Meta AI Creative Strategy | What’s Working Now (Secrets From $100M+ Spend This Quarter)

21:09Olly:i think if i look at like sales conversations when we're coming into contact with new businesses over the last three months the problem it's always easier to do more than it is to do better and I think if I was looking over the last three to six months for new clients who were having those combos with them like are we a good fit the majority of them are treating problems by throwing more crap at the wall and not going deep enough on the quality yeah volume I still believe in creative volume but it has to be done in a way that doesn't create a decrease or diminishing quality over time which is hard but that's linked to the other points you just touched on yeah that makes a lot of sense and then it's about just like how we how do we treat that i guess

21:54Loukas:yeah i think that yeah the difficulty is like this you're going to have so many findings from from doing and that's just one part you've then got the whole research phase that like we look that's the visible data like the research is the the hidden data that we we don't look at but i think a simple framework with with any strategy is like start with the outcome so what what we're to achieve over the next three months as a result you know maybe that's we want to um have x amount of revenue growth or increase spend by 20 in three months at this cap target um then based on that outcome like what is the binding constraint that is if we change that one thing that is going to give us the best chance of achieving that outcome and then that's where from that point we build the

22:38Olly:strategy that solves that binding constraint yeah makes sense i want to i want to just speak to two ways that we can treat that i think let's leave research out because research i think is like message how do we find data that on customers that gives us an idea of like what message we should we should touch on i think we've just we covered at this very start quite a few different like formats um i want to touch on format again um because i think one of the big things you see in ad accounts especially those maybe those brands watching that are like sub 10 mil maybe even sub 20 mil is like they do statics and ugc and nothing else really quick one before we get back into the episode every business spends money to make money ads software flights travel the hundred unavoidable little things to keep the lights on.

23:30Olly:All that spending every month and none of it comes back to you. It just goes out the door. Incard is a financial platform built for digital businesses and it flips that on its head. They're giving 2 % cash back for your first year on ads, SaaS, travel and everyday business spend and it's all uncapped. The same money you were spending anyway except now a slice of it lands back in your account. You also get multi-currency accounts corporate cards and shopify meta stripe and zero connected in one place a true source of truth but honestly you're spending regardless so you may as well earn on it sign up for free use the code swu2k26 links in the description now back to the episode or they'll maybe do statics ugc and then founders ads but there are loads of other formats big swings as we like to call them internally where we're trying to take a reference point from maybe like a youtube or a um another brand or organic and pull it into a dr paid context and um we've seen an example of this recently i'll give you an example to articulate to to so you can envision this better we've recently for a pet brand that we work with found a comedy angle where we set up a like a comedy sketch where uh or comedy sketch sketch scene and then allowed the brands to address like awkward pain points or funnier pain points um without being salesy or preachy so like and what i think what that's a good example of if it is like picking the right vehicle for the swing because it matches the emotional tone of the asset.

25:16Loukas:Yeah.

25:16Olly:And therefore it doesn't appear like an ad. So when we talk about big swings and we talk about high prod, I think if I was to say high prod to like most of the audience, the two things that will come to mind is always founders ads and podcast. It's like they're easy to do, but I think what the issue when you try and do podcast and don't think about the emotional resonance and the emotional tone of the concept and whether that matches a podcast is you end up creating a load of slop assets that don't work.

25:44Loukas:Yeah.

25:45Olly:I think that's a really important thing to think about when you think in big swing, like based on your audience and what you're trying to convey as a message, how do we find a format that will allow us to do that? And it has a high probability of working.

25:58Loukas:Yeah, I've got a good example of that as well. like um we were selling to a kind of gen z into millennial audience like the biggest opportunity was uh 25 to 34 so we were looking at organic as a form of research to uncover the types of things that they are consuming uh that maybe could translate into a new big swing format and one of the biggest things that was coming up for that market when we were looking at consumable so something that you drink or take um was this organic trend that you might have seen where you've kind of got uh someone a load of cardboard and it's facing the person the drinking they're drinking they're drinking something from and they have to guess what drink is that could be like a load of coffees and it's someone that really knows their coffee so it's like starbucks

26:53Olly:and costa and the funny one that everyone should go and look because it's hilarious is the italian bloke that smokes 10 cigs and then names every single brand it's like cig connoisseur it's so funny that's hilarious like a six-year-old bloke who just loves loves cigarettes yeah that's so

27:12Loukas:funny but like that so that is an interesting insight right because like you can translate that into like an us versus them format and i saw then off the back of that there was and someone shared this it's like really great example of taking that format which was like something we've seen organically on tiktok for probably a year or two now it's something that i've always been aware of um but looking at okay how can you translate that into a format that's maybe working within an ad account like an us versus them there was this really good example like i can share a bit of it on the screen just so you see the framing where you've got a girl on the left a guy on the rye and cardboard in the middle of them so they can't see each other um and essentially it was like guess that item so the girl would be like um she was the them and he was the us so she basically was like i'm 1300 milligrams of electrolytes and she was holding up like a 12 pack of gatorade and then the guy's like i'm 1300 milligrams of electrolytes and is holding up one sachet of um of of an electrolyte and then they kept doing that through like you know in terms of comparing the usps of of each but it was just a great visualization way to like take something like an ass versus them but do like a big big swing variation of it another example i've seen

28:31Olly:of going through that same process recently is there's a tiktok trend going around that's like five things that are actually worth the money and the person usually it'll be like someone within a certain demo maybe like a lot of the couple that i saw was like quite like menswear fashion forward males who were who were doing it within a fashion context they were listing like certain shoes certain watches trousers and i think that was a format that i think could i haven't we haven't done this but that was one that i look at i'm like really really framework driven really easy to weave in a product could do it in loads of different niches and categories and it's quite an easy narrative to put together it's like 30 second video yeah um i think tiktok and organic is so good for finding those for for the younger demos and then looking a little bit broader for for the older demos um another one which i actually wanted to talk about is just like science and data back creative um i just think that experts in in videos is just like continuing to crush and probably i'd say probably perform like aggregate level like becoming more and more meaningful in terms of spend level yeah um whether that's like dermatologists whether that's doctors whether that's experts that like aren't an expert in the traditional sense but have authority with the audience so um let me think of an example of that someone who works in a dental clinic and has been in a denticle clinic but isn't the dentist and has just finished.

30:13Olly:I saw an ad that was a dentist where she'd finished it. She clearly worked on reception but was talking in the third person about people who come into the clinic. I think there's so many ways to interpret expert that's a little bit creative.

30:26Loukas:Another good example of that I saw in a native static format was if you've seen those carousels before it was like I worked at Burger King for 10 years here's all their secrets. so it's again like doing it through the lens of like you work somewhere that that has authority within whatever that is but translating it to like a native format yeah i think that's really

30:47Olly:i actually think that's super that's that for me works even better with younger demos as well where it's like it's almost more native more organic feeling than if just slapping a doctor

30:58Loukas:in front of someone yeah um i think the problem with when i see expert ads that really work versus don't the ones that really work are the ones that are very specific about like okay what is the what what is the unique problem mechanism so it's like the aha behind okay here's what the problem mechanism is so for example you know if we're talking to um women in creatine and that women don't get enough creatine because they don't produce as much as men um if we speak to that in the hook and we've identified what the problem mechanism is we can then come in with what the solution mechanism is that then ties the two together i think a expert can do that with like a lot of weight because it almost is like that did you know that uh you know the the reason we

31:49Olly:suffer with x problem is because of y yeah yeah i've definitely seen that play out um i also think you can do there's so many ways to do experts i think people think of it a bit too simply i guess that's how I'd put it it's like the two a little bit too rigid with the execution it's like no let's try and just break the frame of how we do it slightly differently white green screen overlay is another great example um but yeah just seeing that pop up more and more um another have you seen this new ai tool called infinite ugc no by any chance okay so this is super interesting

32:23Loukas:sounds scary because my uh we had some shout you know who you are shout out um for for this but we had one of our strats that like briefed into our ai videographer a concept that you might have seen it on twitter before i can just maybe overlay here just an example of what it is but like it almost is like an ai animation ad but it kind of keeps one single shot it kind of it almost like runs through the frame and has this this really like smooth way that is transitioning scenes i think i do know um and but it's that i've seen seeing those pop up and like um turns out that you can't do it on higgs field like it's very specific to this infinite ugc that can produce these style of ads um and someone within our team because basically i was i was quite our ai videographer was like can we start using this new tool when we're spending thousands a month already on like a Higgs field, I kind of wanted to set some constraints about not having to get more and more of these tools.

33:25Loukas:Whereas actually like looking into it and the CS tested it, then I tested it. If you give, if you go in and you want, you can literally one shot a prompt and it will build the whole end to end scene just by giving it a script. And it's more like those educational explainers and it literally within like 15 minutes created a one minute 30 ad that was almost perfect um all by just giving it the script and it puts it into this format so that's definitely something to look at as a as a way to do you know video ads very low touch that don't require the same setup as Higgsfield which is more niggly and requires much more time to build out the scene structures and actually be able to replicate a character or avatar throughout all of the scenes yeah it's so interesting i'm

34:16Olly:gonna have a i'm looking at it right now it looks you're thinking i'm gonna rip these just gonna have a go it's a great idea um super interesting i need to check i guess the i think we spoke with about this when alex was on about the the skill for those like ai i'll call it ai slot because it is kind of like its own category of it's its own vehicle i think at this stage like ai you've got like pixar you've got so many different types it's always pointless me reading the whole list but you you there's infinite amount of formats there especially as you start to like blend them with UGC blend them with some of the other styles it can just so much so much more runway um I still think the key there is is firstly great scripts but then the ability to turn that into good visuals that's the skill um the skill gap for many people exactly um because I've seen that more and more in accounts where they're just throwing so many of those ads in and they're doing nothing yeah i think it's all because either the script's poor itself or the ability to visualize the script is a completely different skill uh in some way i i haven't i haven't built that yet i need to i'm practicing um just not got that creative muscle yet so i'm trying to try to build

35:32Loukas:that more this looks really interesting i think um just one thing to mention as well because we're going through with our strats a lot of upskilling now around um other growth pillars and how they can impact creative performance um so you know looking at conversion distribution offer landing page um and you know not just jumping to conclusions on creative performance by what you're directly seeing within the ad account because i've seen quite a few examples with brands that we are saw aren't doing the the paid uh the growth strategy for where the problem actually hasn't come down to the creative and by digging a bit deeper it's been either post click um or it's it's it's come down to something to do with you know how they are test their testing environment within the ad account and then they're actually pushing enough spend through creative to before turning off those assets.

36:31Loukas:Or attribution window was another one that I saw actually impacted the performance of the creative because it was optimized for lower funnel versus when we started introducing top of funnel.

36:43Olly:Yeah, if you have like one day view through in and no exclusions, which is like, please go check that in your ad accounts because you'd be surprised how many people still have both of those things just burning cash. for him. It's a great segue to landing pages. I'm conscious we need to get onto Yappers. I'm not going to talk about this for too long. I'll talk about it for like two or three minutes.

37:02Loukas:Could just be a two-parter

37:04Olly:and we go into Yappers

37:06Loukas:in part two. Perhaps, yeah.

37:08Olly:Yappers could be a whole episode, to be honest. Landing pages are, again, I think we touched on this at the very end of the last one of these we did. I started to speak about landing pages, but I just feel like it's the next wave of competency it's like it's taken a couple of years to if we look at our own business two two took about two years ago maybe two and a half years ago we started really pushing performance creative two and a half years later i think it's like competency level in businesses team structure processed um output is is massive there's just so much greater there's so much depth there's so much more than there was then when we were we were basically just like ugc statics super early very formulate ugc like direct problem solution ads with products at the start um feels like we're about we're starting to enter that place for landing pages and it's really clear to see the impact of being really good at post click on ad accounts and whether that's aggregate or whether that's product specific or even on an ad level now.

38:23Olly:I've seen examples where changing the lander will either shift the audience of the ad or just completely change the performance of a concept through that congruency of message. um and uh i think there's the the brands in the us have have probably a little bit further along in this journey in terms of like creating a process that connects um whether that's angle persona or ad level to different landing pages so you look at like im8 you look at gruns marsmen And this is just me naming a few off the top of my head. Kind patches, I think, do quite well. We've done a few breakdowns of these. Maybe we can put some below.

39:13Olly:But it's clear to see that gap and that ability. If you are able to build that muscle, your creative process will be better. You will get more out of your ads, especially for top of funnel. But what I would say is sometimes we'll have this conversation with brands and then they'll go from zero to 100. and they'll go from making like no landers to like 50 i think the the hard thing to do is to like with creative like you just mentioned like what's the right strategy what should i do next

39:43Loukas:based on the data yeah so i was actually going to turn this question to you then similarly to how you asked me on creative like if you were a brand and you heard what you just said you know and you were to then go away and you mentioned, you know, you can approach this through the format of Landa, the persona, congruency, angle, pain point specific, so many different directions you can take. So, you know, how is a brand, what do they need to do to go away and identify what they should prioritise there and why?

40:19Olly:I think of it through like impact, probably impact, confidence and effort. I just think that's the best way to look at strategy for anything where you could do loads of things and you try to condense it down to one. Quickest thing is, so you've got like layers to it in terms of how hard it is to change, but also the impact that the change could have. So biggest impact would be like offer. Lowest impact would probably be like messaging change, but the time and complexity of those are like inverted. I think every brand should be looking at like top ads in the ad account and just changing the PDP headline and images based on the message in those ads.

41:05Olly:Like we've seen 20, 30 % RPS, revenue possession improvements from just changing the headline and the imagery. Usually that happens when the product is selling to two very different personas. We like to use creatine as an example because it's easy to articulate, but that's a really good one. where it's like if we're selling to gym bros and we're selling to 45 plus menopausal women common sense pdp should be different for those thing even things like the text size should be bigger for people who are over the age of 45 because your mum probably uses her phone on massive text and like you go onto your website and she's probably looking at like pdp images with like size 10 font it's like just they're not reading that info um so i think it's starting with that a lot of quick ways you can do that is i think we've shared this before maybe we can put it in the link sorry sorry team for all of these um is we've shared an ad to landing page congruency process where you just upload all of your top ads into gemini and just give it a prompt and it'll tell you the personas and you could compare that to um and all of them it'll also pull out the message by persona and like what you should maybe change um so i think that's a big quick win yeah

42:18Loukas:Yeah, so prioritizing user experience for your highest leverage personas within your account.

42:25Olly:And again, what I've just articulated there for the older demo is like taking a step back. Yeah, it's the message, but it's also like, right, what does someone, what trust signals do they prefer? So older demos would resonate with things like payment icons below the add to cart, bigger guarantees, different trust signals. So maybe like a lot of what works with them is Facebook comment screenshots rather than Trustpilot. I'm trying to think of other ones off the top of my head. Facts, experts, whereas if I'm going after someone under the age of 25, it's more like UGC, influencers, creators, celebrities, things that they see as aspirational or trustworthy, don't need payment icons because they trust the internet as a good place to shop.

43:17Olly:things like that which are getting a little bit more common sense but you apply that on top of the um on top of the message and you can get some really big wins um second to that would be like right where are we at in the ad spend journey and where are we at relative to our market sophistication and then what has our ctr cpm and cac trend looked like over time and what is our and also probably CPMR. So if we're starting to push some real volume in a market and maybe our ad account is quite heavily focused on sort of solution-aware, product-aware ads, and our hit rate across problem-aware or unaware has started to decrease over time.

44:04Olly:And I appreciate that's quite a lot of data to pull out. The question I would be asking there is, like are we expecting too much of our ad creative and is this product too complex to convey effectively on just a two-minute ad and a pdp for this persona and then if that's the case i'm probably added in a step page and that there's so many i'm not going to go into that because that's like a wormhole of like things to talk through but listicle advertorial five reasons why seven reasons why ten reasons why whatever however many reasons quiz funnel um I'd be starting to think about testing those and focusing those on those unaware ads that have really, unaware problem aware ads with really high CTR but low C click to purchase and trying to connect the gap between the two.

44:51Olly:And often you'll find that if you can get a really good one, it will allow you to unlock so much more spend. And maybe that funnel won't be at the same CAC, but the whole ad account will just be able to spend like way more.

45:06Loukas:Yeah.

45:07Olly:So those are two like creative to post-click steps. And then the third one is just offer and perceive value, which is another, probably an episode in itself. But I think offer is always construed or thought about in e-commerce discount. But I do think the way you frame an offer, the way I spoke about this on the episode with Carl Weish, probably two episodes ago now, but the way you present the offer and your ability to create higher value perception of your product through the messaging, through the association, through the structure, through the anchoring to other products or other alternative ways to buy that product can massively shift perceived value and your ability to generate revenue.

45:54Olly:So I think that's like a whole wave and we're starting to see it impact the ad account because we're starting to see as you get better there, your creative delivers more efficiently your cpm comes down it starts to compound on the front end as well yeah but it can be a real challenge if you if you for for brands um who are often those that are in the less sophisticated markets like i'll come back to creatine creatine probably doesn't need a step page it's firstly it's cheap and secondly it's pretty simple so again that's that's probably doesn't need some crazy landing experience probably just send them to the products

46:34Loukas:and sell it in the ad um that's that's a big thing even with creative strategy that i think people aren't thinking about enough like what's the price point of the product what's the aov um because like yeah if if if it's if it's a if it's a product that's that's less expensive then a lot of the time it won't necessarily require the same level of uh strategy that a product that is double triple quadruple the price would yeah i think i think that's really important

47:08Olly:that's a bit of like nuance that's that's the sort of nuance that's hard to like formalize yeah being able to look and be like probably probably don't have enough room to create a step page here because i think the efficiency loss will probably mean we don't hit our cack target so i'm not going to bother um the the example i have on that is you'll know the brand i'm talking about but again a supplement brand where the team went deep into like micro

47:35Loukas:personas um and the messaging almost they they thought that was the binding constraint in terms of like driving incremental reach um but the messaging became so specific that actually it reduce the TAM and what was working was, was something that was more like benefit stacking. Um, and you know, being a little bit broader with the personas that we're speaking to, because although yes, it, you know, we had the specificity of like speaking to a one type of person and one fear, one motivator, we almost didn't need to go to that level of depth because the product was £12 and actually you know the that wasn't necessarily the binding constraint that at that given time or the right lever to pull within that account yeah yeah that's um I definitely

48:28Olly:think that can be a problem um yeah you get that with landers and offers as well but you could treat that by say the offer is um comes back to like what what are we working against as we spend more through meta over time it's like you you assume every year it gets 10 % more expensive and the audience you're getting in front of 12 months down the line is significantly worse than the audience 12 months ago because that's just how advertising works like you move further away from this golden cohort of really easy to convert consumers towards this harder to convert audience so you either need to get better execution which we spoke spoke about or we need to fundamentally change our ability to pay for those customers so we can just continue to outbid people in the auction and that's why gross margin unit economics offer is so important because if you just are able to spend more to acquire a customer then nobody can beat you in the auction and you win at an aggregate level there's ways to engineer that through offer like example being creating like maybe you go and sell a year's supply of creating you create an offer that's now 50 quid and therefore you can afford to outbid everybody on first purchase um and then you can afford to put a step page in as well for those people who need that education um so i don't think it's necessarily as simple as i articulated maybe but yeah i think it's picking picking the right lever at the right time is really important and to summarize i just think people need to be making more landers um making more landers uh but try to keep it simple and not make 50 might start with a couple a month and do and also when you make final point when you make them iterate don't just throw it in and sack it off like put it in watch the heat maps watch the data see how people are interacting with it are they opening the accordions not buying maybe you should change the messaging things like that don't just like make five a month and just sack each one off because that's not

50:31Loukas:not the way to approach it yeah and i think um you know it's that thing of like not always reinventing the wheel like it's what is like the lowest hanging what is the lowest hanging opportunity right now you know that that you could be focusing on from like a persona messaging standpoint post click yeah rather than you know trying to reinvent what you think it should be like literally just use the data in front of you and start there honestly that's why i like five

51:00Olly:five to ten reasons why whatever number you want to go for because it's like once you've built the page it's all about the message and it's quite easy to just like sit down and be like right what should those five reasons why for be for my product usually it should or seven it should probably be all of the benefits then like how many people have bought it before some of the guarantees and maybe a few people who really like it that maybe they'd know and that's like seven reasons formulas built and then you can just iterate on it um and then you can build out more complexity from there but yeah i think that's the the right right right way to think about it um

51:37Loukas:nice well i think we should come back to yappers in part two yeah of this episode so please give this a like subscribe to the channel uh if you haven't because next week is an absolute banger we're going into much deeper into yapper ads as well as following on this format of like things that are just working right now, but wanted to kind of dedicate a little bit more time to this one. So yeah, we will see you on the next episode.

52:08Olly:Thanks for watching.