Sprints with Evan: Creative Testing Best Practices | Patrick Gatterbauer

Patrick: for recording everybody.

Evan: Awesome. So we will get started in a couple minutes here and while everyone's throwing in the chat where they're from, love to see it. Love to see it. I love to see all the even new market call outs. That's amazing. Uh, what I'd also love to know is we have just kicked off or I guess a month into it now Q four. So if you have an emoji in mind to describe your Q four, what what would that be everybody? How are we are we on top of things? Is the room on fire? Uh, Patrick, for you, what's an emoji that would describe how you're feeling about Q four so far?

Patrick: Have a look. Roller coaster.

Evan: Now to be honest, I think this this Q four, just my opinion, will be the hardest ever.

Evan: It's the first Q four after Corona. We were to be honest, gold diggers the past two years. Um, yeah, that's just true. And see how it goes. But I think, um, it's getting very wild this Q four. And for some of our clients, we're using a strategy, we keep normal messaging, normal ads, and just do Black Friday sales on page. So we try to get quality traffic and convert them as a sweet like now in trick or treat.

Evan: This is why I'm so excited for everybody. The recording and everything. Cuz right away, you see how you jump in? Like this is why. Um, so I'm gonna go ahead and share my screen now so we can kick this bad boy off.

But welcome to the party, everybody. Patrick and I are super happy to have you here today. And like what I'm really excited about is I've had the privilege of of knowing Patrick and really respecting him for the past couple years. So I'm excited for everyone to just get like a peek into his brain. He thinks about things in a different way, which I'm really excited to uh, for him to dive into further. But one of the things that I wanted to mention is, um, if anybody has any questions throughout the entire presentation, please just throw it into the chat. Don't use our Q and A function, just straight into the chat. And what I'm going to do is if it's relevant, I'll make sure to pepper it in throughout, but if not, we'll make sure to tackle it at the end. And shout out to everybody here. You all know your stuff too. So if you have any answers for people who dive into the chat, throw those answers in there too. Amazing. Well, to kick us off, um, where I'd first love to start is like, I keep talking about how great Patrick is and it's on multiple facets, right?

So like a big thing for me, uh, with Patrick and myself is he's a well-rounded human at the end of the day. So he spends a lot of time and now he's in Dubai. But at the end of the day, he has a couple kids, always learning, always a student, which is something I really admire.

And how this is backed up is just by that professional expertise. So something about Patrick is he's been in marketing for 16 plus years and everybody sees the accolades associated with this. He takes on clients who are again, looking to scale and make a big difference. And it's only a handful of clients at the end of the day to make sure the job is done extremely, extremely well. And he always goes out of his effort or out of his way to talk to some of the best in the game. So people you see on Twitter, people you might not even heard about to really learn. And what I've learned is Patrick is is truly a master, which is this is a master class that we're going to try to get to today. Hopefully I did you justice there, sir. Hopefully I did you justice there.

Patrick: Thanks, Evan.

Evan: I got you. I got you. Okay. So, um, to kick, to kick this off, before we actually get into the actual process of, and I'll stop sharing my screen here. But before we start getting into the different steps within creative strategy that you hear us talk about all the time, I would just love to know overall, how do you view paid social channels in the grand scheme of things?

Patrick: Um, yeah, I see a very big shift. Um, and I think it is two years now, more than two years now. I made a joke, which is, um, 10 years ago we were able to scale businesses with over billboards. And it seems like we're going back to this times again. And if you have a look at the current trends, tracking loss, um, browser with cookie blocker, um, what Apple is doing now again, I'm not sure if you guys read it, but yeah, it's getting harder definitely. Um, but anyway, that's a good chance, uh, to go back to marketing. Uh, we all had the luck in the past five years, at least the past two years are harder, but three years before, letting the algorithm do the work. So, and a lot of brands or agencies still stuck in this times, um, doing more offers, higher offers, more bundles, different bundles instead of focusing on the message. Um, just to save the return. But this is not going to work anymore. And the biggest shift I see, um, from the technical part, so we are moving in a in a black box now and it's like throw everything in and see what comes out. The biggest thing is every push channel or at least Facebook, TikTok, TikTok and and Facebook is the hardest impact turning to um, pull channels. So people are actively searching on social platforms. And I think this is the biggest shift. So I'm curious how because also Google is aware of this trend, what's their solution? And I have a quote from Google's vice president, um, who is responsible for search results. Um, what's their solution, what's their way to catch up? Because at the moment, most of the young people and also people like us starting searching on TikTok for restaurants or recommendations, on Facebook for restaurants, um, Instagram for places, whatever. And this is going to be a big part also in our research. Um, yeah, and I don't want to tell too much, but yeah, the biggest shift is is definitely, um, push channels, uh, are the next pull channels. So we need to think about this in every situation when the user start also searching on this channels, how can we use this information?

Evan: That's a really big point. I think like understanding your push and pull channels, what that starts to mean. And because you have that definition, you can begin to manipulate is probably the wrong word, but be able to pull the strings you need to get the necessary outcomes. So, uh, I like that you've set the stage that way. And I think like a really good place to jump from here is we know that the arguably most important spot is the research that goes into things. So you and I have had extensive conversations about like ways we've seen a lot of people talk about research, but I also know there's other ways that you dive into things really deep. So I'm wondering, would you be able to walk people through the process that you go through when you're conducting your research if you onboard a brand, you work with a brand, all that kind of stuff?

Patrick: Definitely, yeah. Should I share my screen now or just

Evan: Yeah, let's do it. Let's take them through it.

Patrick: Um, so as said and mentioned, for me, one of the biggest things is, so sorry, just Evan is in my slideshow. So can you see my screen? Can you confirm, Evan?

Evan: Yep, you're good to go.

Patrick: Okay, great. Um, push is going to be the new pull. And if you don't adapt your processes to this point, you will lose in long term. Um, in two 19, 19, Facebook announced this, reach people who are actively shopping with ads in Facebook search results. It's very simple. It's, uh, when people are using the search, they place your ad in their search results. And this was 2 19. So I was like, okay, Facebook is adding a new placement. There's a there's a reason why they are doing this. So I was constantly thinking about how can we use this information to do something. And this is the quote I mentioned before. Um, this is a study of Google. In our studies, something like almost 40% of young people when they're looking for a place for lunch, they don't go to Google Maps or search, he said, they go to TikTok or Instagram. And this is the senior vice president, um, of Google's knowledge and information division. And what I meant then, he added, um, keeps learning over and over again. So Google also needs to improve. And this is an interesting game at the moment. I think there's a huge change and as mentioned, it started in 2 19. So see how fast Google can adapt. I think they Facebook is not the biggest competitor for them, but TikTok definitely is going to be a big challenge. Um, and TikTok's answer was, I think one month or two months ago, they up their video description from 300 to 2,200 characters. So with that said, they are actively pushing the search market for themselves. So because when you have a 300 description, letter description, and you turn to 2,200, there is a reason to do this. So it's like, I'm sorry to say that, but it's like writing a blog post on TikTok. Makes no sense for me now, but this there's a lot of money in the background. They have a reason to do this. And so the biggest value pump for for me or in general, I think in in this time, in this area is there is one solution which solves all our problems in research. If we can identify the search terms, because when we know the user search terms, then we have a retrospective on the problem. So we address the problem and can showcase the solution much easier. And this is a big thing. Always when we do research, we are heavy on the point where I say, okay, what are they looking for? And when I say what are they looking for, what's their search term? What are they typing? And we try to figure this out. And this is a thing we started in 2 19 as we, yeah, noticed there is this search trend on on social. It's called seeding, social seeding. And basically what social seeding is, you are using different creator channels on YouTube, also YouTube, YouTube, Facebook, Instagram, TikTok, and you try to have organic posts stuffed with keywords and text to rank when they are looking for you. So you need your brand keyword, you need different, uh, keywords to rank for their search result, to to be in the top results of the search. And this is, this is what we call social seeding. And in a group I am, I we had a discussion about a few things, but the main point is this gray here. I think the main benefit of the story is we need to understand that paid social is not just a push channel anymore. It is push plus pull. And we need to be aware of this, use keywords in copies, matching with keywords on the landing page. So in my brain, as I wrote it here, um, matching keywords in a file name, it's it's also a thing we are doing and you will see some tests. Um, perhaps testing keywords on the thumbnail of video ads, etcetera. The last point is really, let's see how how fast the game adapts and what's the next changes in this platforms. Um, I don't think we go back in a time where also on Google we did kind of keyword stuffing, but we need to be very sensible on this point to use the keywords in our copies, to use hashtags with the keywords for the organics because this is basically what the platforms are looking for also with hashtags related to find our product or our brand in the top results. Um, yeah, and that's why we started, um, the seeding strategy for our clients. Because fact is, um, with every dollar we invest in an impression, um, and this there's no matter if we are using conversions or brand awareness objective, we generate awareness for the whole industry and not just for our product and brand. So every time, um, our brands have an impression, it's not just an impression or awareness for their product and their brand. It's also it's the product. And if the user is on Google, on Facebook, on TikTok, on YouTube and search for the product, for a product test or product preview, whatever, and our competitors are in on the top places and not our brand, this is terrible. So think like, um, meta, TikTok, um, how can we improve ranking with our organic posts and especially with the third or with new organic posts and how can we use this knowledge to create better ads? And as mentioned, else other brands will take the conversion for sure, which brings us to the research part. Um, research for me, it's extremely important for me and my team to stick to this process. It's a five step, the fifth point is on the next slide. A five point process where we start with all the data, the brands data from the customer service, sales team, get the three to five top questions asked. So we have a feeling for the brand, what's the customer's intent or also three to five questions why they didn't buy. This is also, this is gold because then we can try to we if we identify this points, we can try to speak about it in the ads in a way we solve this problems up front. And the next thing is search engine. So basically looking for articles, blogs, magazines, everything on on Google or Microsoft, what you prefer, but basically Google because it's the highest volume. And why we are starting with the first two points is basically psychology, we want our brains to be primed. We want to be deep in the product from the brand's perspective, number one, and from a market perspective, number two, because if you have top blog articles, you get a lot of knowledge from different perspectives and different people, different magazines. But also this is the first step in identifying concepts or angles, however you name it. And the next thing is reviews. Reviews, I have examples, so one example, always search for the reason why. It makes no sense to take every five star review, put it in and say, hey, nice, a five star review. That's shit. That's totally shit. And I prefer to do it manually and not to automate it. This simple reason, if you have automation like with web scraping or something, you end up with a lot of of reviews. And so you have no selection. And this is why the first two points are so important for me to start with, because with point three, we we primed our brain and on point three, we know, okay, what we are looking for, we're looking for the why is in the review, is there a solution mentioned? Is there a problem mentioned? Or is it just a nice review? Or is it just a bad review? So either way, if there is no why, if there's no solution, if there's no problem, it's not worth anything. And then for sure, the social search, what are they talking about and how they are talking about? So we need to use their words. Read every comment related to your product, to your brand. And then, um, store your results and keep it simple and stupid. So this is our agency template. Uh, we will share it after, um, the presentation. I don't want to open it now because it takes a bit of time, but I will definitely also provide a loom record to walk you through. We really try to or I'm a big fan of try to keep it simple and stupid. So everyone can adapt and everyone don't and so we don't waste too much time in documents with 20, 30 columns. I saw so many things and really try to improve it. And the biggest thing for everyone, keep in mind, it's their mind, not yours. So when you do research, um, you need to think what they need, what they are looking for and not I'm the brand, I know what's my product, I know what I can, I'm so good. Uh, yeah, you think it and but what's the truth is what's the user thinking about it? And so to all brands and agency, avoid involving your ego, just listen to your customers and do your research. This is so important. If we have our ego on the side and go in this research, it's a way better outcome. And so this is the first gold. When I say digging for gold is when you do the research in search in push channels on TikTok, this is the example TikTok from ASOS. It's not our client, but I I take it because it really represents the the point perfectly. If you type in a brand or a product name or something or a hashtag and you see all these numbers, especially on brand profiles, and this is what we do when we're looking for competitors. And you see this brand profiles at the first point, it's very common. There's nothing special. But then when I highlight the numbers, this is very interesting. You see their average is, I would say something like 30 K when you walk through the whole account or 50 K. And then you see, okay, this the numbers above are pretty high, yeah, it's nice, but look at these numbers, they are huge compared to the other one. So there are two reasons. The content went viral, which is possible, but in this amount of video views, I think unrealistic but possible. And so my guess is they are they used the three ad, um, organic posts for spark ads. So sorry, for spark ads. And this is very interesting because now we can analyze this three organic posts and think about what's the hook? So what's the thumbnail? What's the hook? What's the message? What what is happening in this whole video? And maybe or hopefully we can find some interesting insights and improve if we are a fashion label, something, improve our ads or at least we know, okay, ASOS had spark ads on this three or invested money on this three, there is a reason for it. And so this is very interesting and analyze this organic posts. This is a four step analysis method. So you start with thumbnail, with the hook, analyze the content, and if is there a call to action? So is it a direct response content? Um, if yes, which call to action are they using and where are they using the call to action? Is it just at the end or is there a soft call to action, um, in the first seven seconds? This is also something you should be aware of. Use screen record or use tools like foreplay.co to save the video and write down your insights and learnings and share with your team. It makes no sense just watching it, thinking, okay, this is nice and jump to another research blog. So you lose everything. Um, another great point is, so we, uh, will onboard a dog client, that's why I use this example. YouTube results because YouTube is social search and thumbnail Champions League. You need to write great headlines, great descriptions, but also you need nice thumbnails for curiosity. YouTube for me is a nice channel to get inspiration. And I highlighted here three things or three words because when you have a look at my search term, it was how to wash a dog professionally. And you told me, Patrick, no one says how to wash a dog. It's dog buffing, how to buff a dog or how to buff your dog, but no one says how to wash a dog. And so this is also a nice insight for me. Uh, because general, when you use this search engines, you get suggestions. And this is why I left this in. Um, it's a very simple suggestion, but you will see if you type some search terms, you have this corrections and this is super helpful. Also this one, five minute vegan meals. And now it's a good example to think about angles or hooks. I called it five in my search term, five minute vegan meals. And then there is lazy vegan recipes, crazy easy vegan recipes, mouth watering vegan. So there is definitely, um, the point with laziness and convenience, but also with with the taste. And so we have two angles now. We can say, okay, if we do one creative, we have three. We have this lazy part, people who are lazy, we have this easy part. So for easy is not specifically fast. It's just it's, you know, the difference. And so a lazy person, easy and mouth watering. So this is also super interesting. And then look at the thumbnails. Can you use the thumbnail for an ad? I definitely say so, especially the the woman with the timer or different meals here. And this is gold. What's also nice when you search for brands, they tell you usually, especially the bigger one, your bigger competitors, they tell you their USP in Google in the Google snippets. So if you have the search results on Google, identify the USP of other brands by just putting in their name, their product and see what are they using? What are they writing here? How are they describing themselves? And so we have three different, um, examples here. And then, so you know, okay, this is their USP, you won't use it because you cannot stand out if they are placed here. This is their USP, you need another one. So why are you better? How can you compete? This is super interesting. And so this is also a point that we are doing. We have a look what are the competitors doing, what are they saying about themselves or how do they see themselves? What's their positioning? And so we can find a better position or at least a different position.

Evan: Patrick, I love that. And I think it's so important like being able to really treat it like push and pull so you can see that information. So I'm almost curious, like now that we're able to to look at some different ways that we can conduct our research. I know there's the basics that everyone we're we're aware about, we've chatted about a lot. Now that you have this information, how do you translate it into action? Meaning how do you start to translate that into a brief of what to do next? What does that look like for you?

Patrick: Um, this is this is in our brief, in our one more. I'll as I said, I'll share it with everyone. Here is a document where we store all these results. And how we do it is at the end of every column, so we have different tabs in Google document or Google sheets. And what we are doing is at the end of every, say, we do, we start with the customers, then with the blog section. We're using the link, we're putting in the link, the name of the blog, and the important paragraph for us. And at the end of the column, there is a column named tag cloud. And tag cloud is basically, value is basically is it positive for our brand or negative or is it general positive for the product or negative? Sometimes you have products, sorry to say it, but in one research yesterday, um, it was it can cause cancer. And so this is really shit. You need to talk to your client, hey, is this really, is this true? How can we solve this problem? What's what's the reason behind it? And then you get the generic answer, hey, we need to put the sticker on it because this is a kind of of rule, this is on every product in our niche, it's the case. And so we have the values positive or negative. We have the headline, we have the important paragraphs. And at the end there is a tag cloud. And when we look at the example from YouTube, then the tag cloud is like lazy in the next line, it's easy and, uh, water mouth watering. So basically the tag cloud in the next step decides what angles we are using. Is it mentioned often, say appearance, design, quality, or are the colors because maybe we have brands, they have seven or I think the one nine different colors on a t-shirt. And so color is a big thing for them. It just works because of the color. And the next thing is design quality and then you end up with sustainability. Every sustainable brand thinks sustainability is a big thing. It's not. In your target group, sustainability, you cannot win a match with sustainability because they are looking for sustainable, it's clear for them. It's necessary. So it's design quality or the color choice.

Evan: And I think the Lego is literally a perfect representation of that, right? Like we've all been there either for our kids or as yourself as a as a child, it's like assembling it as you see fit. And then that way, if you pull it apart, you can kind of put it back together a different way. So now that we've been able to start briefing our teams who are going to start spitting out different variations and all of that kind of stuff, what I'm curious about now and like to really start digging deep into your mind is the launch strategy more than anything. So when we're ready to launch this within our paid ads, first and foremost, talk to us about, I know a big thing that we keep talking about is push and pull as the first thing. But the other thing that I'm so curious about is just like how are you doing and launching your creative tests?

Patrick: Yeah. So I'll start here. So the launch definitely is one of the biggest thing here because we identified, um, a nice thing. And short recap to the search in push channels. I was like, everything and everyone was going broad. So it started, so everyone was telling you need to go broad, just use 10% look alikes, interest stacks. So it started, I think two years, 1.5, two years ago. So it was really mainstream to do this. Um, we started a bit before this time with, uh, the creative testing at least to try this. We still used at this time, so 2.5 years ago, 3% uh, or 8% look alikes. But testing was mainly broad. It was this broad or a 10% look alike. And so I was, how can we help the algorithm finding best better customers up front? Because when everyone is going broad and everything is going broad, Facebook is a black box, um, giving away targetings. If you think about detailed targeting, what was possible two years ago and what's possible now, it's nuts. It's crazy. Um, so it's exactly the way it's it's broad. Facebook is like or TikTok is like, um, give us so many data, so many data as possible or so the biggest audience, uh, you can have and throw in creatives. So all the part is done with the creative over creatives. And then it's very simple. You have a review process and basically Facebook and TikTok are reviewing your images, scanning the text, scanning the image, video, especially thumbnails, and the landing pages. This is where most ads get rejected at the end with the landing page problem. But what there must be more because this is what everyone, this is what Facebook was doing and this was kind of no surprise. Everyone knows this. And then I was, okay, how was Google showing images and videos in search results? And how can we use this for paid social? And then I ended up with my first test. And the test was in this structure, we still use this. Um, one ad set and in each ad set was just one ad. So it was exactly the same ad set, the same audience, everything. In my first test, we tested this on different accounts. I wanted to really to have the each ad should have, and you see the impression counts, a similar chance. So if you have three ads in the next example, in one ad set, you see end up, um, some ads or one ad ends up with a very less impressions, but see the CPC, have a look at the return. And now think about the following part. The only thing, and it really it was the only thing different in this three ad sets was the file name before we uploaded the creative. And but I think this is not so necessary, but we still doing this. Um, the ad name. So the part is if Facebook is scanning the image, the video, the landing page, I think they are also try to get as much information as they can. When we have a look at the Facebook catalog, they ask you for the Google product category for a reason. The system is not that smart. So they tried to get every data point from you. And if I can help them with the file name, at least what what I did, I ended up with great results and see the difference. And also in the next test, the next test was one ad set with three ads in it. And you see it's the same. It's this ad and everything here, it was very, it was the same. It's a very basic test. Everything was the same, just the file name and the ad name was different. And have a look at the impressions. It's not a five impression versus a 10 impression. It's not any fool you see on on LinkedIn or something. There are a lot of impressions on it. So it's it's for us it worked great. And the only difference, search keywords in file names. So instead of generic naming like date, ad type, info, whatever, this was our or is our structure for the file name and for the ad name. So we have the date, we have the product, we have the type, so it's a video. We have then also the name of the of the creator. Um, what we see on the ad, it's presenting nails covering face. So it's basically just a hand in front of her face. Um, it's user generated content and it's summer color nails landing page. This this was our method. And this was gold for us. It worked insane. I also talked about in a group with a few guys and they tested it, um, different, different, uh, results, but mainly it was a difference. So for them, uh, no one, no one, as far as I remember, denied that it's at least better. Maybe it's not a way better, but at least it was a bit better. And imagine if you have a 100 K or 500 K ad spend a month and you just with this simple technique, you can higher your return 0.2, 0.4, this is insane. This is not nuts, this is insane. And so my important findings on this point, the keywords used, so in the in the ad name, the keywords must match the keywords on the landing page. And this is important because now I go back, what's the difference here? The first ad was with English text. So it was English text but a German landing page. And we also tested it the other way around and had similar results. It never worked. The second is a German file name and the German landing page. So the keywords in my file name match the keywords on the landing page. And we had a great value, great outcome. And also if you see outbound clicks to landing page views, we have here kind of 76% tracking loss. So, yeah, it's crazy.

Evan: When when Patrick showed me this first, it was so interesting because I saw it live while he was running the test like a year and a half ago or something like that. And it was literally every single creative exactly the same. Only difference, ad names and the creative file names themselves. So really, like when he was chatting about the push and pull strategies and treating these social media platforms like search engines at the end of the day, they're going to scrape that data to then determine because we're saying broad audiences, how do we then say within that broad audience, who is the most likely to buy? Of course, we call it the algorithm, but Patrick surely dove into here is like, how does that actually work? How are they actually trying to deliver that? Which is so cool to me. And Patrick, we're running up on time here and I definitely want to give people a chance to ask questions close to the end.

Patrick: Yeah, so sorry, sorry, sorry. So, um, yeah, and this is, so everything here you can read and matching the keywords and I think it's there's a correlation with the meta tags and metadata on the landing page. I think this is the only observation when we saw minimal better results, it was mostly on clients which had not took not the best care on their shops.

Evan: And the testing setup is very simple for us. It's a ABO. We test, um, on ad set level, ad group, broad targeting, um, and three ads. We always run one test with three ads. This is it. The settings, we don't use more ads because when I have five ads in it, two to three ads end up with too less spend to have significance. And then I always have the feeling, maybe they have, they can do better, but Facebook or TikTok, um, did the choice too early. And so I always prefer three ads for one test. It's also very conservative. I know with DCO, but I hate DCO. It's showing you results 7, 9, 14 days. I always have different results. I love it for sales campaigns or general short-term campaigns. This is okay because you don't have a lot of effort, but in general for testing, I don't recommend it. And you see the setting.

Evan: Yeah. Also, sorry to cut you off. I'm just thinking like now that we have this live, I'd love if you could show everyone like some of the metrics you're using to actually evaluate success. And then that way, I think like with five minutes left, we can give give people a chance to ask questions.

Patrick: Nothing. Perfect. So, basically we follow a very basic formula. Grab attention, hold attention, make them click. And so the platform has just, it's it's every platform wants one of those two things, making you happy, bringing you conversions, or making them happy, the users with great user experience. So engagement is king, but conversion is queen. What I'm talking about is don't chase for the highest thumb stop ratio. Don't chase for the highest average watch time. This is nice for engagement ads, but we are in direct response, we want performance ads. So we need a reasonable thumb stop ratio and a reasonable average watch time. And this is the point. You if your ads are too much in education or entertainment, they are watching. Yeah, that's nice. If you have a 40 or 60% hook rate on a broad audience, there are two ways. It's you have shit traffic or you you bring shit people in the funnel because you are just, you just hooked them. It's like clickbait or you have a product everyone needs. There is no other reason for such high thumb stop ratios, especially in niches you usually, um, with broad ad sets, you can't reach this. Um, and we have three levels. The first level is the engagement level. And one I want to present you is post engagement divided by impressions. I call it fatigue because you can see in a post engagement is everything. It's a click, a share, a comment, it's three second video views. There's a lot in it. And if you see fatigue decreasing, you know it's time for new ads. It's definitely time for new ads. Thumb stop ratio, average watch time, everyone knows them. Um, and what we do is kind of video engagement level. It's we have a look at who watched three second, who watched 25%, 25% who made it to 50%, and the whole funnel. So be aware of of this. If there is any significant change, then you have to think about, especially on running ads. If more people are between 50 and 75 than between 25 and 50, they're looking for information because this is how Facebook works. They don't measure the 50%, 75%, they measure where you click on the on the progress bar. So if you click to the end, you are on 75%. So they're looking for information. Um, then we have the second level performance data and there's another, um, self-calculated metric we are using. It's outbound clicks divided by three second video plays because what I want to know is everyone who watched three second of my video, did they visit our website? So this is also super interesting for us. CPC and outbound CTR, everyone knows. The third level is also nothing new. Click to purchase ratio, purchases divided by outbound clicks, cost per results, ROAS. So the third level is a problem because on this level you have this iOS influence. So data may come up to three days, 48, 72 hours delayed. And the main level, this is and these are exactly the slides from our creative SOP. Always remember, challenge the platform, but don't forget the goal. And this is so important. I know we're always talking about attribution tools and so on, but if the platform is telling you this ad is shit, but the tracking tool is telling you this is a ROAS of 3.4, I think there is something wrong. And the other way around, if the platform is telling you this is a great ad, but the attribution tool is telling you, hey, this is a 0.2, you need to dive deep in this because to be honest, I know a lot of people telling you they are completely in 100% data driven, but if we have an ad with great engagement KPIs, thumb stop ratio, promise, this is our our metric, it's called promise. Promise basically, I think a lot of guys call it hold rate. For us it's promise. Um, and we have thumb stop ratio divided by three seconds or 50% video views divided by three second video plays. If we have videos longer than 45 seconds, I definitely use, um, the 50% divided by three second video plays. Yeah, this is, um, the thing. And then this is what we are doing. We have this three reports. We have, this is important, every account has a benchmark has benchmark metrics. You cannot compare your thumb stop ratio, um, with with the industry. That's not so when someone tells you you need a thumb stop ratio of 30 or 40%, that makes no sense because they don't know your product, they don't know your offers. This is you should not chase the high thumb stop ratio. Look at the average of your account and see, okay, what's possible? And in which ratio with a client, it's between 19 and 26% thumb stop ratio, we have the best returns on platform but also on third party tools. And this is what we are doing. We have this three reports. We have, we filter average watch time is greater or equals to three because in this specific client makes totally sense. Then we have the second report and so we have this emojis for motion to have, uh, to find our reports faster. And then we have the second one is outbound CTR is higher than the average. And also thumb stop ratio is higher than 20. This is what we knew, okay, higher than 20 is perfect. This is for us the sweet spot in this account where we don't scam people with like clickbait, but we also doing great money and great return. And another great feature because it's visual, it's the first metrics KPIs I showed you, we have this in exactly in this, um, order in our in our ads manager. And what's also nice in motion is you can have the 25, 50, 75 and 95% video. It's basically you see who watched 25, 50, 75 and it makes no sense in this reports to have 100% because Facebook is telling you 100%, but in reality it's 97% because then they say the roll outs coming and people leaving the ads. So it's really it's the case, it's 97%.

Evan: And I think Patrick, I I I promise I didn't tell Patrick to say that one just about motion. But we have a a couple questions that rolled in that I think we can tackle in the last five minutes. So in the chat, one of the most recent ones came in from Michelle. And what, um, they're asking is if a platform is saying it's a it's a good ad, but metrics are terrible, in your experience, what are some of the possible problems or reasons for this?

Patrick: Um, if the platform is telling you, so we're talking about the soft KPIs are great, but the return is bad, is this correct? Or the platform is telling you in general it's great, but third party is telling you it's

Evan: Engagement metrics are strong, but bottom of funnel isn't great.

Patrick: Okay. Um, I don't see any problem in this. It sounds weird for some guys, but, um, think about this. If you have this great soft KPIs, I call it soft KPIs, and it's a top funnel ad, so UGC or video, you're bringing people in. And sometimes it's exactly this ad bringing new people and they're converting on another ad. We have, um, it in some campaigns, we see one ad has terrible, terrible return metrics in outbound CTR and returns, but high soft KPIs, very high, so really wide above average, we let it run. Because for me, I know, okay, any other ad will convert these users. We bring them in our funnel. This definitely, um, what I what I say and what I prefer is if you have an ad with high soft KPIs, very high, let it run and try to push it.

Evan: Are you ever worried that in that situation, it's a traffic situation? Where it's like the traffic quality might be reduced a little bit with high, um,

Patrick: Definitely, definitely not because, um, if you have a bad return, it seems like they are not clicking. For me, an outbound CTR, if outbound CTR also is high, then you have a problem because then you promise something in the ad and they don't see it on the landing page. But as far as I understood the question, it is if I have a high thumb stop ratio, high average watch time, high hold rate, if this is the case and it's really it's very high, let it run. If you see also you have a high outbound CTR, then you need to turn it off because then exactly is happening what what Evan was telling us, there is a mismatch between the ad and the landing page. In these cases, because we had this one month ago, we tested a new angle and had insane outbound CTRs. Thumb stop ratio was great, average watch time, everything was great. Also thumb stop ratio, it was insane. The ads bombed, but we had nearly zero conversions. So I think it was two to three conversions on 7, 900 spend, kind of like this. And then we switched it off and asked the client to build a new landing page specifically for this angle. And this is killing it. But great question, Evan. Thanks, yeah.

Evan: Of course, of course. And I think with that, I know there's a bunch of questions and everyone could like ask a million of them. Please make sure you connect with Patrick just on LinkedIn, on Twitter, all that good stuff. I'll be sure to include that in the email out tomorrow that has the, um, the recording as well as some of these templates we've been talking about today. But Patrick, really, really appreciate you, my friend. This is this is great. Again, the mad scientist at work, everybody. Thank you so, so much for this. Really appreciate it.

Patrick: Yeah, thanks for having me. And sorry to have too less time to talk more or go a lot deeper, but I think you you got the basics and the understanding for what we are doing and how we're doing it. And stick to the process.

Evan: Entirely. Amazing. Thank you everybody. We'll talk soon. We'll see you at our next one. Connect with Patrick, all that good stuff. See ya.

Patrick: Bye.

Evan: Take care everybody.