A/B Testing Thumbnail Masterclass: How To Use YouTube's A/B Tool Correctly!

Here is A Step-By-Step Guide to YouTube’s Compare & Test Tool To Help You Get More Views.

‍

This is a pretty long email, but it’s one I think you’ll be revisiting over and over again, so be sure to save it!

Over the past year as a YouTube Strategist, I’ve had access to YouTube’s new A/B testing feature.

I’ve used it on a ton of creator’s channels to fine-tune their thumbnails and there are a few crucial things I’ve learned that you absolutely need to know to get more views.

The tips I’m going to share will transform how you use this A/B testing tool as I’ve seen people screw this up and lose out on views through silly mistakes, and I don’t want you to be that person.

‍

I’m going to break this newsletter down into three parts:

1. Understanding The Basics of A/B Testing
2. My Top 3 Tips for Thumbnail Testing that Consistently Get Results
3. Common Mistakes and How to Avoid Them

‍

Part 1: Understanding the Basics

‍

First, let's talk about the basics.

‍

If you haven’t used a thumbnail A/B testing tool before, it's a game-changer for finding the best thumbnail that gets your viewers clicking.

‍

When we A/B test, we’re basically showing two/three different thumbnails to our audience to see which variation is the best at getting people to click.

‍

‍

Some tools have been around for years that have allowed us to do this, but they come at a monthly cost.

The good news is that YouTube has built its free version of these tools, which everyone can access.

If you don’t know where to find this tool, you’ll find it when uploading a new video or in the details page of videos you’ve already uploaded.

All you need to do is scroll down to the thumbnail section, click on test and compare.

‍

‍

This is where you can upload three different thumbnails to be tested.

‍

‍

Once you click done, YouTube will show these three thumbnails to your audience simultaneously and begin showing you the watch-time share of each thumbnail.

If you don’t know what watch-time share means, it just means out of the people who clicked, which thumbnail led to those viewers watching the longest.

‍

‍

Once complete, YouTube will automatically set the winning thumbnail as the default, which everyone will see.

‍

Part 2: Top Three Tips for Thumbnail A/B Testing

‍

Cool, now we’ve got that out of the way, let’s talk about my top three tips…

‍

Tip 1: Screenshot Your Results

‍

This is important because currently, YouTube only shows the latest test results; the second you create a new test, those results are lost forever, so keeping a record of what thumbnails won and why is essential.

‍

The way that I’ve been doing this with my clients is by using this Notion template, which I created that lets you log these screenshots so you can see what thumbnails won and why they won 👇

‍

The Brand Familiarity Notion Template

Enter your email to receive The Brand Familiarity Notion template to change how you view and design your thumbnails (entirely for free) 👇

jamiewhiffen.co.uk/brandfamiliarity

‍

‍

There’s also an area where we have a list of the patterns and trends that we notice results in more clicks with our viewers.

‍

This means that every time we’re in the thumbnail creation process, we have a list of things that work and a list of things we should avoid putting in the thumbnail as we know they’ve not performed well in the past.

‍

If you want to download this FREE template to use, click here.

‍

Tip 2: Test Big Differences First, Refinements Second

If you watch one of the many excellent interviews between Mr. Beast and Jon Youshaei, you’ll see that Jimmy A/B tests by starting the test with significantly different thumbnails to identify the best concept:

‍

‍

And then once he has a winner, only then does he run additional tests with minor variations to refine it:

‍

‍

You don’t want to start off testing the difference between a yellow arrow, a red arrow or a blue arrow as the difference between these things will be really small compared to different thumbnail concepts entirely.

‍

My advice to you would be to really push yourself to get into the habit of creating three different concepts per video and I KNOW that can be super annoying and take up a lot of time, but honestly, it makes such a big difference.

‍

Even if you can only get two thumbnail concepts per video that’s way better than just sticking with one concept that you think is really strong, but actually, it’s kinda poor.

‍

If you don’t have another concept to compare it against, then there’s no way to know if the thumbnail you think is great is ACTUALLY great.

‍

For example, in this video I worked on with GrowVeg, we only had two concepts, and you can see that one of these concepts was a before and after, which we thought would do well.

‍

But, it actually underperformed compared to Ben holding the large ginger up with the promise of growing it at home, and this resulted in the video getting over 1.4M views!

‍

‍

Take this video too, we had completed a first round of thumbnail tests and found that the winning concept was Ben holding a bucket of mushrooms, so we then created three variations:

‍

‍

One of these was the clear winner, and this helped the video get 900k+ views, far above what they typically see on the channel.

‍

‍

So please, make three concepts per video as I’ve seen really surprising results taking this approach, and whilst it can take more time to create completely different thumbnails, it can result in a far higher watch timeshare!

‍

Tip 3: Repeat Tests Periodically

‍

Now, a common mistake that my clients tend to make, which I keep having to remind them not to do, is to stop taking the first result as gospel.

‍

From past experience, we’ve noticed that even if thumbnail A wins in the first test, this result can change to thumbnail B after a few days. For this reason, we want to run tests multiple times.

‍

It's important to remember that in the first 24 hours to a week, data can be skewed by returning viewers.

‍

These viewers are familiar with your content, thumbnail style, the channel name and are more likely to click, watch for longer, and boost watch-time share, which is what YouTube is measuring to select the best thumbnail.

‍

This means that it's essential to retest periodically to adapt to changing viewer preferences and trends. What works in the first week might underperform a month later when shown to new viewers, which is a type of viewer we need to attract if we want to grow our channel.

‍

Also, remember that some videos have thumbnails which perform well on the homepage but poorly in search results, and vice versa.

‍

When you first upload a video and test two different thumbnails straight away, YouTube will show your video to subscribers on the homepage. This might make it look like one thumbnail is better than the other, even if it's not.

‍

To get a more accurate result, you should test your thumbnails multiple times once the traffic volume and sources have stabilised, this will give you a more accurate thumbnail winner.

‍

So, please please please, retest your thumbnails multiple times over weeks, months, or even years later.

‍

I once changed the thumbnail of Ali Abdaal's popular investment video, which already had 2 million views, to make it more eye-catching.

‍

‍

‍

This small change kickstarted the algorithm to show the video again and boosted the video's views from 2 million to 6 million!

‍

‍

So, updating thumbnails, even after a long time, can help your videos get more views.

‍

Part 3 - Common Mistakes

Now let’s wrap up with some rapid-fire common mistakes and how to fix them:

‍

1. Overlooking External Factors

If you have a video that will be watched more in the summer months, like how to grill the perfect steak on a BBQ, don’t test this in the Christmas period when views and watch time could be down across the board.This could mean that YouTube selects what it thinks is the best thumbnail when it could actually be the weaker thumbnail. So consider trending topics, seasonal changes, or events that influence test results.

2. Ignoring Viewer Feedback

Pay close attention to your comments and likes/dislike ratio, as it can reveal if your winning thumbnail is attracting the wrong type of viewer or if your fans feel you're using clickbait tactics, which can damage your relationship with your audience and lead to even fewer clicks in the future. Graham Stephan is a classic example of a creator who uses these negative emotion tactics in his thumbnails constantly, but this leads to a reputation that he can’t 100% be trusted as he’s consistently stating that the markets are going to crash, which eventually waters down his messaging and trust with his audience.

3. Not Verifying Results

As this is still a beta tool, it makes mistakes. A few times, I’ve seen that the tool has selected a thumbnail as the winner despite having a lower watchtime share, so always double-check your results to ensure accuracy and account for potential errors in the testing phase.

‍

I hope you found this newsletter useful!

‍

If you want some more tips from me on how to improve your thumbnails, check out this video on my YouTube Channel:

‍

Have a great weekend! 👋

‍

My Favourite Things This Week

‍

‍

Guide I’m Reading…

‍

YouTube Thumbnail Design – The Ultimate Guide - Great article from Thomas Frank who breaks down his process for creating thumbnails. Highly recommend reading this!

‍

Book a Free 15-Min Call With Me.

‍

This is a quick note to let you know that if you find this newsletter valuable, you can book a free 15-minute call with me here.

‍

I’ll review your channel and give specific advice on how you can generate more views by creating better video ideas with stronger titles and thumbnails.

‍

‍

Video I’m Watching…

‍

10 Mins of Internet Growth and Marketing Masterclass by Mr Beast: - This is a short complilation video of Mr. Beast rattling off advice for those of us who want to grow an audience on the internet, specically YouTube.

‍

‍

Video I’m Watching…

‍

Warren Buffett’s Writing Secrets - Fantastic video from David Perell which I was really blown away by.

‍

If you’re someone who writes newsletters, tweets, YouTube scripts, this is a video you’ll want to check out for the PoP framework that Warren Buffet uses to write engaging words.

‍

Thanks for reading, if you want to reach out to me, just send a tweet to @WhiffenYT.

‍

Jamie