As some of you probably know, I run a YouTube channel for three years now. As a huge introvert who’s also afraid of the camera, I managed to get it to 130K subscribers in that time and millions of views.

The reality of YouTube

Running a YouTube channel is extreme multi-tasking and a lot more than just making videos. Thumbnails are an important part of that and recently they’re an essential part of ANY marketing strategy — even outside of YouTube.

If you want your posts, materials or even stakeholder presentations to convert you need a planned and well executed visual strategy.

I learned a lot in the last year alone, but even with all that knowledge, there are sometimes thumbnails that just don’t work or perform poorly. In that case, it’s necessary to redesign it with different principles in mind.

Don’t learn Design in 2024

Each year I make a series of two videos talking about why NOT to and then why you should learn design that year. The videos cover the same points but from completely different perspectives.
They usually resonate well and spark interesting discussions.

This video was no exception getting over 1000 likes and good engagement. However, I noticed that the thumbnail wasn’t converting that well.

Thumbnail 1

I started with this thumbnail. The idea was simple — use big typography — show UX and 2024 clearly with UX being crossed out in red. That instantly catches attention.
Then it shows me doing a thumbs-down gesture with a background of a shower of sticky notes.

That’s a UX cliche, but it quickly gets associated with the brand.

Some patterns may feel cliché but the stereotypes exist for a reason — we process them extremely fast.

The result? Here it is after a week.

Strategy: Optimise

When there’s no change after a few days I usually redesign the thumbnail.
The thought process is simple — depending on how the first image was designed I think what could’ve led to it not converting higher.

In this case it was quite easy to figure out — while thumbnails like that have worked in the past I realised that it’s a bit too complex visually. In a small size, it looks messy and difficult to focus on.

When testing the thumbnail with an AI eye-tracking simulation it showed a split focus between my face and the white and red contrast of the letters.

In the small-sized one, the visually dense background led to lower clarity. It made the thumbnail look pretty chaotic and the UX cliche of sticky notes wasn’t even that visible. It looked more like some kind of pattern.

Flip it around

Knowing that the CTR is at 5% and outlining the main issues led me to create a quick list of actions to redesign it.
I always start with a list like that before doing any design, as it helps guide the work and everything happens a lot faster.

• It’s too complex — needs to be more minimalistic

• Let’s go with a clean and clear background

• No need to show my face — it takes focus away

• Bigger sticky notes but fewer of them

• Quick and high-contrast message

• Some visual representation of a down-trend

  

Redesign

The starting point was a single-color background. I went with a pretty vibrant red. No gradients, no elements — just a single color.
Then using a bold font I wrote “Don’t do it!”

Initially, the text was taking up 80% of the horizontal space, but I realised that with that much whitespace, the text can actually be smaller — it still is very easy to focus on even in small sizes.

The next part was the association game. We knew that postit notes are associated with UX so I decided to just create two notes as vector images. Then — since a lot of UX is designing apps and websites, I generated a sketch of a mobile phone using Midjourney. After some tweaks and manual edits, I removed what was on the screen, cleaned up some AI added useless elements from the phone, and added a big UX on the screen.

I also added a shadow under the phone to set it up against that simple background and give the design a little depth.

For the cards, I decided to go with one red, but with a pretty strong shadow so it’s still visible on the background, and one orange. That’s a palette that matches well together.
Then I added a down-facing arrow on the red card and a frowning emoji on the orange one. The icons use the exact same color as the Post-its themselves, just darker. That way it fits a lot better than if they were black.

The final part was arranging everything in a way that guides the eye well and well… I posted it.

The results

Just one hour after posting the CTR jumped to 10.2%, which is over 2x of the first thumbnail.
That is a great result, as with YouTube competition getting above 10% is usually reserved for real-time events that everybody wants to know about.

It works exactly the same for your LinkedIn posts, Instagram posts, and all materials that you want more eyes on.

Be ready for change

The issue is that last year the thumbnail style of the first one actually worked very well. That also doesn’t mean that a new, minimal approach will work perfectly next year.
But knowing that an image doesn’t perform as planned should always lead to an analysis and a correction. In my case, it starts with that bullet list and an eye-tracking simulation.

Based on that it’s a lot easier to plan and execute a successful redesign.

Check out my new Instagram account where I share my marketing and sales tips that have led me to selling over $1M of digital products.

Liked the article? Share it!