The YouTube Ads course
Module 9 - Creating versions for A/B testing
To A/B test a YouTube ad properly, keep one script and change only the opening six second hook, so you produce a version A and a version B that differ at the front end and let you see which creative actually works.
Last updated 31 August 2026
Running time 8:07Watch it
What does this video cover?
- 01You do not want to change everything. All you want to test is the opening creative, the six second hook, so to speak.
- 02There are no rules around this. It is just something we have done over many years, and there are lots of different ways to skin a cat.
- 03Take the script you already have, with the hook, the educate section and the reminder of why they should buy, and write two openings against it.
- 04Make version A something pretty standard and easy to do, and make version B a little bit more quirky, because that can be more interesting.
- 05The example used is a pair of walking boots: version A is a realistic scene, version B is more caricature and comedic.
- 06Which one suits you depends on the tone of voice of your business and your brand.
- 07Once you have two openings, you top and tail the script slightly so each opening leads naturally into the same body, and that is the only change you need.
Full transcript
What are you actually testing in an A/B test?
The opening creative, the six second hook, and not much else. In this video I am going to talk about A/B testing, but there is a little bit more to it than that, because what I am also going to do is demonstrate how you can create a really good hook with a couple of options, and how you then use that script to get two versions.
You do not want to change everything. All you want to do is test the opening. I am going to show you how we do that, and how we top and tail the creative. There are no rules around this. This is just something that we have done over many, many years. There are lots of different ways to skin a cat, but this will give you an idea of how you can do it.
How do you set the two versions up on the page?
You take the script you already have and write version A and version B openings against it. Going back into the script where we were, we have got some of the ideas around do you need a new pair of walking boots, do you need them to be sturdy, and so on. We have got the educate section, where we taught them about the new things on the boots, and then we had a reminder of why they should have them.
I am not going to do the calls to action. I am literally just going to do the hooks. That script is essentially fine, but what we are going to do is say version A and version B, and you are going to have two openings. This is where we start to tweak the scripts a little bit. We make life easy. We do not want to be changing everything. We want to be using what we have got to the best of our abilities, so let us get rid of the other stuff we do not need.
What should version A look like?
Something pretty standard and easy to do. What I tend to say to people is do something that is pretty standard and straightforward for version A, and then do something a little bit more quirky for version B, because I think that can be a bit more interesting.
So version A. Let us take the whole premise that somebody needs a new pair of walking boots. Why do they need a new pair of walking boots? Because they will be letting the water in, or they are uncomfortable. Let us call version A the sensible one. It is still a nice little idea, but it is more straightforward.
We see a man sat in the woods, sat down on a log, boots off, massaging his feet, and the voiceover goes, is this you every time you go for a walk? Nice little bit of footage, nice little voiceover, and then you get into it: yeah, do you need a new pair of walking boots? We are assuming they need a new pair of walking boots because they are uncomfortable.
What should version B look like?
Something a bit more creative, with more fun in it. Version A is realistic. That is what really happens to people. But let us have a little bit more fun with version B.
We see a man stood at his front door, wet through, holding his sorry looking boots, and the camera sees them with water flowing out of them and generally looking like a waterfall. He pulls a sad face, and the voiceover asks something a little bit lightweight and fun. He is stood there like this, going, oh, it was leaking out, leaking out. So it could say, are your boots like this after every walk?
It is a bit silly, it is a bit extreme, but you can see there is a difference. There is a definite A/B test. The A is a real life situation that happens often: taking the boots off, massaging your foot, it is uncomfortable, yeah, do you need a new pair of walking boots, yes I do. Version B, which is a bit more creative, is more caricature and comedic. Are your boots like this after every walk? Well, of course they are not, but we are exaggerating and having a little bit of fun.
It depends on the tone of voice of your business and your brand. I am just giving these as little examples, but it is a nice couple of little tests.
How do you make the rest of the script fit both hooks?
You move one small piece of the script so each opening leads into the same body. What I would say is that you just need to tweak your version A around a bit. For version B, I would take that little bit there and move it, so the nod to the waterproof point works with version B. With version A, I would take that out and put it at the end. That is the only change you need.
You can see there that version A is like a little segue. It just needs a new pair of walking boots because, well, actually what you would do is tweak this around. You would probably still say, do you need a new pair of walking boots, but you would put it in a different place. So you are just altering the structure a little bit to suit the idea that you have put in. Not difficult, simple, and the rest of the script is the same.
What you are doing is testing two front ends, which is really important, because you can start to see which creative works. As the rest of the team go through how you build campaigns and how you test, learn and reiterate, you will start to see that you might tweak it out even more as you go forward and have other versions of all of that. But this is a really good example of how you actually put your A/B testing together.
Where next?
The library holds every video we have published. The AEO pages explain the method behind them, what it costs and what the evidence actually supports.