ZOE: Unboxed - The one hook that cut Zoe's CAC in half with Raphael Yarish

Speaker A: That whole hook, it essentially dropped KAK by, like, half.

Speaker B: Wow.

Speaker A: Like, that one video, we made ads with, like, Stephen Bartlett. You know, I was literally holding the iPhone, shooting with him. The brand team at the time maybe had like six hours with him, and I only had 30 minutes. It was still in my, like, probation phase. I need to, like, drop KAC here. I'm getting fired.

Speaker B: Yeah.

Speaker A: When I had found kind of the, like, the CGM M hook of, like, I'm afraid, the anticipation of what's going to happen. What is this device that you're, like, putting into your arm, it essentially dropped KAK by like, half. We just would book a place, like, once a month and we would shoot like 200 videos. If you plan it out properly, you can actually do 500 ads in a day with, like, three people. Just stitch three of your top ads together.

Speaker B: Yeah.

Speaker A: And you'd be surprised the amount of times that works. I'm like, I'm not even joking. The whole brand shoot with, like, a bunch of cameras like this. Yeah, yeah. Uh, and then just me on my iPhone. Yeah. But it's like, actually the performance ads, we're probably seeing 20x times more than any other brand content.

Speaker B: Welcome back to DTC Unboxed, where we unpack the growth secrets behind the world's leading DTC brands. Today's guest is Raf, who spent over $50 million, uh, on Meta over the past five years working for brands like Zoe, Speechify and Mojo, deploying and creating actually thousands of ads himself, with his own two hands, shooting on his iPhone. And, uh, he's now building Ad Manage, which has helped to launch nearly 2 million ads on Meta across their brand portfolio over the past couple months. Today's episode is full of really great tactical insights. If you're looking to kind of scale your creative systems, this episode is a must watch. So, Raf, how did you get into the. What I need to see. Excited to have you here.

Speaker A: Thanks for having me, Ryan. Um, yeah, I got into it maybe about, like, I don't know, eight or nine years ago. I, um, actually started in lead gen. Okay. I worked for one of the biggest, um, lead gen in Europe called mvf. Um, and I did about, like, two and a half years there. And it was probably, like, one of the best places to start, uh, because it was one of these places where they didn't care about brand at all.

Speaker B: Okay. Yeah.

Speaker A: And it was, like, very extreme. It's like you had to get, you know, your margin, your cost per lead. It had to be this or it's like the Whole campaign and that whole vertical, like, wouldn't run at all.

Speaker B: Super tactical stuff.

Speaker A: Yeah. So it was extremely, um, cutthroat in terms of how you achieve performance. And I actually had started on the very niche channels, uh, which was like LinkedIn, Twitter, Snapchat and Pinterest.

Speaker B: Oh, wow.

Speaker A: So I was one of the first people to do ever, probably the first B2B, uh, lead gen on Pinterest.

Speaker B: Yeah. How did that work?

Speaker A: Uh, it was very difficult because it wasn't built at all for that use case. Yeah. Uh, but it was like a bunch of static ads, uh, at the time for asset tracking, which is like, um, used in construction and very popular in the US And I would run that, uh, for people who owned homes, because homes was very. Home inspiration is very popular on Pinterest. So you could get like a side angle, essentially target those people. Um, and, uh, there would be some overlap of people who needed asset tracking or vehicle tracking. Uh, yeah. Basically you just had to get like, extremely creative in how you, you cracked those performance channels because no one at that time had really ever done it. Um, so like, Facebook and Google were definitely more in a solid place, bigger scale. But those channels at the time was called the EPM team Emerging Paid Media. Um, and yeah, I learned like all the ropes there and it made me like such a good marketer into like all the other things I did after that. Um, but yeah, that was a good start. And then after that I went in house, did different startups. I went from agency to in house, uh, did about a year at Mojo. Um, and then after that I also went to Zoe. Uh, at the time after. And before Zoe, I had also done my own e commerce stores. I did like this Covid brand thing, um, and then figured out that I really didn't like, kind of the op side of E comm ever running it myself. Yeah, I was like, oh, ah, this is such a nightmare. And I was like, I'm much better at the marketing side. Um, and Covid was like, was tapering off. So I was like, actually I'm just going to go like full into marketing and try to just be the best at that. At that time I had basically pivoted into Zoe, um, and became essentially the head of performance there. Um, and that was some of the like, most fun work, uh, I've probably done, uh, in terms of pure performance marketing. So, you know, we made ads with like, Stephen Bartlett where, you know, I was literally holding the iPhone, shooting with him. And I was like the brand team at the time maybe had like six hours with him and I only had 30 minutes.

Speaker B: Wow.

Speaker A: So I was like, already you can see the disparity there. But I was like, okay, well, if I only have 30 minutes, all I'm getting is hooks. Yeah. Because that's the most important thing. Yeah. So it's literally like, I was like, steve, we're going to just film like back to back to back to back. He's like. He's like, I got this. Don't worry. Okay. Okay. Um, so we did that M. And yeah, some of those ads worked really well. And then we did another shoot with this Davina, who performed, like, even better. Um, yeah, because Davina was like a really good fit for the Zoe audience. Um, and really well known. And it was almost like pre the time when, um, she had done a lot of ads. So it's still like, it was like fresh entry almost to like the ads market.

Speaker B: Yeah. Because she's done quite a few brands.

Speaker A: She did so many now. Yeah. So I don't know how well they perform now. By the time it was like a really good arbitrage. Um, and. And those, like, absolutely smashed for probably like a year. Uh, and it was hard to replace those actually is a common thing in performance. Um, on, like, how do you beat new ads? Uh, but then, yeah, after that, I had done some like, freelance with Speechify as well. Made, like, went to, uh, took what I learned at Zoe and made to the extreme for speech five, like shooting 500 ads in a day.

Speaker B: Wow.

Speaker A: With like three people. Um, so doing that and just like all about volume. And then, uh, from that transition, uh, starting Ad Manage. And now that's what I do full time. And the co founder of Ad Manage. So.

Speaker B: Yeah, that is quite a record.

Speaker A: Yeah. Yeah.

Speaker B: DTC powers. I mean, lots of different things I want to talk about. I mean, also, it sounds like you came from a technical background as well. Right. I saw you like a physics.

Speaker A: Computer Physics background. Uh, few computer science classes. Um, I had done. I actually did like a UBS, uh, internship when I was 18, like pre university. Um, and I was like, it's not for me. Yeah, I was like. I was like. It was funny because I was doing this like, internship thing and I was there for two months and, um, I was working on this project that wasn't gonna get finished for like two or three years. I was like, what am I doing working on this project? What's the point? I'm not gonna like, see the end of it at that time. Um, and I ended up working for like just five different teams. Because I was like, so bored. I was, I was like, I just finished all my work in like half a day of the whole week. And I was like, what am I meant to do? Uh, so I ended up working for like the cybersecurity team for a bit, like the, the, the ops team, uh, like different functions. Um, and that was really good. Uh, and I also did a lot of like, science, uh, like research internships as well. So like an Imperial at Queen Mary. Um, and that was also before university as well. And then I did physics. Um, and yeah, it was a good degree.

Speaker B: Uh, so, yeah, because you always strive as one of those people that's like, very creative but also very technical, which I feel like is a really difficult, like, it's very rare basically to have the combination of technical side and, yeah, creative side. It's probably like a mixture of all of these experiences that you've had, which is quite cool. Um, so, I mean, I'd love to talk a little bit more about Zoe specifically. I mean, it sounds like you joined, um, that company when it was like five, fairly early on. Like it was maybe not as big or established as it is now. And it was quite a lot of, I guess, category education, I suppose, as far as M. I'm aware because there was a lot of like, stuff you had to do, isn't there, to get the end product in terms of health tests and all that kind of stuff. So be curious to understand what the model looked like when you joined. And.

Speaker A: Yeah, so when I joined, they were running zero paid ads. Uh, they had run them, uh, a year before I joined. And CAC was very high, like unsustainably high. And I was like, I can't believe this run for that long. Uh, and, and then, yeah, when I joined, it was at the time when the wait list was building up. So there's a wait list of about 300,000 people. Um, you know, Tim Spector was going on all the morning shows and things like this and building the hype around there. Um, and when we had launch paid ads, it was actually really easy because it basically all those like, super weightless keen people. Uh, but, uh, it was a hard product to sell because it was like three things in one, which was, uh, like a cgm, a poop, um, test, like a gut health test, and a blood fat test as well. And the idea is basically you take all of these things, um, and you get a personalized score on every single food that is in theory unique to you. Um, I said that in maybe about 30 words or so. Uh, and if you think of an ad, you only really have, like, maybe like three to five to really get someone even less, to be honest. And the visual does a lot of the heavy lifting. Um, so what I had done early on was cool. There was no person, ah, or company that did this still, not to this day. And so we actually pivoted away from this because it is so complicated to explain. But what I had done was I actually looked at, um, the US a lot because CGMs were a lot more popular there. Um, and basically reverse engineered, like, what was the top winning ad, uh, for, ah, three or four of the, like, c. Big CGM companies there and essentially, like, used part, uh, of that hook and then the rest of, like, the Zoe explanation. And that worked really well. And now everyone, like, actually knows Zoe. At the time, anyway, I was like, oh, it's that thing, right? Yeah. And that was, like, the hook that I had taken from the US to the uk. So that's so cool.

Speaker B: Uh, and what was the impact on. On cac? Because obviously you said it was unsustainable. So is that. Was that enough to sort of dial it into the right sort of place where you could start to scale, spend or.

Speaker A: Yeah. So when we had first started, cac, uh, was really good for like, a month because it was all the waitlist people. And then became the challenge of, like, okay, now it's passing the point of, like, unsustainability. So it's like the real performance has got to kick in now. And that's when the whole, like, us thing, uh, with that I was looking into. And, um, I remember at the time it was still in my, like, probation phase. I was like, all right, I need to, like, drop CAC here. I'm getting fired. Yeah. Like, no pressure. Yeah, yeah, yeah, yeah. I was like. I was like, I know how performance works. I was like, it's just. It's just like that. Um, so, uh, when I had found kind of the, like, the CGM hook of, like, I'm afraid, the anticipation of what's gonna happen, what is this device that you're, like, putting into your arm? Uh, that whole hook, uh, it essentially dropped KAK by like, half. Wow. Like that one video.

Speaker B: Yeah, that's so interesting.

Speaker A: And then I was like, yeah, it was actually insane. And then I was like, okay, job secured until the next winner. Uh, and then, um, yeah, it was just about, like, iterating and stuff, stuff like that. But that, like, insight of, like, taking what works. Uh, in this case, it was like, Adjacent, kind of not competitor, but vertical, because they did CGM only. Um, and it was actually really hard for Zoe to eventually, like, pivot away from the CGM image because they, like, work so well. Once you have a winning ad, it's like, you can't turn that off. Uh, so anyways, yeah, that's how we kind of, like, we pinpointed a specific thing that worked from an ad perspective, and then you can kind of go on to explain the rest. But that's how, um, we kind of started that category.

Speaker B: Nice. And sorry. Just for anyone that doesn't know cgm, what's the. What's the acronym?

Speaker A: Good idea. Good point. Yeah. Um, it means continuous glucose monitor. It's like, uh, something for diabetics originally, actually to measure your blood sugar at, uh, any point of the day, basically, so that if it ever goes below a certain level, it will, like, notify you on your phone.

Speaker B: Got it. And so that. That hook that you talked about specifically, um, it was that basically building a bit of a curiosity gap in terms of, like, what is this thing on the arm? Like, what is that doing? Is that kind of the main thing? Or like, what was the kind of breakdown of that hook, do you think, that drove the success?

Speaker A: Yeah. So something I had learned from MVF was like, have you. If you ever hold a small device that looks ambiguous to the camera, it click through. It already goes through the roof. Because it's like, what is that? I need to know. Um, and it's like you put like, some copy of, like, this tiny device, and then it's like, absolutely smashes. So I use that insight from what I had from MVF and then combined it with the anticipation of almost like, like, fear of, like, ooh, is this gonna hurt? Because I personally hate needles. Like, I. I absolutely hate needle. So, like, I could really relate to this hook. Um, and when I first did it, I was, like, really afraid. I was like, oh, man, is it gonna hurt? And then, um, yeah, it's basically you just film that entire journey in two seconds. Yeah, yeah. And then boom. And then like, that's it. And then it's like, oh, it's not so bad. And you get the payoff Im then you go in as well to the story. So.

Speaker B: And did you find that, like, in your customer data or any post purchase stuff that was like a blocker or any, like, surveys?

Speaker A: Not really. Uh, there was the finger prick one, which actually was a blocker. Uh, but that was a lot more like, bloody. Yeah. So harder to show on Facebook. Um, and not as like shocking, like satisfying. It's like just a finger prick. Um, and people have seen that for years, but people haven't seen the cgm. It's like a relatively new device. If you're not. If you don't know someone diabetic, you wouldn't have really seen it. So it was more like the novelty of that. Plus it made like a clicking sound when you did it. So that uh, so it had the sound effect, which is really important in ads actually. It's super underrated. Uh, like the sound effects, um, and then the emotional relief of like, oh, it doesn't hurt. So you get like this massive facial transformation in about one or two seconds.

Speaker B: That's so interesting. So it's actually more like maybe it's like the reaction actually though is the more.

Speaker A: Is the reaction. Yes, exactly.

Speaker B: Super interesting.

Speaker A: Yeah.

Speaker B: Okay, and then so you, you've gone in, you've been able to reduce CAC significantly. What are the next steps to scale that system? Like, how do you then go, okay, how do we ramp up spend? What's. What's next on your agenda?

Speaker A: Yeah, so, um, at that time, uh, this was, I think, I believe at the time when ASC was rolling out as a campaign type in Facebook and you know, all the reps were like, oh, just put in all the ads. Just do it. And I was like, okay, let's try it. So I basically went, uh, uh, and tried to build a system to be able to like produce a lot of ads. Uh, um. And that's really hard if, especially at that time, like it wasn't as common to be honest. Um, um, and had to like, figure out a lot of stuff, uh, on my own and trying to talk to as many different people as I could. But um, yeah, so it's basically. And I'm a big believer in like kind of like lean teams and not over bloating because I've seen it so much at different companies. So I, that was like my team was like way, way, way smaller than the whole brand team, for example, uh, on the performance side. So what I had done was I basically hired, uh, initially Full Stack creators. Because that's kind of like what you're saying is like I get both the like the analytical side and the creative side. I was like, I just need to find more people like that. And I was like, oh, that's really hard. A lot of interviews. No one's like that. I'm like, oh, uh, man. Okay, I need a plan B then. Uh, and plan B at the time was actually to just hire people who were good at content. So I'd actually look for organic creators to hire them in house. Not, not just to make videos for partnerships, ads and stuff like that. That's more the meta now, to be honest. But back then it was more like take them in house and actually just convert them on how to make paid ads. Uh, and I did this with like two or three people. Um, and I had one static person as well at the time. So it's maybe a team of like three or four. Um, and we're making hundreds of ads a month. And basically the, the, the, what we did was we just would book a place like once a month and we would shoot like 200 videos. Um, and it's like if you, if you plan it out properly with like enough hooks, enough scripts, enough visual differences, uh, and if you have at least one or two actors, you can actually do way more than you think, uh, in a short amount of time and a short amount of visual, uh, space as well. You'd be really surprised. Um, so, and then, yeah, we would also go to like real customers, houses as well. Because people like, loved Zoe. They're like, oh, I'm in, like, I'll do it. Uh, so we would go to their house, um, and film some videos with like real customers. And that was actually a lot harder than it was with actors. Uh, because I was like, it was hard to be like, can we retake that again?

Speaker B: Yeah, yeah. The authenticity doesn't quite come through. Even though it is more authentic people in front of camera.

Speaker A: But there's like so much, much waffle that it's like, it's actually, it's actually really tough to be succinct. Um, so anyways, yeah, that, that, uh, but that was the system. It's like we do shoot once a month. We would try to get out as many videos as possible. We had statics going all the time. Uh, and we would just launch these every week. It was like a weekly cadence. It's like had to launch every single week as well. So when we didn't have the shoot, we would do like you just film at home and make your own videos. So everyone at the time was a full stack creator. Which means you research, you film, uh, and you edit and you launch your own ads. Basically.

Speaker B: Was that. Basically a creative strategist, a creator. So doing the actual, you know, ad creation, editing and then launching all in the same.

Speaker A: Yeah, yeah, exactly. Wow.

Speaker B: But yeah, that was super difficult to like put that into one person.

Speaker A: Yeah, yeah. So you had to like you had to make things as easy as possible on the things that didn't matter. So, uh, as a thing that wouldn't bring you the most value. So, uh, and I had the experience doing this myself, so I was like, I could train people to just do this. And the hardest part was always, like, making really good creative. Everything else was a lot easier to teach, which is why I went for people who did organic content, because it's like, they already have a bit of that skill on organic. And it's so much easier to go from organic to paid, because organic is actually way harder than paid. Uh, so, yeah, for sure.

Speaker B: Yeah, 100%. Because you can't actually, like, activate with speed. You have to wait and you have to get the timing. So I totally get that. That's super interesting. So did you, in terms of, like, training and SOPs, did you just have, like, a very systemized notion board that was like, this is the stuff you need to know was like, you just spend time with this person?

Speaker A: Yeah, I spent time. Um, and, you know, for me, it was like, really about building that relationship, actually. Um, I have a lot of good friends who are, like, really good at the sop. Stuff is not my strength, I would say. I would agree with that myself as well.

Speaker B: Yeah, I'm very bad at that.

Speaker A: Yeah. Um, so for me, it's more like, like, building a lot of trust with someone and like, just trying to teach everything I know and then for them, putting their own spin, uh, on it, because I can't have them making the same stuff I'm making, uh, because I need it to be different, because I need to find more winners. Like, so he's going back to, like, I need to find more winners. Um, so, yeah, it's just, honestly, I didn't have a big team, so it was a lot easier to just go one on one with everyone. So. Yeah.

Speaker B: And how did. So when you say winners, you need constant winners, which is obviously what everyone is trying to do. What did you quantify that? Was it like a number of winners that you were aiming for on a monthly basis or, like a weekly basis that you knew was enough to fuel what you needed to fuel?

Speaker A: Yeah, I'd say it was like, there's different levels of winner, but, like, generally, because everyone has different definitions, I think the easiest one is just like, is it a top spender and is it like a top three or five spender in the account? Like, it's as simple as that. Because there's too many metrics. Otherwise we would try to aim for at least like one a month. And we wouldn't hit that a lot and maybe would be one every two or three months in terms of like the top five winners. But like, the more ads you do, uh, the more you can have a chance of finding those winners. So even if you think like the ad is not great, it's like there is like zero, almost virtually zero risk of just launching it and trying it. If it's not good, Facebook won't spend on it. It's like a dark post. So people on the Internet won't really see it unless they're marketers and they're checking in the ads library. And then at that point if they copy a bad ad, it's like sucks to be them. Then they waste all their time on the research. So it's like the risk is really low. And this is always something I had to like discuss with the brand team.

Speaker B: I was just about to ask you about that because, ah, you feel like a very sort of like tactical performance performance marketer. And obviously Zoe, they were very brand protective, especially I imagine around Steven maybe as well. It's like, how do you, how do you balance those two things and like, how did you kind of navigate those two worlds?

Speaker A: Yeah, I mean there's like a simple like, equation you can show people which is like out of the hundred ads, it's like 95 of them will get maybe like 2,000 impressions. Like, I don't know, I can't remember the exact numbers, but it was like something really low. Right. Uh, and it's like you just show that like actually the risk is really low on this. Uh, but, uh, and like some people would get that, but then, but then it's like it was a really hard battle. And it's actually. I went to the Meta Summit in uh, San Jose like last, was it last week or two weeks ago. Uh, and they were giving everyone the Alex Schultz book, CMO of Meta.

Speaker B: Yeah.

Speaker A: And he actually had like this one great line. Like a lot of the book was, was okay. But there's, there's one great line which was like he doesn't uh, put uh, his internal team. He does, he never calls it performance marketing, uh, because that implies that everything else that is not performance marketing is lower than performance marketing because it's not performance. So he would put as direct response marketing and brand marketing. So it's more on a level playing field because this always happens in companies where it's like the performance and brand team like kind of hate each other 100%.

Speaker B: Uh, I see it all the time.

Speaker A: Yeah.

Speaker B: Every company I talk to is. Has some sort of tension between those two teams.

Speaker A: Yes.

Speaker B: Yeah.

Speaker A: So relabel it as direct response marketing for internal purposes and, uh, let me know if that works.

Speaker B: Yeah. Uh, reframing it, repackaging it.

Speaker A: Yeah, because. Yeah, exactly. Um, but then, yeah, in terms of doing that, you really need to basically get the buy in. Um, and then you need to give your. And then it's like, check back it in three months. And then it's like, I've actually made a lot of mistakes with, uh, that. It's like there's a lot of parts where it's like. Oh. Like actually, um, we had one where, uh, it was like the cgm, uh, uh, company was like, really mad at us for. For an ad we did. And, uh, I was like, okay. And then we had like, the whole, um, ASA complaint as well because of the. Because we didn't put in the ad and no one ever did before of, like, uh, Stephen, uh, Barley is an investor. Uh, and then, so we got slapped on the wrist for that as well.

Speaker B: That was quite a big deal, actually, wasn't.

Speaker A: Yeah, Fuel and Zoe. Yeah, were both, uh. But no ads had ever done it before either. And it was very much something that was never really hidden. It was like he spoke about it very publicly, as well. Um, so anyways, yeah, there's that. Ah. Um, what else? Also, it's like, I got a lot of slack for, um, the CGM because it was really hard to pivot away from that. And it's like, oh, Zoe's not just cgm.

Speaker B: Yes. Yeah.

Speaker A: So that was like another, like. That was like, the downside of, like, oh, when that, like, you know, edgy performance angle does work, then it's like, actually really hard for the performance team m to take it down. But those situations are rare, so it's like, it's not. It's not worth it.

Speaker B: But yeah, uh, my. My. Perspective on that is basically like the. What works in terms of acquisition, what hooks people in, isn't necessarily like, what someone buys as well. Like, you want to. You want to maximize your ctr, thumbstop rate. All those things drive cheapest traffic, which it sounds like that's exactly what you did.

Speaker A: Yeah.

Speaker B: And then really it's about like, I don't know. In my mind, the brand team's job is to really make sure that that customer has. Has the maximum value possible to the brand of the company. And therefore it's about sort of telling the story and buying the right products. But there's Lots of our clients that basically like, will come in on one product or one angle and then go and browse and buy something else. Like, yeah, what's your PV on that? And like that acquisition message versus like the retention message and everything that happens after that. I guess.

Speaker A: Yeah. So I mean that's probably one of the, the big, I don't know if it's unspoken, but things that aren't focused on enough, which is like the offer, the landing page, kind of like these big other levers that like creative is a massive one. But there's these two other ones that like people really like don't spend enough time on and they're not as synergized with the rest of the performance team, I'd say generally, uh, maybe it's like a CRO person or CRO agency, whatever. Um, so yeah, that part is hard to bring together. And then in terms of like what's Brand's job is kind of, if I summarize your question, um, is really, I mean, so Zoe was like, uh, like a really good brand and it was really good to see how that kind of operated. Um, you know, they had the top branded podcast in the world. They cracked that. They um, were actually incredible at PR basically. Uh, anything that they did was like prepackaged and ready for a journalist to talk about every single week. Um, and it was like actually Brand was really good at these like other channels. Um, that was not performance. So that's why like it worked kind of so well because you had both of these firing at like top tier levels. Um, and for Zoe it was really like at ah, the time it was like one product. Um, so there wasn't much like browsing, going to something else. Um, so we didn't really do as much of that as really like we had the quiz, the funnel and it was like what was the cost per purchase or subscription or whatever or trial. Um, and uh, yeah, and Brand, like what we saw honestly was just like it would decrease your CPA and increase your conversion rate ultimately. Because it's like when we then did launch in the U.S. it's like, oh, see, obviously U.S. is more expensive market. But it was also like disproportionate to the CPMs that we were paying, uh, on the traffic there. So it's like, oh, we didn't have as much brand presence there. You could really see the impact of it. Uh, whereas Zoe had such strong brand presence. It would like really reduce cpa.

Speaker B: Yeah, completely. I see that in our ad accounts, but anyone that's running like any above the line activity, any, anything that like major brand activity wise has a, like a huge impact in terms of cpa, uh, efficiency. I was talking to Huell not too long ago and they were having the same, well, it's the same conversation. It's like they've got this whole brand legacy piece. They're UK founded, their British brand. Scaling in the UK was relatively straightforward compared to scaling in the us which was eventually much more difficult. Yeah, I think they're still sort of figuring that out now. Um, I guess in your mind, how much should a brand be spending on brand metrics? And again, maybe specifically in the ad accounts there's stuff like ad recall, reach, et cetera, versus some of that just lower funnel tactical type.

Speaker A: Yeah, uh, it's probably one of the hardest things because it's like the typical journey is it's like cool, you might spend 100k and then it's like, oh, you don't see anything. And then it's like, well that was a waste. Uh, and uh, this happens all the time. And the reason is so hard is because there's no good KPIs for brand in comparison to performance anyways. You can get stuff, but it's really hard to have a fast feedback loop in brand, uh, which makes it really hard to hire someone. And basically your turn, uh, to find someone good is like minimum, like six months usually for like, uh, whereas a good performance person you can usually tell in the first month. So you could basically get six performance people in this in the same time that you do one brand person and then they have immeasurable impact on the brand, which is always really hard to tell. Um, so honestly I have not seen that many good uh, brand people because you just don't get the same feedback loop. Um, so I actually don't think I can give like a good answer on this, uh, because it's just so difficult. And like Zoe, we're really uniquely positioned. Unlike they were doing leading science, which like journalists love to talk about. Uh, they crushed it on the podcast. They were early on that front, which had actually like a huge impact. I think there was one stat, uh, Rich, uh, Willen who kind of like did was the whole behind the scenes of it. Um, he, I, he, I told me the stat of like, I think at some point maybe like 50% of all purchasers had listened to the podcast, like something crazy like that. And there's like this cool pixel technology. Actually I can't remember the tool of like, you can sort of track, uh, podcast for purchases it was really cool. Yeah, I think I told you. Yeah.

Speaker B: Because I think Stephen was doing this for the first time. Wasn't it his podcast?

Speaker A: Yeah, yeah.

Speaker B: Ah, yeah, I remember.

Speaker A: Um, so anyways, yeah, that, that was really interesting. But then it kind of like, I don't know, analytics, uh, or like attribution I think could be like a big black hole is like, doesn't give you the answers you actually thought it would. Um, but I don't know. Yeah, it doesn't really answer your brand question though. Yes. Really, it's great.

Speaker B: I mean, yeah, some interesting stuff there, I guess. Like from our perspective, like in an ad account we might spend 10% of the overall budget on like ad ad recall for example. And then obviously look at like the, the lift or impact on cpa. So like we always look at it in the context performance market, but obviously we're very much a performance agency. But I'm guessing from a brand perspective, different kind of POV M. Um, I'm curious about attribution. Yeah. How, how much time did you spend on that, Zoe? Maybe some of your other experiences.

Speaker A: Yeah, yeah. Uh, I have I think two really good examples for this that are, uh, so Zoe was originally basically like a data science company. So all the attribution was actually built in house. Uh, and it was actually really good, really good in the sense that it was good enough for me to take action on, uh, which is ultimately like every analytics is like if it doesn't tell you something on like what you should take action on, it's a waste of time. Uh, so Zoe was actually really good at that and basically had a good idea of like how long it would take roughly for someone to convert. We basically had all the different touch points utms, uh, that were attributed to a single person, um, throughout their journey, pre and post purchase. Um, and we can then calculate basically how long it would take, uh, and how many roughly touch points, uh, which doesn't actually matter that much in terms of performance marketing, uh, in terms of what action you can take. Uh, and then you basically at the end of the day always end up with a multiple.

Speaker B: Yeah, yeah, yeah, yeah.

Speaker A: Which is like just get to the multiple that's as close as possible and then like check back in like a month or two. Whereas Zoe, I had the luxury of like I can basically check every week, uh, but it didn't change that much, which is like, okay, cool. Uh, and you essentially have like your, your, your top line baseline and your performance baseline and you would essentially have a multiple difference between the two and you can do for anyone listening, you can basically do like easily a GEO holdout in a country if you're international enough. And that's like the easiest best way. So it's like, okay, we don't spend ads for a month in Canada. Uh, what is our organic baseline? Let's turn it back on. Check the pre and post and then you can get a multiple and then do it again in two or three months and see if the multiple changes a lot. And just operate on that and make sure cash flow and everything is good. It's like really, it's like, it's that simple. It's like just. And I think the multiple at the time was maybe like 1.7 or something like that. Uh, and you would know enough. And like literally all of my dashboards would have a lot of that baked in. I would have the original, the Facebook one and the multiple version, and I would just operate on that, make decisions like that.

Speaker B: Super interesting. And then you would reverse engineer that back down into your CAC that you're aiming for.

Speaker A: Yeah, exactly.

Speaker B: Super interesting. Okay, and how much did you look at, like just P and L or like marketing efficiency, you know, mar and stuff like that? Or was it sort of. Was that your kind of hero?

Speaker A: That was the hero one.

Speaker B: Interesting.

Speaker A: Uh, because it was really hard otherwise. Yeah, yeah, yeah, yeah. And it's like, again, it can be a black hole. Um, and another kind of like, good example is more on like the speechify side, which is like, actually their CFO is their main media buyer, which is crazy because I haven't really seen any other company operate like that. But it actually makes the most sense because they're the ones that are closest to the financial numbers and can really actually see the impact of a holdout in a country because they know all the numbers off the back of their hand.

Speaker B: Like on a daily, they can be looking at that stuff coming in and going, oh, that's interesting.

Speaker A: Yeah, you don't need to. Weekly is like at most where you can kind of see impact. Uh, but yeah, so he was kind of like the main, uh, person on spend decisions, uh, in terms of country allocation and things like that. And I was like, oh. And it was the same thing. It's like you calculate a multiple and you move on. So it's like really like it ends up at the same thing. But, uh, they were able to operate on like a much faster, uh, like timeline because it was like the CFO actually making the decisions.

Speaker B: Because that's one of the biggest things I think brands get Stuck into. I mean, again, when I spoke to the AG one, um, eximo as you want, he was saying the same thing. It's like he spent so much time managing internal stakeholders and trying to, like, um, you know, tell them how much a channel was performing or, you know, something like that is it. It does feel like having a, um, metric that is just the voice of truth. It, like, gives everyone clarity, and then you can unlock, spend, and you can make decisions fast and you can move at speed, which is ultimately what scales you.

Speaker A: Right, yeah.

Speaker B: Um, that would be my perception of it.

Speaker A: Yeah.

Speaker B: What was it like making ads with Stephen Bartlett? How did that look like?

Speaker A: I mean, it's pretty fun. He's a good marketer, so it was actually very easy. Um, and yeah, uh, it was really funny because the whole brand shoot was like a bunch of cameras like this. Uh, and then it's just me on my iPhone. Uh, and it was like such a contrast. But it's like actually the performance ads we're probably seeing, you know, 20x times more than any other brand content.

Speaker B: I was gonna say the ROI of that iPhone.

Speaker A: Yeah, it might have been the same iPhone, actually. I've had the same one for a while. Um, but, yeah, he was really good. Could, uh, do most things in like, one take, one or two takes max. Um, and I've worked with a lot of people. I've made thousands of ads. Um, so it's like you can tell in the first, like, I'd say minute or two, like, how often you'll need to retake something. And that's actually like, one of the biggest, like, other tips I would say, is make sure your performance person, the one who actually, like, looks at the data in Facebook Ads Manager or whatever analytics tool, um, is on set. Because they're the mini Facebook, they're the pre Facebook of, like, you want someone to basically have a qc, like a quality check. Uh, it's like, is this good enough to pass my benchmark? Because you never know what will perform on Facebook. Like, nobody can predict that really, but you can know when an ad is bad. And it's like, you should absolutely retake that. Retake that shot. Sorry. Um, so that I was that role basically, and it was really helpful. And it's like when there was an actor who wasn't as good, it's like, you know, we'd have to retake like three or four times until we got the shot. And then it's like that your success rate of finding a winner from those shoots was actually, like, way higher. Um, so that's like the one tip is like, whoever's looking at a lot of Facebook ads and a lot of competitor ads and a lot of research is I make sure they're actually on set.

Speaker B: Yeah, yeah. So I'm almost positive brands will not be doing that. But that's a really good tip because, yeah, uh, literally bringing the ad account to the creative rather. Because you always have this disconnect between. Well, when I speak to brands that aren't, you know, working with an agency like ours, for example, they're doing creative normally in this kind of ad hoc way, working with lots of different partners. Um, and there is this like siloed effect because they're not connected to the outcomes and the data and therefore they're not producing with that in mind. So like, I can see a lot of brands falling into that trap of, um, yeah, not having all the contextual understanding of why and how and what, which is really important. I think when you're making ads, you

Speaker A: train a mini ML system in your brain of what ads work, what doesn't because you've seen so many.

Speaker B: Exactly.

Speaker A: And I was actually talking to um, Ben yesterday, uh, the founder of Mob, uh, and I was like, I feel like the best performance people, uh, because we were having this debate about media buyer versus uh, creative strategists and kind of how they're separate rules. But I'm, I'm really like actually like the best performance people actually kind of have both. Because if you've been a media buyer or creative strategist on either end, you will have had to dip into the other side.

Speaker B: Totally.

Speaker A: It's like, so if you're a media buyer, it's like, oh shit, my account's not doing well. Like, I need to make a better ad. It's like you jump in and it's like creative strategists like, oh, why my ads not getting tested properly? It's, you know, 100%. So it's like you have to have this, like anyone who's good enough will do that jump and that crossover. Um, whereas like, I think a lot of people make the mistake of like hiring like a very specialty role. Um, but again it's, it's much harder to find those people who do both and it's a lot harder to like scale those teams. So there's trade offs with both.

Speaker B: But my perception would be that it's actually much more difficult to find great creative strategists. If I was hiring from my agency, I'd probably prioritize the creative strategy skill Set and then upskill on media buying, particularly because media buying has become a lot more simple and automated.

Speaker A: Yeah.

Speaker B: Um, would you, would you agree with that?

Speaker A: That, I mean that's exactly what I did as Zoe is like, I'd, I'd find the people who are good at organic convert to paid and teach them media buying interest is this like it's so much easier if you have good fundamentals in the account and you have a good cadence. Uh, so it's like you can launch a lot of ads. You have a good testing set up and a good scaling setup. That's like all you need really. Um, so yeah, honestly. Yeah. Ah. And the creative is like, it's really hard to teach. Uh, you need like some innate, I

Speaker B: think of it like taste is quite difficult to. Well, almost impossible to train, isn't it? Or like that kind of gut on the creative side is hard.

Speaker A: Yeah, yeah, yeah, exactly.

Speaker B: Yeah. So, so what did it look like, I guess moving from Zoe to Speechify? Like what, what were the different things that you saw there versus your time at Zoe out of interest?

Speaker A: I mean Speechify is operated like so differently. Yeah, yeah, yeah, yeah. Ah. Uh, but these guys are like, they're also killers in their own way. Uh, such a, such a cool company.

Speaker B: Um, what is it for anyone who doesn't know?

Speaker A: Oh, it's basically, um, essentially it's like text to speech. But any piece of text that you have, it will transform it essentially into an audiobook for you. So anyone who's like either dyslexic ADHD or prefers to actually listen to stuff on the go or before bed or anything is really good at that use case. And they've really like conquered the B2C market, um, on voices. So, um. And yeah, so it's like an app product and Zoe was like half app, half D2C. It was like very unique combo actually. But yeah, speech fight. Um, I did more freelance. I'm basically more on like the creative side and some of the media buying side on their operations. Um, so they were one of the first actually ad managed customers because they're, because we, we took them basically from doing, I don't know like 50 ads a month to like 2,000.

Speaker B: So um, that jumping system must have had an incredible impact.

Speaker A: Yeah, yeah. And I was only like a piece of that. I didn't do the whole thing. Uh, just to clarify. Um, and obviously there's a team that now runs that and like I'm not involved uh, anymore. But um, at the time it was basically like I would Go with Cliff, who's the founder. And we would basically do the same kind of thing as Zoe. It's like we do these shoot days, but instead of doing like 150, 200 videos, we do like 5 or 700 in a day. Uh, and he was. He's like a really good, like, he's really good at improv. Uh, and, uh, his. His sister as well. Geffen. Really good as well. And like, some of the top ads are still some of the ads that I filmed, uh, from back then.

Speaker B: And so were you just like turn up this house, like, no scripts and just as you run around shooting?

Speaker A: No, no, we, we, um, We. We wouldn't really have scripts because I know they were really good at improv and they could basically pitch the product every single time, like without fail. So it's more actually finding really good inspiration, like winning inspiration. Um, and this is when we actually use like, ad scan to basically find like 150 different winning ads of, uh, just totally different verticals. And would just like, like I could. I look at an ad, I'd be like, okay, I could see how I could use that for speech fire. And we just have 150 of those. And each of those videos would turn into like five or 10 videos, basically, because you improv is slightly different every time. Um, and you just go like back to back to back to back. Uh, like insane volume.

Speaker B: Wow.

Speaker A: So, um, and it's like basically just like a canva board of like a bunch of ad inspo and like the one liner that could. You could take from that to turn into speechify.

Speaker B: So what, like, what does post production look like after a shit like that as well?

Speaker A: Like.

Speaker B: Yeah, do you just literally great, like, chop and change it all together or like, is it organized? Like, how do you do it?

Speaker A: Yeah. So at the time it was like I would, uh, label every single video. It was really long. Now this. I can do that. Uh, at the time it was like too expensive or not good enough. Um, and, um, yeah, basically I would like, label all the different, like, files from my phone on my laptop and then put into like a Google Drive. And they had. We like, we helped them hire like a team of like 20 editors. Because it was like, that was actually the bottleneck.

Speaker B: Yeah.

Speaker A: Yeah. So it's like just hiring like a bunch of video editors. And it was like. I remember it was like we did this like, round interview of like 30 people, then maybe like five states. And you did that a month later. A month later, then you get like 15 people. Um, and that was actually like the massive unlock to get a lot of volume was just like, because you can rehash a video a lot of times. Um, and I know there's the whole talk about like, oh, it has to be really different.

Speaker B: Yeah.

Speaker A: And I keep asking people, I'm like, how different though? Like, how different does it really have to be?

Speaker B: Yeah.

Speaker A: Uh, and it's like cool if you just change the intro hook or the like visually, if you can make it look different, I'm pretty sure it'll pass. Uh, although now I'm like, does the messaging get combined? But then it's like the messaging is kind of always very similar because it's the same product.

Speaker B: There's only so many ways you can

Speaker A: re frame the messaging and you can have different like Personas or audiences basically that you're targeting and different characters that you're using. Um, but anyways, yeah, so you like video editors was the big bottleneck and then, you know, help solve that part. Um, and then, then they were able to do like thousands and thousands of ads and were able to get like so many winners and scale so much. So. Yeah.

Speaker B: So it's actually all in systemizing the, the content you got from the shoot, like organizing in a way, packaging in a way that enabled them to.

Speaker A: Yeah, exactly.

Speaker B: Stuff out fast. Super interesting. It's very similar to what we do actually on this piece around creative diversity. I guess it relates back to how you were thinking about coming up with concepts. Like, we think about like iterating and looking at competitors is obviously a great source of inspiration. But ultimately like, you're going to be churning out a lot of the same stuff that everyone else is doing and therefore people are seeing the same types of frameworks. Like, how do you balance, I guess, like iterating versus innovating, like coming up with new stuff. And um, like, how do you kind

Speaker A: of see that impact the net new versus iteration debate? Yeah, um, I, uh, actually think like when I was at Zoe, it would swing a lot where it was kind of like a necessity almost of like, oh, we've had a winner. Like, we've just had a winner. Let's iterate and like juice everything out of it versus like, okay, now it's been two months. It's like, we need a new winner. Yeah, it's like, let's go back to like 80% net new until we find something and then it goes back to like 80% iterations. Um, so that was my system back then, which in retrospect is probably not great because you're very like reactive still. Um, and you would have still like a creative roadmap of like, you know, I want to try to get all these different types of things done. But like that didn't work that well. And then it's like again, it's like, I don't know, a lot of the ads I had made that were winners, a lot of it just came from improv. It was like on set, I'm like, oh, wait, actually like, what if we just try this? And it's like that would work and that was never in the script. Right. Um, so I think you just have to force yourself to make new content. Even if it's an iteration, if you film it new again, you can find new ways to do it. And I think it's like you just have to get in the habit of like just producing a lot. Um, and then you'll naturally find variations and like new concepts from that.

Speaker B: And is that visually or is that like messaging in terms of differences? Like, is it mostly like filming in a different setting with a different creator or is it like different script with the same hook? Like, how are you viewing that?

Speaker A: Yeah, so every house or actor would be different every time. So you already have that differences, which is enough for Facebook. Uh, but the scripts would be similar because you kind of know what works in terms of like the body of the ad. Uh, the hook would be very different every time though. Uh, but you would still like every shoot. It's like cool, we're filming our top 10 winners again, like every time. Because it's like you never know, you might find like a variation of it. Uh, and like that was like a non negotiable. Uh, and then the rest of the time you would try to get new ads or it's like you would splice things in, whatever. And you could do a lot on the posts, production side, on the editing. Um, but yeah, it's like you always got to refilm your winners.

Speaker B: Yeah, yeah, yeah.

Speaker A: Because it's like you might find a slightly different way or it might get inspiration on set to be like, oh, actually this. So that's like. I think actually I don't know if anyone. I haven't heard anyone really talk about that. I just realized I was like, oh yeah, that's like so obvious. But maybe a lot of people don't do that. I don't know.

Speaker B: No, no, I m mean it's interesting. I mean for us as an agency, we focus a lot on new ideas because I think like we're always trying to. We're like optimizing towards finding unicorns, basically. Uh, but then actually I think we probably could do more on iterating on like best performance because in theory is much easier. Much lower hanging.

Speaker A: Yeah.

Speaker B: Uh, um, but it's actually like systemizing it I think is actually the more difficult thing.

Speaker A: Yeah.

Speaker B: Um, so it's interesting to hear about.

Speaker A: Yeah. I'd say if you start any, like, role in the performance world, it's like try to just make an iteration of the top winner because it's your highest chance of success. Yeah, yeah, yeah. And like, just do a slight tweak on it, like enough that it's like, you know, passes the Facebook test. Uh, and it's like, maybe you target a different Persona, but it's like the same frame by frame. And also it's like people don't copy enough like frame by frame. Uh, and they don't think enough about like, uh, frame one. Like those 1080 by 19, 20 pixels is like everything. Uh, it's like what's on screen at 000, basically. Exactly. Yes. And it's kind of like, you know, a lot of people already know this on YouTube because YouTube has this whole thumbnail system. But like, nobody really does it on Facebook. Uh, actually, no, I say like a few people actually really focus on it. But in terms of pure video, it's not really, uh, talked about as much. And you can actually set a different Facebook frame. But I actually like to think about it more just like, what's the first frame of the video? Uh, and every shot I set up with my phone, I'm like, is this good? Like with this? You know, I'm like, I have the QC in my head. I'm like, is this good enough? I love that. So, yeah, first frame, so important.

Speaker B: I totally agree with that as well. It's so funny. Like, when I first started, that was all I used to talk to my team about was 00, 0.

Speaker A: Yeah.

Speaker B: And we actually used to do custom thumbnails on Facebook. Yeah, I don't see as much now because I think it like, we haven't seen that correlation, I guess, with performance. Like, we probably do the same as you said. Now we just prioritize what's on screen rather than like doing a custom thumbnail like YouTube.

Speaker A: But. But I think, I mean, I see a lot of people. We. We actually have a feature for this at uh, Ad Manager because like, we've had this requested a lot. And if, you know, you have like a good first frame that works in a video, you can actually increase Your chance of success on your other videos because you've taken a proven first frame.

Speaker B: Yeah.

Speaker A: So it's actually like reusing. It is actually really powerful and you can give an artificial boost basically, uh, to your video. Um, so I remember Marin, uh, he asked us, uh, for this actually. Uh, and I was like, all right, you got it. We just shipped it in like the same week. So.

Speaker B: Yeah, I love that. Definitely hear many people talking about that, for sure. I guess I'm curious also about your. So, like, we found a winner. It's a unicorn. We want to iterate on it. What do you find? Like, the win rate is off the iterations. Like, actually, is it. Do you know what I mean? Is it a case of. Because we tend to find that once we found this unicorn, it's actually quite difficult to get the same level of performance again. Um, and again, maybe this is because we need to think more about systemizing this. But, um. Yeah, I'm curious, I guess on your experience there and what you found.

Speaker A: Yeah, it's a great question. It's honestly like really hit or miss.

Speaker B: That's exactly what you find.

Speaker A: Yeah, yeah. Is what I've personally experienced and seen. Um, and I think it's the wrong way to think about, like, you want to replace the winner. It's more like you want to actually supplement the ad account. That's like, it's. It's quite. It can be quite dangerous to just think about, like this one ad is the whole ad account, whereas you actually have like a thousand other ads in play. Uh, and you actually want to look at like total spend of the account and where is it distributing to the ads. And if some of those iterations are even getting like 5% of the spend, like, that's great because you already know that winning ad is good. And if it has a good CPA or cost per result, whatever metric, um, like that. That's like a good. It's not a unicorn winner, but it's a good win for the account. Um, and so I think it's really hard to replace a true winner. Um, I've seen it a handful of times. Um, and it honestly comes more from like, I'd say the common journey I've seen is like, you see a good organic video, turn it into a paid ad, does really well, then you really performance the shit out of it, and then it goes really big and you really engineer it in that way. And I've seen that work. I've seen that happen a handful of times. But it usually comes from that organic inspector and it's like you slowly convert it more into a doctor ad.

Speaker B: Interesting. I guess we think about it in terms of the components of the winner as well. Right. It's like, who was the creator? What was the core message? What was the core theme? Was there a core Persona that we went after? Was there something visual that we did? Do you also break it down into its raw parts or do you see it holistically as a winner and you kind of see it as a concept that you need to replicate?

Speaker A: Um, both. I've seen the components. This is actually a really popular winning ad format is literally just stitch three of your top ads to together.

Speaker B: Yeah.

Speaker A: And you'd be surprised the amount of times that works. I'm like, I'm not even joking. Um, so I think there's. There's definitely some truth to that. And then reusing. But you already do that in all your videos. It's like all the body is usually kind of the same.

Speaker B: Yeah. Yeah.

Speaker A: So it's like you already do this natively when you make the ads. So I'd say you're. You're already doing that. Um, and then. Yeah. In terms of like, tweaking it slightly. Um, it's always worth doing because it's so easy to do. Uh, it's so low lift. Exactly. For the. Your EV is exceptionally high on that type of system. Uh, so you should just do it anyways. Even if it doesn't do anything, it'll probably at least have one winner. Uh, that will spend a decent amount. That's worth it.

Speaker B: Okay, cool. And then I guess wrapping all of that now into the world of Andromeda. We, uh, are viewing things quite differently in terms of this idea around winners. We're seeing like, full funnel creative units are basically working better.

Speaker A: A unit. Yeah.

Speaker B: Do you know what I mean? So it's like you're using creator stuff. Maybe you're mid to upper funnel using different types of assets for. You're kind of like treating all different levels of awareness in one package. How are you? Well, that's how we're seeing anyway, what we see working. How are you thinking about ads now in the context of Andromeda? Any of the stuff you've just talked about change or like, has it changed in the context of that update over the past 12 months?

Speaker A: I think Andromeda has just made it more clear to people what always worked.

Speaker B: Yeah.

Speaker A: Like if you're good enough at that time. And so you already knew that. So I like, I actually don't think, uh. And draw. I think it just made it more public. It's almost like the education piece basically like that that's what Facebook did there. Uh, so obviously there's some now it's actually gem and there's probably a new one now. Uh, and like you know they have these updates every year but I, if you were good at performance, you were always doing that anyways. Um, I think those principles. Yeah, exactly. And I think like what's actually the shift that I'm seeing now is a lot more on like the AI ads. And you see it basically like there's a trend. It's like the affiliate and lead gen marketers are usually first to experiment like a new tech and we'll figure it out and we'll trickle down to apps, then dtc, then maybe big brands eventually. Um, and so uh, at admanage we have all kinds of different people. Anyone who launches a lot of ads basically uses a system like ad manage. Um, and it's really interesting to see those ones who are on the leading edge, the bleeding edge almost I'd say. Um, and I'm seeing a lot more like AI ads like actually work.

Speaker B: What type of AI ads that have interest.

Speaker A: So both like uh, the fake real ugc, whatever it's called, uh, and the Claymation using Suno to actually uh, make a music video that has insane engagement and if you can tweak it, engineer it in a way uh, that will work really well. Um, and then you can also just like non obvious ones is like AI B roll, uh, is one of them uh, iterations, statics. Honestly it's like proliferating through the ad space and it's like coming whether you like it or not. Um, and, but it's like um, I'm starting to see it win more and more but right now it's working for people who don't care about brand. So it's like it's like this really thin line.

Speaker B: Yeah.

Speaker A: Uh, and eventually it will work for people who care a lot about brand and it's just like a matter of time until the tech gets better. You know it's a new video model every three months that will improve it. Um, and yeah, it's literally just a matter of time. And I think like in the future it'll be a mix of like you have your real content and your AI content. So like Speechi5 for example, they do a thousand AI ads a day.

Speaker B: Wow.

Speaker A: A day. A day. Oh my God. Um, how, how is that possible? Uh, they, well they, they had, they hired like this whole B team basically to just tackle this.

Speaker B: Wow.

Speaker A: Like we're gonna do a thousand AI ads a day. Yeah, because it's like we don't want to shift focus from the main team. So it's like we'll just hire a separate team to do this and try to beat the main team. So you have this internal competition essentially. Um, and yeah, so they do a thousand AI as a day. They launch like. Yeah, everything through Ad Manage is actually like insane volume. Um, and yeah, it's so, it's so cool to see. But they're like, they're very experimental and like willing to do kind of what it takes to get to the next performance level. So. But yeah, it's coming. Um, and it's very hard to set up and do right now. I'd say do well. It's like, takes so much context to get a really good AI ad. Um, and it still requires at this moment like a lot of editing still. You still need those 20 video editors if you want to do like skills, scale. Ah. So it's just like imagine it's like it's just AI footage inside your Google Drive.

Speaker B: So yeah, interesting. I mean, definitely there's all sorts of stuff happening with AI that I find really interesting. The stuff that we see performing consistently is still the human to human stuff. And then at some point there will be this convergence because it is like getting there. There's some specific ad types that I think are really cool. This kind of cartoony style is also.

Speaker A: I see. Yeah. The claymation. If you just look up like what are the 20 most popular animation styles? It's like just run all of those. Yeah, yeah, yeah.

Speaker B: That's. That's actually a really powerful unlock because those would have taken like an amount of time that wouldn't have made it a feasible.

Speaker A: You would have never done it. Yeah, exactly. So you're not Pixar.

Speaker B: Totally.

Speaker A: Yeah.

Speaker B: Ad, um, Manage. You obviously have visibility in a lot of different ad accounts and what you're doing is obviously helping brands deploy ads a lot faster. And unlocking creative as a bottleneck is my perception. Um, and I guess in all the stuff that you've learned across all the brands that there's like loads of really interesting nuggets there. But like, if you were to boil it down in terms of, across the data you see at Ad Manage from your experience historically, what are maybe the three key things, key levers that brands want to think, uh, about or need to be thinking about today to really drive success in their performance?

Speaker A: Three things. Okay. Um, so one is creative volume is actually very important. The data heavily suggests whether it's correlation or causation. The more ads you have and do, the more you spend generally. And people generally will not spend unless it's profitable or it's like to some allowable degree. So it's like that is very clear in the data that we see. Um, so that one is still like fundamental. Um, so that's one. It's like creative volume is still really important because it ultimately comes down to how many winning ads can you find. Um, so tip number two is like, yeah, people don't do enough on the offers and the landing page and the CRO experience. Um, it's like you know, an unlock there of like even 5, 10% can then like what you will see normally is it will. You'll have the same cpa. You'll just be able to spend more. It's like you can take the other trade off of like actually want more efficiency. But then it's like when you start going the restricting Facebook more, it ends up worse a lot of the time and it's like, oh, I'm in a worse spot than if I just left it. Uh, so that's tip number two. It's like people don't do enough of that. And, and then it's like actually KPIs for the team is like, is so not. It's very common but it's like not well done I would say in a lot of places. Uh, and it's like you need to keep it like stupid simple. Yeah, that is like anyone can understand, anyone can do it. Everyone has visibility and make uh, sure like your creative strategists are in like Ads Manager or ad Manager, um, to make sure that they understand like what Facebook is deploying money to the North Star.

Speaker B: Everybody is the same North Star.

Speaker A: But it's really hard to enforce I would say. Um, and what I've found, that's why I had initially done the full stack because it's on your responsibility to make a winning ad. And it's like if you want to make a winning ad, you need to know the data to then re input into the start of the ideation to know what works. Um, so it's like you make that a forcing function. Um, so to recap, it's like creative volume is actually still massively important. Uh, and diversify via like Persona ad formats. Uh, we didn't even talk about ad formats. But yeah, it's such a massive one. That is super underrated too. Uh, like offers and landing page is absolutely massive. And then really good, like, hiring. And KPIs is massive, so. Yeah.

Speaker B: Amazing. That's awesome. Yeah. This has been great, great conversation you've had, lots of wisdom to, uh, impart on me and, uh, the audience. So thank you so much.

Speaker A: Yeah, of course. Great to be here as well. Thanks, Ryan.

Speaker B: Cheers, Ralph.