A/B test thumbnails and titles the right way

Last updated 13 Oct 2026

Don't guess which is better — change one thing at a time and let the numbers decide.

Course · Stage 6: Publish, measure, multiply · Lesson 27/40 · 8 min

Why it matters

The same video can get very different click-through rates from different thumbnails; you only find out by testing.

Change several things at once and you can't tell which one moved the number.

Step by step

1. Make 2 versions that differ in one element only: facial expression, background color or the text on the image.
2. Preview at phone size: is the text readable, is the subject clear?
3. Use the platform's built-in A/B test if available; otherwise swap thumbnails after a fixed period and compare click-through.
4. Compare only after enough impressions; too few and you can't conclude.
5. Log the winner and why, and apply it to the next video.

Common mistakes

• Changing image, text and color together.
• Calling a winner after a few dozen impressions.
• Not writing down the lesson.

Ready-made prompt (ChatGPT / Claude / Gemini)

Video: [title + summary]. Current thumbnail: [description]. Suggest 2 thumbnail options that differ in exactly one element (the one most likely to matter), explain the hypothesis, and keep on-image text to 4 words max.

Exercise

Run one A/B test on your next video and log the result.

Done when: Two versions differing in exactly one element, enough impressions before deciding, the winner logged.

Need help? ngocdiep3135@gmail.com