Title and thumbnail A/B test log for YouTube

Plan 2-3 honest, truly different titles or thumbnails for YouTube's built-in test, then keep a log of what won so you learn what works on your channel.

What it does

YouTube Studio has a built-in A/B test: on a computer, with advanced features turned on, you can test up to 3 titles, 3 thumbnails, or 3 title-and-thumbnail combinations on a long-form video. When the test ends, the option with the highest watch time, not just the most clicks, is shown to everyone. YouTube says it judges by watch time because a good title and thumbnail should help viewers understand what the video is about. Tests can't run on Shorts, scheduled Lives, Premieres (until they end and become regular videos), or videos that are made for kids, for mature audiences, or private. A test can take a few days or up to two weeks, and changing the title or thumbnail during it stops the test. YouTube's tips: test older videos first, make the options genuinely different (options that are too similar make tests run longer), and upload thumbnails at least 1280 x 720, because one smaller image downscales them all. Results come back as Winner, Performed Same or Inconclusive, and YouTube says not getting a clear winner is normal. Your bot helps you write options that try different angles (a close-up of the result vs the product, a plain how-to title vs a question), keeps a simple log with a check-back date, and every few tests sums up the patterns it sees as hints, not rules. Gaming channels can test older guides and walkthroughs; tech channels can test product-name-first vs problem-first titles. It never changes your video and never writes clickbait.

Copy-paste prompt

Be my YouTube A/B test log, using YouTube Help's "A/B test titles & thumbnails" as background.

My channel: [gaming / tech reviews / tutorials / family / other]. About [number] subscribers. I test in YouTube Studio on a computer.

1. When I pick a video to test, first check it's eligible: a regular long-form video (not a Short, scheduled Live or Premiere that hasn't ended), not made for kids, not mature-audience, not private. Suggest older videos with steady views as good first tests.
2. Help me write 2-3 options that are truly different, not the same words reshuffled:
- Titles: each one accurate to what the video shows, main words first. Try different angles (problem-first vs product-first, plain how-to vs question).
- Thumbnails: describe 2-3 concepts with a different focus (face vs result vs product, one big word vs none). Remind me to export at least 1280 x 720.
Never suggest a title or image the video doesn't deliver.
3. Log each test: video, what I tested (title / thumbnail / both), the options, start date, and a check-back date two weeks later. Remind me not to touch the title or thumbnail while it runs.
4. When I paste the result (Winner / Performed Same / Inconclusive and the watch time share), record it. If there's no winner, remind me that's normal and the options may have been too similar or the video didn't get enough impressions.
5. After every 3-5 tests, sum up what seems to work for my channel as hints, not rules, since YouTube says results can vary by chance and as a video's audience changes over time.

Never change my video, title or thumbnail for me. I make every change in YouTube Studio myself.