Need Advice YouTube Experiments - How To Experiment, Analyze, & Adjust?

[NanoniumTheDude]

It's always great to experiment with new things whether that is changing up thumbnail style, changing the editing, or even trying YT shorts

MY QUESTION - How do you know how long to experiment before you analyze and adjust? For reference, I upload biweekly but I am experimenting with thumbnails, shorts, and editing styles all at once and I'm wondering at what point do I pause and reflect on the effectiveness of all this before deciding to implement a "permanent" change?

 

[SILTHW]

Personally, I am focused on Average View Duration and tend to immediately double-down on a style, length and presentation structure that drove high AVD. I should do that more with CTR - by that I mean looking at why certain thumbnails do better, but I feel like my focus needs to be AVD so I make those adjustments constantly.

 

[Ragulsundaram]

Use the Tubebuddy tool wisely. Use it's SEO studio wisely and the keyword explorer is the best feature in the tool. Search for trending topics and make your videos based on them

 

[NanoniumTheDude]

Personally, I am focused on Average View Duration and tend to immediately double-down on a style, length and presentation structure that drove high AVD. I should do that more with CTR - by that I mean looking at why certain thumbnails do better, but I feel like my focus needs to be AVD so I make those adjustments constantly.

So essentially you keep doing what gets results regardless of how long you're experimenting with new styles/thumbnails/videos?

 

[Ragulsundaram]

So essentially you keep doing what gets results regardless of how long you're experimenting with new styles/thumbnails/videos?

That's the intelligent way to go. You have to give what people love. You cannot force them what you want them to have

 

[Beanie Draws]

Generally I go with the CTR% if I'm experimenting with new thumbnail approaches. And obviously click through rate, overall views, and average view duration will tell me if a content change experiment is working on or not. If the views are growing as with the other stats, I know to continue, if the data is the same, then that tells me it's not necessarily a bad idea, but not something to totally pivot towards.

I'm doing a few experiments with shorts now. I always like to experiment with new ideas. Even if the experiments aren't successful, it keeps the art of creation fresh and exciting.

 

[Stanley | Team TB]

As a manager for various businesses I was taught early on to trust my staff. It is the well-oiled machine that drives the furthest, and I would advise you to do the same. If you use TubeBuddy A/B testing let it do its job; put the blinders on, ignore your personal feelings and run with the best results

If you do not don't worry, the same principle holds true. Fix the worst performing thumbnails/videos and then move on. If you are perpetually fixing the worst end of your library the overall performance of the channel will always be improving. You likely won't stop doing that until you hit a point where subscribers make up enough of your views that you no longer worry about search ranks.