no matter how hard it is to prove correlation between the creative metrics and the performance of the Ross it's easy to see that like your best ads have the best creative metrics in most cases great hook rate great hold rate and great average watch time here's how to actually analyze the results of your creatives to actually extract more learnings from it to actually make more winning ads join me in the full creative analysis process so in my weekly creative report recap I want to see what are the learnings from the test that I ran so
I can actually make more iterations so in other words I want to double down on a specific concept angle or Creator that's winning or I want to see which ads work the best so I can repurpose them for Tik Tok tests so how to actually do creative analysis the basis for this is to have a proper naming convention this allows you to compare your performance per angle per Creator per style whether it's wh listing or not whether you use the different video editor in this case you can see part of our naming convention and what
are different tags that we are using to actually evaluating our tests in motion how does this actually look like first report is actually the report of the best ads so you could see different ads and they Ross in this case this is something that's running on retargeting so probably it's not uh really correct but you could see here this one spend the most this one spend uh decently but it also it doesn't have a good Ross but the thing is this is pulling the data from Facebook ads manager while will we are often making decision
based on Triple Veil on norbe but it's easy to spot okay this one has potential it did not spend much so okay let's try to push it or maybe we are overspending on this one recently motion introduced option to pull your nor beam data in their dashboard so that's something that would be pretty useful down the road so that's first part of the weekly report the second part is that we want to compare performance per product you can see here different products and their performance this one is is something that is getting lot of the
sales and amazing Ross but it's not something that necessarily scalable cuz this is potentially running on retargeting this is obviously the product that we are pushing the most and it's most scalable so when you see a report like this you can easily understand okay which products we want to push which we want to pause or which we want to like double down potentially after that we also tracking the performance SP our Creative Source whether it's an influencer whether it's a ugc whether it's an agency in this case you can see that performance is pretty stable
between the influencers and ugc creators also proper naming convention allow us to compare performance per different Assets in this case videos images and GIF again images are mostly doing retargeting so that's why we are not pushing them so much in this case but this gives you like clear representation of which assets potentially make sense to double down this all is possible with a proper naming convention so obviously this is part of the equation what we all want to know is how to iterate on a winner so I'm always making my decisions based on the data
never based on the Assumption so we are making decision based on Thum stop ratio and hold rate thumb stop ratio is something you all know how to calculate 3C video view divided with impression this tells us how good the video is at grabbing the attention after that we have a hold rate a lot of people are calculating hold rate as through play divided with impression but I don't like that because then your good hook could impact hold rate because then hold rate would also measure partially hook because you're dividing with Impressions so that's why I'm
dividing my hook rate with as a through play divided with 3 second video view this allows me to compare performance of the hook and then hold rate is comparing performance after third until the 15 second which is actually when through play this fire I think this is a clearer representation what is happening after the hook cuz often you could have a amazing hook and then your hold rate would good but it's not actually hold rate that's good because it's hook that that has a great percentage that you're including into your equation so how to actually
iterate on a winner uh number one is obviously the easiest thing to do is iterate on a hook have the same creator with the same messaging to film the video in a different setting let's say he filmed the one in the kitchen the other he can film Outdoors if he filmed in a car he can film it in a room or something like that cuz often you would have a messaging or a script that's working the best so it makes sense to try to replicate it you can also do slight iteration of of the same
message by having different variations of that messaging which is trying to communicate the same angle but in a slightly different words because then Facebook counts that as a new version of the video here's the example of what I meant so in this case we find out that like this kind of hook is performing the best we initially said the professor from X University now we put like additional variation Professor from Duke University this is story of how four students from Duke University a retired Professor from Duke University so basically we found winning hook we made
like lot of variations out of this seven probably like you would get one or two that would actually be great performer but it's definitely worthwhile testing that but only when you have a proven winner it doesn't make sense to test different 10 variation initially before you even know that ad is winning one other thing how to iterate when you have a winning video is to have a message overlay number one would be to have a meme style Banner you can see here fine lines and wrinkles putting you down so this is part of the angle
that is mentioned in the video now we put that as a banner you can put it in the first couple of seconds or uh stay or leave it there for the whole video the other thing is to have a fake Tik Tok comment fidget here again putting this there or like communicating the angle through the comment uh what I found is that works the best either communicating the angle or calling out the Target customers in those meme Styles and comment widgets the third would be to have the different creator with the same script as I
said it's usually the script that's winning one so it makes sense for you to get that script to other ugc creators or influencer to try to recreate it and Ford uh which is probably the easiest one to speed up the video by 10% again Facebook count that as a new asset you often get good results just by increasing the spend of the video by 10% or maybe add some additional variation change the b rolls and stuff like that because you have a venix script you want to make more variations out of it that's it guys
this was a short video on how to iterate on your winner I want to emphasize that you start making these iterations only when you have a winner if you want me to make additional videos how I'm actually analyzing this metric in Facebook ads manager or how I'm actually analyzing the creatives when it's a image just comment below if you are spending at least 50k and you want me to implement this process in your ad account so we can actually make more winners out of your existing winners book a call below cheers
