A thumbnail that looks great in your editor can fall apart when it is shown at the size of a stamp, next to other videos. The Thumbnail Preview and A/B Compare tool helps you catch that before you publish. It also helps to know what it cannot do, and where to go for a real test.
How to compare two YouTube thumbnails with the preview tool
- Upload thumbnail A and thumbnail B. They stay in your browser.
- Type a title for each, and set your channel name.
- Switch between Dark feed and Light feed, because many viewers use one or the other.
- Look at both on a phone-sized screen, and at arm's length on a desktop. Can you read the words? Is the face or object clear?
- Pick the one that is clearer and truer to the video, or go back and redesign.
What to look for
- Readable text. Three to five big words at most. If you squint and cannot read it, shorten it.
- One focal point. The eye should go to one place first.
- Contrast against both feeds. A dark thumbnail with a dark border can vanish on the dark feed. A pale one can vanish on the light feed.
- A title and thumbnail that work together. The thumbnail should add something to the title, not repeat it word for word.
- An honest promise. YouTube's click-through rate FAQ says clickbait videos tend to have low average view duration, so they are less likely to be recommended.
What this tool is not
It is a mock-up. It does not use YouTube's data, and it cannot tell you which thumbnail gets more clicks. A nicer-looking preview does not mean a better click-through rate. The tool helps you avoid obvious problems, such as unreadable text or a thumbnail that disappears in dark mode.
YouTube's own Test and Compare
For real data, YouTube has a feature called A/B testing in Studio. Here is what YouTube's help page says today:
- You can test up to 3 titles and thumbnails, as title only, thumbnail only, or both.
- At the end, the option with the highest watch time share is shown to all viewers. If there is no clear winner, the first option you uploaded is used.
- It is currently available on computers, through YouTube Studio, and needs advanced features enabled.
- You cannot test Shorts, scheduled lives or premieres (a premiere becomes eligible once it ends and converts to a regular video). Kids content, mature audiences content and private videos are not eligible.
- A test can take a few days to 2 weeks. If a title or thumbnail is changed during the test, the test stops.
- YouTube recommends testing older videos first and making the options clearly different from each other, because options that are too similar take longer to produce a result.
- If any test thumbnail is below 720p (1280 x 720), all of the test thumbnails are downscaled to 480p (854 x 480), so upload high-resolution images.
Notice that YouTube picks by watch time, not by clicks. A thumbnail that gets clicks and short views is not necessarily the winner.
Use both, in order
- Use the preview tool while you design, to avoid readability problems.
- Publish the best option.
- If you want proof, run a Test and Compare in Studio on an eligible video and let it finish.
YouTube's click-through rate FAQ also warns against testing several thumbnails or titles on the same video by hand, because differences may come from traffic sources rather than the change itself. That is one reason the built-in test is a better choice than swapping thumbnails yourself.
FAQ
Can I A/B test thumbnails on YouTube?
Yes, in YouTube Studio on a computer, if your channel has advanced features and the video is eligible. You can test up to three titles and thumbnails.
Does this tool tell me which thumbnail will get more clicks?
No. It shows how each one looks in a mock feed. Only real viewers can tell you which gets more clicks.
How long does a YouTube thumbnail test take?
YouTube says it can take a few days up to 2 weeks.
What is the winner based on?
YouTube says it is the option with the highest watch time share, not the highest click rate.
Are my images uploaded to a server?
No. They are processed in your browser.