The point of a second thumbnail is to give yourself a real choice, and a real choice means a genuinely different idea, not the same thumbnail with the background recoloured. Two variations on one concept teach you almost nothing. Two different concepts teach you which idea your audience responds to, which is the only thing worth learning here.
Most creators make thumbnail two by nudging thumbnail one: same shot, bluer background, bigger arrow. That is not a test, it is a tweak. When you only have small variations, you learn which shade performs slightly better, while the bigger question, is this even the right image, goes unanswered.
Change the concept, not the colour
If thumbnail one is a close-up of your face reacting, thumbnail two should be structurally different: the object, the result, the before-and-after, the place. You are testing two ways of selling the same video, so the two should be far enough apart that picking between them actually means something.
| If thumbnail one is... | Make thumbnail two... |
|---|---|
| Your face reacting | The object or result, no face |
| The finished result | The messy starting point |
| A wide establishing shot | A tight, single-detail close-up |
| Text-led | Image-led, almost no text |
The further apart the two ideas, the more you learn from comparing them, and the more likely one of them is genuinely better than the first instinct you started with.
Two is the sweet spot, not ten
You do not need eight thumbnails. Two strong, different concepts is the right amount: enough to give you a real decision, few enough that you actually finish them and film the next video. Thumbnail-making can expand to eat a whole day, and a day on thumbnails for a video that took an afternoon to film is the wrong ratio.
Make two you would genuinely be happy to publish, then move on to deciding between them. The goal is a real choice, made quickly, not a gallery you agonise over.
Where Chewbr fits
Make thumbnail two is step 21 of the 47, the middle of the three thumbnail steps. It exists because a single thumbnail gives you nothing to compare, and comparison is how you avoid publishing your first guess just because it was your only one.
Keep reading
With two real options, shrink-test them at phone size to see which survives. Choosing the winner is covered in why your second-best thumbnail is usually the one to publish, and both were built on the principles in thumbnail one.