer understand what the video is about so that they don't waste their time clicking on the wrong videos.

Why are there conflicting thumbnail test results?

There may be cases where multiple thumbnail tests will lead to different results: 

  • Different results using the same video: Thumbnail test results for a given video may vary due to the statistical variation that exists in any real-world experiment, similar to flipping a coin that results in different outcomes each time due to chance. Test results may also vary due to natural changes in a video’s audience composition over time. For example, early impressions are more likely to come from viewers who are already familiar with your channel, while later impressions are more likely to include viewers who haven't yet watched your channel.
  • YouTube’s thumbnail test results compared to third-party thumbnail test results: When using tools not native to YouTube, you may receive conflicting thumbnail test results. YouTube’s thumbnail experiments are true, concurrent A/B/C thumbnail tests. This means all thumbnail variations are shown to viewers at the same time. Many third-party tools run thumbnail tests sequentially, and thus may generate different results. These tools often optimize solely for click-through rate, which may determine a different winning thumbnail than measuring by watch time share. This may be due to videos with high overall engagement (strong click-through rate, watch time share, and other factors) are shown more frequently. We believe that evaluating winning thumbnails by watch time share will best support creators’ growth.

What is the recommended thumbnail resolution for this feature?

We recommend using the highest resolution for all thumbnails. If the resolution of any thumbnail is lower than 720p (1280 x 720), all experiment thumbnails will be downscaled to 480p (854 x 480). Learn more about custom thumbnail best practices.

How can I provide feedback?

The data from Test & Compare will be used to evaluate your thumbnail performance. Our team is focused on improving our systems to make them accurate and useful for our creators. You can provide feedback about your experience using the feature within the test report. 

Was this helpful?

How can we improve it?
Clear search
Close search
Google apps
Main menu