Subscribe
Share
Share
Embed
Targeting and bidding are automated, so creative is the new lever. Meta marketing scientist Tommy Changaris on creative diversity, fatigue, and building a system that keeps producing winners.
Video is the most powerful tool in your marketing stack, and most of it is being wasted. Not because the footage is bad, but because no one’s thinking about the business outcome it needs to achieve. Everyone’s still stuck in the "fast, good, cheap: pick two" triangle. We say f*** the triangle.
We're David and Jess Siciliano, co-founders of Savvy Studios. We started in the traditional agency world and left it behind because, guess what? The model is broken.
Whether you're an agency or a brand, each episode you'll walk away knowing exactly what to ask for, what to push back on, and how to make video work harder for you.
We believe you can create great ads and brand videos with any budget. People just don’t know how to make it work. After listening, you will.
bumper-verify-8e6c1345
00:00:00.040 — 00:00:10.840 · Tommy
Targeting is automated, bidding is automated. The biggest lever that you really have at your disposal is creative. The official line is creative is the new targeting, but it's going to be the people of the best creative operating system.
00:00:11.360 — 00:00:12.760 · David
O creative operating system.
00:00:12.800 — 00:00:14.360 · Tommy
You can steal that, can I?
00:00:14.400 — 00:00:15.320 · David
I might steal that.
00:00:17.280 — 00:00:19.720 · Jess
This is Call Us disruptors.
00:00:19.760 — 00:00:31.480 · Speaker 4
A show for the agencies, the marketers, the brand side leaders who are done pretending that pretty video equals performing video. Hi, I'm Jess. I'm David and this is class disruptors.
00:00:31.520 — 00:00:33.080 · David
We have Tommy singer song today.
00:00:33.240 — 00:00:33.880 · Speaker 4
Yes.
00:00:33.920 — 00:01:05.410 · Tommy
Hello. That one, that one. That was a good one. That one. All right. Hey, everybody. I'm marketing science. Science partner at meta. I help larger brands measure the effectiveness of their ads, figure out what's working, what's not. And I come at this from that angle. Uh, there's a lot of data on what works and what doesn't.
And one of the biggest levers that, you know, the data is constantly shouting from all sorts of different angles, is that creative is the biggest lever that advertisers have available to them.
00:01:05.690 — 00:01:08.090 · David
I have a question right off the bat. What were you saying?
00:01:08.130 — 00:01:09.170 · Speaker 5
I'm really hot.
00:01:09.330 — 00:01:09.930 · David
I know, me too.
00:01:09.970 — 00:01:11.730 · Speaker 5
You put the icon on the top.
00:01:12.010 — 00:01:13.210 · David
Did you finish fixing it?
00:01:13.730 — 00:01:14.850 · Tommy
Oh, is that what a gentleman was here?
00:01:15.330 — 00:01:16.650 · Speaker 5
Yeah. Gentleman.
00:01:16.850 — 00:01:18.890 · Tommy
The gentleman. Young.
00:01:19.690 — 00:01:21.250 · Speaker 5
I know you are beating.
00:01:21.250 — 00:01:22.330 · David
I mean, I'm always sweating.
00:01:22.330 — 00:01:25.290 · Speaker 5
So you are literally beating. Okay. Sorry. What was your.
00:01:25.330 — 00:01:43.050 · David
My question was just yesterday in the research, I was given this fact that the life cycle of a winning meta ad has shrunk from three months to 14 days, and the recommended cadence is now 10 to 20 new creative variants a week for 10-K per month spenders. I don't know if that's a good segue into the first thing you see wrong right now.
00:01:43.090 — 00:03:13.140 · Tommy
Yeah, yeah. No, that's that's a good uh, I'm not familiar with that specific, uh, data point, but I mean, I it's kind of not, uh, surprising to me. Um, I mean, the gist of it is a lot of marketers still use audience segmentation as like how they do their targeting. Um, and then they also here create a diversity here.
We need new creative. And they think that, uh, minor variations on the same ad of the same concept count as diversity, creative, diverse and creative variety. Um, and I think both those things are pretty wrong. Um, I see advertisers who are running, you know, 50 campaigns with ten ad sets, uh, each with five different custom audiences, and each one of those, um, to try to, like, really hyper target their customers who, you know, they have good reason to believe, like, oh, this is what our customers were, what our customers look like.
Um, and at the same time, thinking they need creative diversity and having 50 different variations on one ad that are minor changes. Uh, different font, different, slightly different copy, slightly different call to action. And those aren't really actually diversified. Um, so, uh, yeah, that's I think again, um, some of what you pointed out in terms of like the life cycle and stuff that aligns to what I would say typically.
Um, but yeah, those are the two major problems I see.
00:03:13.180 — 00:03:16.580 · David
Why why why is that a problem now? So we used to do that, right?
00:03:16.620 — 00:03:26.139 · Tommy
Yeah. Yeah. And it's very and I think there was a time particularly like the 20 tens where like doing that was probably the right call. Um, because there was so much
00:03:27.220 — 00:03:53.340 · Tommy
fine grained targeting that was available. And, you know, the algorithm, our ads delivery algorithm, I feel save the algorithm. And there's a lot of different algorithms, but that one was not as mature as it is now. Um, and so I think that there was a time where that was that made sense, but especially now in the advantage plus area era, that's kind of the default, um, campaign creation for lower funnel objectives and,
00:03:55.020 — 00:03:55.500 · Tommy
uh,
00:03:56.670 — 00:04:36.590 · Tommy
Targeting is automated, bidding is automated. The biggest lever that you really have at your disposal is creative, and it's really the way to the main input and differentiator that you can have to, um, you know, find your way in front of the right customer, at the right place, at the right time. Um, you know, the we're personalizing, I think, 3 billion feeds per day.
Um, that's crazy. And, uh, using AI and, you know, um, creative and creative diversity is how you really tap into that level of personalization we're doing. I think about half of all like ad out not excuse me, not ad outcomes, auction outcomes.
00:04:36.630 — 00:04:38.070 · Speaker 5
Um, what is an auction outcome.
00:04:38.110 — 00:05:14.880 · Tommy
Auction outcome like i.e. uh, my ad, you know, uh, won or didn't win a specific auction for a specific placement, uh, in someone's feed. Okay. Um, about half of those are attributed. Half of that outcome is attributable. Attributable to creative. Um, other things obviously go into it. The bid um, the person their estimated action rate.
These are all like factors estimate an action rate is kind of affected by the creative. Um, but there's all these other factors that do play a role. But creative is like the single biggest lever, um, that, that actually has an outcome on that.
00:05:14.920 — 00:05:16.560 · David
So. Okay, so let me make sure I understand so far.
00:05:16.600 — 00:05:17.200 · Speaker 5
Yeah.
00:05:17.640 — 00:05:29.800 · David
Advantage plus means AI is going to place the ad for you. Yes. No longer are we doing like I'm targeting women in there who are 35, who live in Vista like you. Now it's just I'm making creative, especially 34.
00:05:30.040 — 00:05:30.920 · Tommy
Very specific.
00:05:31.120 — 00:05:45.040 · David
To 28.5. Uh, now it's Advantage Plus is gonna put the creative in front of that person if it makes sense for them, or they're going to bid to put it, the different creatives are going to bid to put it in that spot because Advantage Plus is said, this is the right ad for this person, right?
00:05:45.080 — 00:06:12.360 · Speaker 4
Yeah. It's it's um, there's there's a lot to it. But yeah, that's, that's that's definitely the gist of it. Um, and it's more like you want to think about rather than be like, hey, I'm going to make an ad and then tell the person to tell the algorithm to deliver it to this specific type of person. I'm going to make an ad that speaks to that specific type of person in terms of the creative, in terms of the motivators and the hook and all that.
00:06:12.400 — 00:06:42.880 · Tommy
What's actually in the ad? Uh, let the algorithm then find that person and people similar to them who might not exactly fit the exact same criteria, like when I've done in this example, like women who are 34 and within Vista. Um, again, some specific San Diego geography here, but like, uh, hey, maybe that same ad would be very would also do well with, uh, women who are slightly outside of that age range, who live in Oceanside and Carlsbad and, uh,
00:06:43.960 — 00:06:57.770 · Tommy
Bonneville or whatever, you know, whatever the surrounding area. And by saying, hey, this specific person, I've now limited the amount of people who. That I could be served to. Even though it probably would work on all these other people too. Right.
00:06:57.810 — 00:07:01.090 · David
Um, or maybe there's a random community in Truckee.
00:07:01.290 — 00:07:02.250 · Tommy
Also has the.
00:07:02.290 — 00:07:10.850 · David
Same relation and advantage. Uh, advantage because these people are similar. Let's throw it there. To which we wouldn't have targeted if we were doing it manually.
00:07:10.890 — 00:07:16.969 · Tommy
Exactly, exactly. Yeah. Um, so that's, I think, a big piece of it. And the
00:07:18.010 — 00:07:32.690 · Tommy
I think the other element to think about is like, uh, and we'll touch on this, I think a little bit, but like when, um, I have a really small audience, uh, I'm just kind of adding constraints to a system that are
00:07:33.770 — 00:07:43.290 · Tommy
maybe don't seem arbitrary, but like to the system are arbitrary of like, okay, cool. I'm just going to ignore everything outside of this now, even though there's not a good reason to write.
00:07:43.290 — 00:07:54.180 · David
Um, as opposed to is it maybe more? Would you say call it more holistic to say I'm looking for sales in general, so go find the best sales as opposed to putting in all the input and saying, don't look at these for sales, right.
00:07:54.220 — 00:08:16.539 · Tommy
Yeah, yeah, yeah. I mean it's it's like, um, hey, I've said, don't look at anyone outside of this specific group of people when, because I think that this specific group of people that I've predefined are my customer, and my customer might actually be slightly different than that. And maybe they have this.
There's a, you know, I've defined them by demographics or whatever in this case. But like there's interests and
00:08:17.700 — 00:09:07.420 · Tommy
other things about them that, um, you know, algorithm can infer by all their activity that they do on the platform, uh, that, hey, I wouldn't even think of that. Like, oh, actually, women below this age, you know, might be interested in this too. Or, uh, men might be like, I see this all the time where advertisers will want to be like, oh, I just like, it's women's clothes.
And I just want to show the ads to women, okay? Men buy clothes for women in their lives. So it's there's a likely there are other likely buyers besides just women. In that case, where if I say, hey, all the way over to women, like, okay, we just took out half, you know, probably not half, but like a chunk of people who would otherwise also be interested in buy your product, uh, just because of an assumption you made about who would actually want to buy this.
00:09:07.460 — 00:09:16.620 · David
Right. And then maybe that maybe, I mean, maybe I'm making a story, but maybe that customer, like, needs less ads fed to them because they're like, that looks nice on my wife. I'll buy that. Right. Like, maybe they're a better sale.
00:09:16.660 — 00:09:22.540 · Tommy
They could, they might, they might be a much easier sale, right? You don't you don't know. Uh, but the system can figure that out.
00:09:22.540 — 00:09:35.260 · David
Okay. Okay. So that's that's a big thing that's changed, right? We we went from a human operated to now there's a system doing the testing and evaluating and placing better than we can. Yeah, right. That's a big shift.
00:09:35.300 — 00:10:29.910 · Tommy
Yeah, it's really big. But it's also like to me it kind of it's like back to before social media advertising and like when you, you know, there was all this specific data where you had to just, hey, I'm going to put billboards up. Not that billboards don't exist anymore, but like, hey, like when I was like the main form, you know that or like, magazines or whatever.
Print. Um. And I can't target. I can't like micro target, any specific audience. I just have to make creative. That goes out in the world that will reach the type of person who, yeah, you know, who is my buyer? Um, and so it's sort of almost like back to that where it's like, hey, that kind of day trading, micro adjustments is no longer what we do actually go back to, like, hey, actually, just having good creative and good ideas is what is going to drive the best results for you and and get you in front of the right person, in the right place, at the right time.
00:10:30.790 — 00:10:34.830 · David
And we're talking about creativity, but we're kind of leaning into creative testing, right?
00:10:34.910 — 00:10:35.230 · Tommy
Yeah.
00:10:35.590 — 00:10:40.950 · David
My follow up question naturally would be if we're not then saying who to target, how does it know who to target?
00:10:41.230 — 00:11:16.439 · Tommy
Uh, so there's a testing element to that. But I think the big thing is that, you know, whenever, um, you have an ad and you have people who buy, you know, there's ad, you launch an ad. People buy from that ad. Um, and the system is then looking for people who are similar to the buyers to show them, to show those people the ads.
If you've said, hey, the only you've kind of predefined, like the only people who might be similar to my buyers are fit this specific category. Again, you're limiting the system in its ability to cheaply
00:11:17.480 — 00:11:48.850 · Tommy
try to serve ads to people who are similar in ways that maybe you don't see or know about, but, uh, are true. Um, but especially in like, the types of things that you can, you know, specifically like call out whether it's demographics or even interests in things, um, that are like predefined where, you know, it's, it can find similar people who have who exhibit similar behaviors as your the buyers already by using that creative.
00:11:49.130 — 00:11:55.090 · David
Um. Okay, so then it's going in the creative. You're saying we're putting that targeting information in and we're putting this all the time?
00:11:55.130 — 00:12:42.770 · Tommy
Yeah, we're putting the targeting into the creative, basically. Like that's that's a good way to think about it. Um, example. Um, so let's say you have a beach blanket company. Okay. It's a very specific example, but it's one that I've used before. And, uh, there are young you think that your main customer is a young, single people who go to the beach, but there are also families who want to buy your blankets.
And so if you only ever advertise, put in your ads, uh, imagery and videos and stuff of like young single people using your blanket, there probably still are some, you know, family type buyers who will see that, but it won't necessarily go to them as much. If you have a similar concept. People at the beach enjoying themselves in the blanket there or whatever.
Um,
00:12:44.210 — 00:13:29.370 · Tommy
if you have hey single, if you have one concept where, hey, it's single people playing Frisbee and then another one where it's family building a sandcastle. Um, and they, you know, feel functionally different. Hey, you're now segmenting. You've now. Ah, you know, those two, you know, buying buyer segments that you have are now being effectively will now be effectively reached by those two ads.
Uh, much better than if hey, I said people in these age groups who are likely to have kids and these sorts of things, and people in these age groups who are unlikely to have kids and live in these areas. Um, I've now basically given the algorithm a lot more freedom to try to go after these two groups, and maybe in ways that are not the ways that I would have predefined those groups with whatever my own assumptions are.
00:13:29.410 — 00:14:19.580 · David
But I think it's interesting. We we've talked about this a lot is like, um, that work, you know, in the 20 tens even used to be, okay, we're going to target women who are single in their 20s, who wants this beach blanket, and then, okay, we go hire their video production company. Hey, you know, we want the lead character to be single looking women in her 20s who like this beach blanket.
Here's some, you know, and then there's a creative process and blah, blah, blah, blah, blah. But now if the if the creative is doing that targeting, not the strategist, that means the video person has to know that and they have to start putting. They have to be asking, which is what we do. Well, okay. So who are all the people we're targeting?
Well, a lot of our customers are single women in the 20s, but there's also some families, there's some some men. There's like some older women in their 60s for some reason and, you know, the southwest that really like this product. Okay. So then we're maybe casting three different heroes.
00:14:19.620 — 00:14:21.700 · Tommy
Probably probably a few. Yeah. Yeah. Exactly. Yeah.
00:14:21.700 — 00:14:30.620 · David
And it's no longer like we're going to do make one ads with a bunch of different hook variations. It's we're actually going to make significant variations to casting. And that's the differentiation in the creative.
00:14:30.620 — 00:15:16.910 · Tommy
Right. Yeah I think the the way that you want to vary these things is across like truly structural dimensions. So uh, hook motivator format and pacing, you know, human presence, brain integration. Picking a few of those buckets and then varying those as opposed to like, hey, we made the ad with this CTA called action versus this one.
And like, okay, cool, we got variance. Like, no, like it actually probably needs to be like, hey, we made this same ad with a family doing X activity and then and this motivator and then hey young single people with this motivator and this example, like hey older person with this motivator and like varying it across all three of those things.
Yeah.
00:15:17.150 — 00:15:35.470 · David
I mean the first thing I could hear a brand saying, listen to this is like, that sounds expensive, but I don't think it has to be. I mean, sounds like if you're already doing the creative and you're you're changing something significant like casting or the scenario a little bit, then it could be done pretty efficiently.
And you, you kind of have to because what if it isn't? What if the one you do that day, the one ad you create, doesn't perform well?
00:15:35.710 — 00:16:15.680 · Tommy
Yeah. Let me say that that would be that'd be a bummer, especially if you if you kind of put all that effort into, like, one thing. Um, so yeah, I think and I think it's like one of those things that needs to start at, like at, like the beginning. Yeah. Where you have to be planning based on that, not, hey, we're going to make this one ad and then this general and then hyper target of these people.
It's actually we're going to keep our audience really broad so we can go to anyone. But then, hey, we know that we have some sense of these are the motivators. These are what people are saying, why they like it. These are the types of people who like it, build things for them, and then we'll get those people, plus people who are similar in ways that we maybe didn't think of.
00:16:15.720 — 00:17:10.240 · David
Yeah. I think what's fun about this is like, um, this is a whole different conversation around creative and video production than, I don't know, it's just it's changed significantly. Um, and I think it's important, I, I really I was really excited to talk about this because I think it's important that marketers, video producers, brands out there see that the reason why budgets are shrinking, quote unquote, they're not really shrinking.
You just need more variation. And so they're trying to. They're trying to take the same budget they had before and make a bunch of more things out of it, but it just means we have to work differently. They need to get similar ideas. It needs to be more adaptable, more modular. I don't know, the budgets aren't necessarily shrinking.
We just we need to do different. We need to approach our production a not shooting one thing, you know, like, yeah, that's kind of a waste. You need more. You need to try different things because the testing is going to be in the creative itself. That is a huge shift than the way it used to be ten years ago.
00:17:10.560 — 00:17:22.920 · Tommy
Yeah. That's you know, I'm less well versed in like the production side of like it just happens all the time where I'm there, like, like, you know, people would be like, oh, so we need to make a bunch different things. I'm like, yeah, probably like it will work better. Yeah.
00:17:23.120 — 00:17:39.280 · David
They go to the video press company and they're like, cool. So that same I the number I would say is 100 grand project for one video. And they're like no, no, no, no, no, we can't do that. And they're like, wait, what happened? We used to do that ten years ago, right? We can't do that anymore. It's now got to be these like micro shoots that you're doing quarterly or whatever.
00:17:39.370 — 00:18:10.370 · Tommy
Yeah, I think having them having thinking about it as like a system of making different creative all the time is way better than thinking about it on like, an individual day's shoot or day's production basis. Um, and, you know, you guys are no better than me about like, what goes into all the, like that back end, uh, production stuff.
Um, but yeah, I mean, the the better advertisers are the ones who are going to consistently have a variety of creative.
00:18:10.410 — 00:18:13.450 · David
Yeah. So creative is the new test, the new targeting.
00:18:13.450 — 00:18:14.090 · Tommy
The new targeting.
00:18:14.130 — 00:18:15.370 · David
Yeah. University is the new target.
00:18:15.650 — 00:18:42.380 · Tommy
Yeah, I would say the official line is the creative is the new targeting. But to do that, to do to target different audiences, you then must have diversity, right. Um, and then and then the, the other note on that is like and thus keep audiences broad. And we can get into that in a second in terms of like creative fatigue and like testing is like the the way that you let that creative do its job is by having a broad audience.
So. Sorry. Yeah.
00:18:42.380 — 00:18:50.180 · David
So when this says, you know, 10 to 20 new creative variants a week, that's not like changing out the text in the first three seconds of a video.
00:18:50.220 — 00:19:03.260 · Tommy
No, that would be that would be different. A completely different motivator, different hook, different concepts. Um, visually distinct. Uh, that's another another important piece. Um, yeah. Yeah. Not not just different CTAs.
00:19:03.300 — 00:19:32.620 · David
Right. So for testing creatively now it becomes like okay, so maybe our our scene number two and three is the, the woman with the beach blanket unfolding it. And you see how soft the fabric is and how durable it is and all that stuff. But the first shot, maybe we film three different versions of like she's like different reasons why she'd go to this project.
Sure, that would be a creative, diverse test, even though it's the same concept, you're having significantly different looking motivators or looking locations or whatever in that first shot. Right.
00:19:32.660 — 00:19:37.699 · Tommy
Yeah. Even if like the concept is the same of like women with blanket as a concept
00:19:38.820 — 00:19:54.460 · Tommy
that like, yeah, there could be a lot. There could be a few concepts within that of, like, it was easy to pack. Right. It looks good at it looks esthetically pleasing. Yeah. It's easy to clean or, you know, whatever the case may be. Yeah. Yeah, yeah.
00:19:54.500 — 00:19:55.500 · David
Negative three out of one.
00:19:55.700 — 00:20:10.020 · Tommy
Yeah. Yeah. Exactly. Now you have three sorted and she's doing three different things there of like whatever like easily putting in the car like it's out somewhere in nice setting or hey she's packing up and like. Right. Sand falls off really easily or whatever.
00:20:10.020 — 00:20:21.100 · David
The right. And then they're asking me the same. And this is not changing the text that's in the first. No, no, I can't tell you how many times we've been asked, can we have a couple variations where the CTA changes in that first opening text changes and it's like, what are we testing you?
00:20:21.140 — 00:20:26.420 · Tommy
I mean, you can it's just the especially if like everything else is visually similar, like it's the same.
00:20:26.460 — 00:20:27.300 · David
Add to the algorithm.
00:20:27.580 — 00:20:42.350 · Tommy
Yeah. So yeah, it's it's it's it is not different. It is not differentiated enough to be treated in a way that it will serve to two different people and be an effective like segmentation and targeting tool. Uh, I love it. Yeah.
00:20:42.390 — 00:20:44.070 · David
Okay. Talk to me about creative fatigue.
00:20:44.110 — 00:22:19.480 · Tommy
Creative fatigue. Uh, a lot of people, you get good creative. You. Congratulations. You did it. Uh. It's working well. Um. That's great. You should want that. But I see a lot of advertisers who will get a piece of winning creative, and then kind of run it to the ground and have not, like, prepared anything in the meantime, and then just kind of scramble reactively to, like, find some new stuff.
Um, by the time, like, performance is actually dropping, like, it's kind of too late. Like you're now in a reactive mode. Um, and there's also, like, you're probably ignoring some early signals, um, that, hey, this was starting to fatigue. We had a great piece of creative like who we did had a good piece of creative.
It worked really well. We got really good results from it. Awesome. But then we just kind of keep running that we don't replace it. And by the time it runs in the ground, we haven't actually prepared anything new to replace it. Um, and, you know, frankly, by the time that, uh, performance drops, you end up, like, scrambling to replace it and then go through a period of worse performance while you're still replacing it.
Um, there also, in general, if that happens, they've ignored some early signals. Um, we actually have some very specific definitions of creative fatigue, um, that will be surfaced to you in ads manager in the delivery column. Uh, there's Creative Limited, which is when, uh, the ads, like cost per result starts to increase above, uh, or exceeds previous ads, and then you actually get a big creative fatigue, uh,
00:22:20.760 — 00:22:23.520 · Tommy
warning badge essentially. Uh, when.
00:22:23.520 — 00:22:23.680 · Speaker 5
The.
00:22:23.920 — 00:24:03.770 · Tommy
Book. Yeah, it's a little it's like a little bubble like thing. It's in the delivery column. Um, and, uh, if you're seeing those, It's it's already kind of too late. Um, and so there are good things to watch out for, but you also do want to stay ahead of the problem. Um, other big things is fatigue is compounded.
But I'm going to keep coming back to this because it's really important by smaller audiences. There's just less. I've made the audience smaller. There's less people who I can show it to who have a, you know, high propensity to click and buy and take or take whatever my desired action is. Um, and as I run out of those people, uh, it gets harder and harder to get in front of more people who are maybe likely to do that.
And so the costs start going up. So again, the narrower the audience, the worse that problem is going to be. Um, and then by the time you're feeling it and you're like your actual top line metrics like Roas, you've probably been leaving performance on the table for weeks. Um, the big thing that, you know, I say is like, build a like you should have like a, a production, um, like cadence, not just a production like sprint.
Like every so often. Um, creative should get refreshed every. The rule of thumb I usually use 2 to 4 weeks for some time. Sometimes it's a little bit longer, sometimes it's shorter, depending on how much spend you have. Your audience, uh, you know, what kind of offer you have, that kind of thing. Um, and so, you know, those creative fatigue indicators can be really helpful and like, actually look at but like even those are kind of, you know, I'd say Creative Limited is a little bit of a leading indicator.
But creative fatigue is now very much a lagging indicator. Yeah. Um so.
00:24:03.890 — 00:24:04.810 · David
Interesting. Yeah.
00:24:04.850 — 00:24:18.810 · Tommy
That and then the other big thing is like, hey, just back to what you're talking about earlier. Oh, we'll just change out the CTA or we'll change out the font or color or something like that. Um, what won't help the problem?
00:24:18.810 — 00:24:19.410 · David
Uh, same.
00:24:19.410 — 00:24:20.250 · Tommy
As the same ad.
00:24:20.570 — 00:24:40.380 · David
Interesting, because I don't know. And I don't know if you can speak to this. This is also maybe not a fair question because you haven't seen any of these ads I'm going to refer to. Yeah. There's been a few times where we've made, um, CTV ads and then they're like, can we have the social version of it? And of course the CTV ad, you don't need to hook them in three seconds.
So you should probably, um, they can't skip it. So it's a very different.
00:24:40.420 — 00:24:42.020 · Tommy
It's a different medium. Yeah. Yeah for sure.
00:24:42.060 — 00:24:56.660 · David
And so we'll reformat it for 916. It's the 32nd CTV ad and it like blows everything else out of the water. It's like the top earner. Um, and I'm as I'm listening to you, obviously you can't confirm or deny. Probably. But I'm wondering how much of that is just like, it's just a completely different ad than what they've been running.
00:24:56.660 — 00:25:41.190 · Tommy
So, yeah, I mean, it's it is probably completely different than what they've been running. I think the everything when I talk about like creative diversity is like within the meta platform. So like, hey, if you poured over an ad and again, you probably you made the appropriate changes to nine by 16 and the right pacing and timing and all the other things I would tell you from like a TV, you know, from a TV ad.
Oops, sorry. Um, if you do that, it is a genuinely new ad to our system. Yeah. Um, but, uh, so, you know, I think it's not unreasonable to do that kind of thing. I do see advertisers do it a lot. Or, you know, have them tell me like, oh, we don't really have any brand assets. And then I see them like run a TV, I see a TV ad, and I was like, oh, you're.
00:25:41.190 — 00:25:41.270 · David
All.
00:25:41.270 — 00:25:45.510 · Tommy
Those. What's that? That's a brand that. That's a brand asset. Like what's funny?
00:25:45.750 — 00:26:50.790 · David
What's funny about this conversation is that like, um, I know from traditional video production world, a lot of video producers will make the 32nd AD and then hold on to all the raw material and you get the 1/32 AD. So if they want more, they got to go back and ask for more revisions. Oh, interesting or whatever.
And so what's interesting is you've said two things that we do, but I just want to it's like validating. And I want everyone to know. Yeah. Because we should all be doing it. Um, it's hey, we may we shot all the. We shot this whole ad. There's probably all this other footage. Right. So that should be made available to iteratively test quickly because it has to be quick.
And then number two, um, it needs to be in the production flow. It needs to be in a system where we're doing where we're refreshing it and it's more fluid, and we're doing it on a either quarterly or every two months or whatever, so that you can make variations quickly. But the current, the current process is and I can hear brands saying this as they listen to you.
Well, how am I supposed to afford to create, you know, ten variations of creative that's significant enough for testing and do it every month. Well, we need to change how we're approaching our shoot days. It can't just be one ad one day.
00:26:50.830 — 00:27:31.280 · Tommy
Sure. Yeah, right. Yeah, I think that's that's super true. Um, and, uh, yeah, it's I would say it's interesting to me the thing about like, uh, that a production company might, like, give you the your 32nd TV spot and then none of the other, uh, material from 30% or charge you for that. That's actually. It's funny because I get why they would charge for it, because it's that that is valuable material to have.
Because you can make, again, like social ads or cut other versions of the same ad from that. Um, again, depending on like, you know, what's conceptually different in there. But yeah, that's why I, I didn't I didn't I didn't know that that's something that happened so that that might be part of it.
00:27:31.440 — 00:27:47.800 · David
This is my opinion. I'm that is I will be clear about this, but I, I mean, they used to do that because they'd make one ad for the year and run it on TV all year. And so like that would be a big up, you know, it's like, well, but how am I supposed to get recurring revenue from you? But now it's no, no, no, we'll come back next month.
Have it like we're going to do it again in two months. Like yeah.
00:27:47.800 — 00:27:47.920 · Tommy
Yeah.
00:27:47.920 — 00:27:55.640 · David
Yeah. Have all the material like that's the change. And I just don't know if I don't know if all three sides of the industry has caught up the agencies, the brands and the video production people.
00:27:55.680 — 00:28:17.120 · Tommy
Yeah. No, that that toy makes sense. Um, and yeah, that's so funny because like I said, it is like a, like, you definitely have the assets, but maybe they don't. Maybe they don't have the assets, as it were. They have the TV shot spot and then they don't have anything else to again be able to oh, cut it to the right format and size and whatnot for social.
So that's that's interesting.
00:28:17.160 — 00:28:19.320 · David
Like no, no don't look at the man man. We don't have that footage.
00:28:19.360 — 00:28:19.960 · Tommy
Yeah we don't, we don't.
00:28:20.120 — 00:28:20.560 · David
Footage is.
00:28:20.560 — 00:29:18.290 · Tommy
Too ugly. We threw it away. Yeah. Or it's oh yeah. Or it's ugly. It's like unpolished. It's like sometimes that works great. Yeah, great. Sometimes that's awesome. That can be amazing. Totally. Uh, depending on depending on the brand and, like, exactly what you're doing, that can be awesome. Um, so anything else?
Yeah. Let me say we touched on this. So just refreshing the like for fatigue, you need to refresh the concepts, not just the color or the call to action or a couple pieces of the copy or whatever. Um, and then I think the other thing too is like, um, we touched on this a little bit format is a big one of like, you know, I think for you guys like, hey, we made a video ad, there might be stills in there that are great for, uh, static ads.
And having the mix of appropriately sized video and static is also that's another piece of the diversity that's really important to be including and thinking about. And that will again help with fatigue and that kind of thing.
00:29:18.330 — 00:29:29.140 · Speaker 5
I mean, we added to our offerings stills, didn't we? Yeah. Um, which I think is really why not? We're already here. That's already set up. It's lit. There's. We've got talent. Yeah.
00:29:29.180 — 00:29:45.460 · Tommy
When I'm. I mean, again, I'm not the video production person, but I'm imagining it. We took a bunch of video that are all just frames of the way that video works. Is taking a bunch of tiny a bunch of pictures really, really fast. And somewhere in there, I'm betting there's some good pictures that are good static cats, too.
Uh, for the same reason.
00:29:45.460 — 00:29:49.540 · David
The marketing science partner meta is saying, just grab a screengrab and turn it into stills.
00:29:49.580 — 00:30:05.980 · Tommy
I wouldn't say, yeah, I mean, there there are scenarios where I'm like, probably don't just willy nilly do that, but like, I'm sure that there are scenarios where like, actually there's some great. Like again, single screen grabs to like use as static ads and that you've already shot, you already did so.
00:30:06.020 — 00:30:12.900 · David
But to just I mean, I mean to be transparent are we added that into our system because you mentioned you said that at the gym.
00:30:12.940 — 00:30:13.420 · Tommy
Yeah.
00:30:14.060 — 00:30:32.220 · David
Yeah. We should bring our photo camera and just take a couple quick photos that you set up. And now they have like multiplicative ads. They have the video ad we just shot. Plus like if it's three scenes they have three more ads at least, right? So now we've made at least four ads, and I'm sure we could find ways to make 10 to 20 on the one shoot day for the for their month or whatever.
00:30:32.260 — 00:31:01.140 · Tommy
Yeah. The for the format is another big just piece of like and it comes down to like different people who are maybe in other similar ways. Maybe one spends a lot of time in Instagram feed and see static stuff. Someone else who's similar maybe spends a lot of time in reels. They see more videos, and so being able to have the right format that reaches both those people where they spend a lot of their time on platform, like that's really important.
So.
00:31:02.620 — 00:31:03.220 · David
Okay.
00:31:03.820 — 00:31:05.580 · Tommy
Uh, testing. I know you want to talk about that.
00:31:05.620 — 00:31:05.780 · David
Do you.
00:31:05.780 — 00:31:47.670 · Tommy
Want to. Because testing. Testing, I mean, testing, this is a big part of my job. Like this is I do this a lot, um, in not just creative, but a lot of other a lot of other sorts of things, but definitely creative. And I see kind of two ways that people go about this. I see plenty of people do it well, so let's just say that.
But I see two kind of main ways that people go about this the wrong way. And one is, uh, having no, like, principled, um, rigorous way to like, test and isolate, like different parts of different concepts to kind of see what's working the best, um, or who test super obsessively and they, they run, you know, split tests or AB tests on
00:31:48.830 — 00:33:20.120 · Tommy
blue versus red button on purchase here versus buy now CTAs and like insane stuff like that. And neither of those approaches are very actionable. In the one case, you don't really know what's working and what's not. And the other case, you're way too hyper focused on things that don't really drive a meaningful difference.
Um, and so I see both those things. Um, some workarounds would be, uh, Meta's creative testing feature. This was launched, I think, to 100% of advertisers in January of 2025. So it's been around for a little bit. Um, and it lets you test up to five variants of an add in, like outside of your main campaign. Um, so you can put, you know, your five variants in there.
It runs with its own budget, uh, doesn't disrupt your, your primary campaign. And then you get winners out of that that you can then, you know, move to your primary campaign. Um, really important even within that is like, hey, not doing the, again, different CTAs, but like, hey, having different concepts that have that vary across, you know, structural dimensions.
But hey, where we've maybe isolated those structural dimensions in terms of this hook versus that hook, this motivator versus that motivator, uh, this, you know, um, type of scene versus that type of scene, uh, that kind of thing. Um, so, you know, utilizing that is great as well as, you know, hey, more formal structured a, B testing again, when we're at the concept level, not at the red versus blue button level.
00:33:20.280 — 00:33:37.960 · David
Yes, I know you're. It's funny because you called it insane, but how many times have we done CTA swaps and whatever? I'm saying, we do it all the time. It's like a standard deliverable, and we're like, okay, sure, sure, sure, sure, sure. But I know because the médecine, the marketing science partners at meta, says that is not a test.
00:33:38.000 — 00:33:53.440 · Tommy
Yeah, it's I mean, it's it it is a test on a variable that makes such a small difference that it's just it's not, um, it's I'm, it's not going to show a big difference in most cases, so.
00:33:56.400 — 00:33:58.160 · David
Um, okay. What else?
00:33:58.280 — 00:34:55.810 · Tommy
Uh, I mean, the other thing I said, I mentioned the creative testing feature, um, concepts. And I just, I think, I do think there's an important element of like, always be testing and iterating. This is I think this is a really important one. It is true for creative. It's true for other things. But um, having kind of a mindset of continuous iteration, continuous learning and constantly testing, but constantly testing in like principled ways where, hey, I'm isolating a variable, a meaningful variable, like like a motivator or something like that.
Um, and tracking. Hey, here's what I've tested, here's what has worked, here's what has not worked. Um, to be able to kind of come up with new hypotheses and then test those and having a really structured a way I like to think about it is like, uh, hey, I'm going to test these two things at the end. I always want to be able to say like, hey, when I changed x variable,
00:34:57.370 — 00:35:21.260 · Tommy
uh, I saw this percent difference in CPA or, you know, whatever the relevant KPI, uh, for this type of, you know, like, have a really structured sentence that you want to be able to say at the end and then, oh, hey, I can, you know, structure my testing and my experiments or what have you based on that. Like, I need to be able to make this statement at the end and then that helps me back into, oh, how do I structure my testing?
00:35:21.300 — 00:35:22.980 · David
Can you give an example? I think I blacked out.
00:35:23.180 — 00:35:47.300 · Tommy
Uh, yeah. So it would be like, uh, an example would be like, hey, you know, like a unprincipled thing would be like, hey, back to the, like, the blanket example or whatever. Um, would be like, uh oh, I've got some different, you know, concepts. I haven't really looked deeply into what I mean by concept or what those are, but I don't want to see which one does better.
So I'm going to test those all against each other, um, and see which one does better.
00:35:48.420 — 00:36:32.030 · Tommy
There's a test that comes out of that. But better would be to like say um, of these I've got three different concepts that are, uh, vary by this variable, by the, you know, the the motivator. Back to the the girl who cares whether she cares about the way, how it packs, uh, how it looks or, uh, how, you know, easily how easy it is to clean or whatever.
Um, so I've got, I've got three different concepts which vary by this motivator. And I want and I'm at the end of this test, I want to be able to say this motivator drives an x percent better CPA. Uh, where cost per like this better cost per purchase than the than the others because everything.
00:36:32.030 — 00:36:34.310 · David
Else was the same. Because everything else was experimenting.
00:36:34.350 — 00:36:56.309 · Tommy
Yeah, that's how experimenting works. So I varied. But again I chose the meaningful variable as opposed to I'm going to test which CTA, uh, did, you know, drives that same thing. So there's a little bit of nuance there for sure of like it because it can be like you can get you can go down the rabbit hole of like, what is this motivator really that different or, you know, that kind of thing.
Um, but
00:36:57.470 — 00:37:19.230 · Tommy
I think there's the sort of there's a, there's a balance somewhere in the middle between, hey, I'm just kind of throwing at the wall and seeing what sticks, and I'm obsessively varying and I'm obsessively isolating variables that are like, not super important. Somewhere in the middle is, hey, I'm testing at a decent pace continuously.
And I'm isolating meaningful variables.
00:37:19.230 — 00:37:20.030 · David
Interesting.
00:37:21.630 — 00:38:05.310 · David
Yeah. Because we're working with, um, I don't know if we wanna take this, but we're working with a supplement brand and they did some, like, storytelling ish CTV ads, and, uh, it performed well on retargeting the most because they've already been hit with, like, this is what this product is. This is what it does.
The sales wasn't as good. So they said, okay, well it's works for retargeting. So let's hit something that's a sales ad. So we want to do the doctor talking about I've made this product and we said cool. And we're like so what should we do. There's a couple of ways we could do this. And we said, let's do three different ways and let's do one.
Is the doctor's in a doctor's place working with patients, one is a doctor's in a home place, but he's clearly a doctor talking about the product still. And then the third is it's a testimonial of customers talking about how great this doctor was. And we see the doctor at the end is that.
00:38:05.350 — 00:38:10.230 · Tommy
That's that's that's good. Yeah. That's that's, that's that's differentiated. Yes. That is, that is diverse creative right there. Right.
00:38:10.230 — 00:38:20.120 · David
Because then we'll know for them down the line okay. Maybe do maybe don't do this one but do more of these These kinds were never right, as opposed to being like. Sure, let's pick one of those and see how it goes. Now they're risking it all in one approach.
00:38:20.160 — 00:38:36.680 · Tommy
Yeah, exactly. Yeah for sure. Because one of those you might get lucky and one works well, or you might get lucky. They'll work well with different and maybe with slightly different audiences. Um, but yeah, it's it's definitely like, that's like worth, like testing. Yeah.
00:38:37.800 — 00:38:38.400 · David
Okay.
00:38:38.440 — 00:38:52.039 · Speaker 5
I like bringing it back to bringing us on right at the very beginning, because all of these things are talked about like during building the concepts, the strategy behind like why concepts are
00:38:53.280 — 00:38:54.120 · Speaker 5
chosen.
00:38:54.160 — 00:38:54.440 · David
Yeah.
00:38:54.480 — 00:38:55.600 · Speaker 5
And how they're made.
00:38:55.640 — 00:39:05.320 · David
We're always having to remind brands like, so what's working like, what have you tried. Like why are we doing this? Like who is your target audience? Where do they live? And they're like, oh yeah. Well, you know. Yeah. Uh, it's like, what do you mean? Well, yeah.
00:39:05.400 — 00:39:06.200 · Tommy
Yeah, I would.
00:39:06.240 — 00:39:06.560 · David
Almost.
00:39:06.960 — 00:39:35.209 · Tommy
Say like that should be like in the, in terms of the creative process, like that should be that like kind of segmentation, customer info should be like embedded in the process. Like at the beginning when you're starting to talk about like the ideas of creative, because that flows directly from what you're like, what you're just talking about in terms of like, hey, who's the customer?
And like, what do they care about? Where are they? Like that kind of stuff. Like, that's not for
00:39:36.250 — 00:39:44.490 · Tommy
that's not that's not something you should do. Later when you're like and set up like audiences to do that. Like you should do that. That should be how your creative gets made, right.
00:39:44.530 — 00:39:48.130 · David
Part of the concept being in the creative brief, some would say, which is the last episode was about.
00:39:48.290 — 00:39:49.890 · Tommy
Oh, okay, there we go. All right.
00:39:49.890 — 00:39:50.890 · Speaker 5
Well, yes.
00:39:51.210 — 00:39:52.490 · Tommy
To cut to that real quick.
00:39:52.690 — 00:39:53.130 · Speaker 5
Yeah.
00:39:53.770 — 00:39:54.290 · Tommy
Yeah.
00:39:54.450 — 00:40:03.410 · David
Um, okay. I love it. Okay. So to summarize, we need one system. Oh, wait. What? I didn't actually read that before I said it.
00:40:04.530 — 00:40:45.620 · Tommy
Yeah. I mean, the way I'd phrased that was like, uh, we kind of talked about three different problems, but it's really one system. Okay. Um, where? diversity, creative diversity feeds the algorithm. Yeah. The algorithm uses the diversity as targeting. Uh, if you don't refresh that, eventually it's going to die off.
Um, and then testing is the pipeline that kind of helps you stay ahead of that fatigue and know how to be consistently producing good creatives. Um, it's definitely the people who are going to the advertisers who are going to do well. The brands who are going to do well are not the people who are the most clever in terms of their custom audiences or things like that.
But it's going to be the people with the best creative operating system.
00:40:46.100 — 00:40:47.540 · David
Ooh, creative operating system.
00:40:47.580 — 00:40:48.420 · Tommy
You can you can steal that.
00:40:48.620 — 00:40:50.220 · David
Can I? I might steal that.
00:40:50.300 — 00:40:51.180 · Speaker 5
Would that be us?
00:40:51.220 — 00:40:52.140 · David
Yeah. Well, that's.
00:40:52.140 — 00:40:54.420 · Speaker 5
How the creative operations.
00:40:56.020 — 00:41:02.540 · David
All that work happens in pre-production and concept for the shoot. Not after. That's a huge change. I love that. Okay, so should we do a rapid fire round?
00:41:02.540 — 00:41:03.500 · Speaker 5
Yeah. You want to pass it to me?
00:41:03.540 — 00:41:04.940 · Tommy
Yeah. Oh. Hit me.
00:41:05.460 — 00:41:07.980 · Speaker 5
Okay, so this section is what would you do?
00:41:08.020 — 00:41:08.500 · Tommy
Okay.
00:41:08.540 — 00:41:15.060 · Speaker 5
Okay. I've got 50,000 to spend. Do I make one amazing ad or ten decent ones?
00:41:15.460 — 00:41:24.980 · Tommy
Uh, I would say make ten. Hopefully slightly better than decent ones, but but yeah, the more, more, uh, concepts, more diversity would be better.
00:41:25.020 — 00:41:31.100 · Speaker 5
Yeah, I love it. I only have budget for one day of production. What's what's my priority? What should be my priority?
00:41:31.500 — 00:41:38.420 · Tommy
Uh, getting as many as much material that works for different concepts as possible.
00:41:38.580 — 00:41:45.380 · David
Ooh, that's a clever distinction. I'm just gonna jump in there because it's not like getting different angles of the same thing over and over again. It's like different setups.
00:41:45.420 — 00:41:55.020 · Tommy
Yeah, yeah. Not just yeah, not just not just an angle of the car 50 times. Yeah. For not not just. Oh we got a, we got the thing out of the car from the front view or from the rear view of the car like you know.
00:41:55.220 — 00:41:56.180 · David
Yeah. Slow. No, no.
00:41:56.220 — 00:41:56.860 · Tommy
Yeah.
00:41:57.140 — 00:42:00.980 · Speaker 5
Get it out of the car. Get out of the cupboard. Get it out of the bag a bag.
00:42:01.020 — 00:42:02.060 · David
Yeah, yeah.
00:42:02.340 — 00:42:03.980 · Speaker 5
Okay. My next, uh,
00:42:05.740 — 00:42:21.630 · Speaker 5
chapter is called. Is this actually Diversified creative. Okay, I changed the opening text on the first, right? No, I changed the CTA from Learn More to shop More. No. I swapped the spokesperson, but kept the script.
00:42:21.910 — 00:42:36.310 · Tommy
Maybe if, like, the spokesperson looks very different from the other spokesperson and is maybe in a slightly different environment, and there's other kind of visually distinct things, even if they're saying some of the same stuff, that that would be more diversified.
00:42:36.390 — 00:42:37.670 · David
You said only maybe, though.
00:42:37.830 — 00:42:42.470 · Tommy
I said maybe it's I would say that there's some of these things are tricky of. Yeah. Uh,
00:42:43.830 — 00:43:05.070 · Tommy
it can be, you know, hey, if it's, if it's two if it's like two on the nose, the same. And we've just changed colors or the color of the background around. I know, probably not. But if it's hey, it's kind of some of the same general, uh, motivators, but in a very different environment. It's visually distinct from the other environment, including the spokesperson that would that would be more diversified.
00:43:05.950 — 00:43:09.320 · Speaker 5
Uh, I made one version funny and one version Serious?
00:43:09.360 — 00:43:12.840 · Tommy
That's. Yeah. That's good. Yeah. That's that's. Yeah.
00:43:13.200 — 00:43:34.440 · Speaker 5
Yeah. We just did that with a, uh. I'm going to I guess I'm going to mention this in every single podcast episode, but we just did a shoot for prostate, uh, fun, a supplement, and we did one that was funny, where it's from the dog's point of view, and one where it's just the man can't fall asleep because he keeps needing to pee.
00:43:34.480 — 00:43:57.080 · Tommy
Oh, yeah. That's that's great. That's. I was saying that those are two very different concepts. Definitely diversified. And just you're going to reach different people. There's some people who really have problem. The second problem and really, you know, feel that and it's going to resonate with them.
And there's some people who maybe are feeling a bit more silly about it than the other one is going to resonate with them.
00:43:57.320 — 00:43:58.920 · Speaker 5
I said his mom, like.
00:43:59.160 — 00:43:59.640 · David
I love the.
00:43:59.640 — 00:44:01.160 · Speaker 5
Dog and his mom would love the dog,
00:44:02.720 — 00:44:15.080 · Speaker 5
which is funny because she's not the one that would she would buy the pills for his dad, you know. Okay. This is. Would you rather one great concept or ten average concepts?
00:44:15.240 — 00:44:17.480 · Tommy
Uh, ten again, hopefully better than average.
00:44:18.120 — 00:44:22.720 · Speaker 5
A great offer with a mediocre, as I say, that mediocre. Mediocre.
00:44:22.760 — 00:44:23.160 · David
Yeah, I know.
00:44:23.160 — 00:44:30.840 · Speaker 5
That sounds strange in my accent. A great offer with mediocre creative or amazing creative with a mediocre offer.
00:44:31.000 — 00:44:56.960 · Tommy
Oh, as much as we've talked about the importance of creative, I'd say a great offer with mediocre creative over that. If you have great creative, but the offer is not good, uh, you might get a lot of quick. You might have a high click through rate, you might get a lot of traffic. Um, but if it doesn't sell, it doesn't sell.
So, um, you know, you can only you can only fix that problem so much. Um, yeah.
00:44:57.000 — 00:44:58.120 · Speaker 5
So make sure your product.
00:44:58.120 — 00:44:59.040 · Tommy
Make sure your product actually.
00:44:59.080 — 00:44:59.760 · Speaker 5
Doesn't suck.
00:44:59.800 — 00:45:15.330 · Tommy
That's the that's that's before anything we've talked about. That's definitely stage one. Once you have a good product then your differentiator is going to be your creative. But that's way back in the marketing funnel. Just is the product good? Is probably still still has always been important and remains important.
00:45:15.370 — 00:45:20.850 · Speaker 5
Yeah. The goal isn't to find a winner. It's to oh wait. Yeah, it's to.
00:45:21.050 — 00:45:25.130 · Tommy
Uh, I would say create a system that consistently produces winners.
00:45:25.690 — 00:45:35.090 · Speaker 5
I love it. Um, I skipped this one. Uh, this is finish the sentence. Um, the most overrated metric in marketing is.
00:45:35.130 — 00:46:04.010 · Tommy
Oh, frequency. People always care about that a lot. And it's a useful diagnostic metric. It's a useful output. Uh, there's not a lot of great direct levers to adjust that. Um, besides, maybe, you know, broadening your audience, which I'll just keep going back to, uh. But, yeah, that's that one is always people always have questions and it's and wanted want to really manually adjust that.
And it's it's not super easy to manually adjust. Mhm.
00:46:04.450 — 00:46:11.140 · Speaker 5
Okay. Next one. Brands should stop obsessing over and start focusing on.
00:46:11.500 — 00:46:18.740 · Tommy
Uh, they should stop obsessing over, uh, niche audiences and start focusing on diversified creative.
00:46:19.340 — 00:46:20.980 · David
I like that in the settings.
00:46:21.020 — 00:46:34.060 · Tommy
Yes. Yeah. Niche. When I say niche audiences, I mean in the like in my ads manager settings, I'm going in. I'm creating a small audience because the the creative will help me find the right audience within a larger set. Okay.
00:46:35.380 — 00:46:41.420 · Speaker 5
I love it. Okay. Let's do a true or false. The algorithm knows my audience better than I do.
00:46:41.860 — 00:46:58.820 · Tommy
Uh. You saw. True, true. I'd say. Well, no. False when you raise that. False. Because, uh, you probably know you almost certainly know your audience better, but the algorithm knows how to find them better, especially when you give it the right inputs in terms of creative. Mm.
00:46:58.820 — 00:47:01.980 · Speaker 5
Mhm. One winning ad can carry a business.
00:47:03.590 — 00:47:08.270 · Tommy
Uh, probably for a short amount of time. But in the long term, no.
00:47:08.350 — 00:47:08.630 · Speaker 5
So.
00:47:08.630 — 00:47:10.230 · Tommy
False. False. Yeah.
00:47:10.790 — 00:47:12.870 · Speaker 5
The problem is usually targeting.
00:47:13.950 — 00:47:18.510 · Tommy
Uh, true. And that the targeting is usually too narrow.
00:47:19.910 — 00:47:22.230 · Speaker 5
Refresh creative every month.
00:47:22.550 — 00:47:31.310 · Tommy
It's a good rule of thumb. True. I'd say true. Uh, for some brands more. Some brands slightly less. But that's probably a good rule of thumb.
00:47:31.550 — 00:47:34.830 · Speaker 5
Creating different ads for different stages of the funnel.
00:47:35.110 — 00:47:35.750 · Tommy
True.
00:47:36.630 — 00:47:40.270 · Speaker 5
Uh, targeting is overrated. Creative is the new targeting.
00:47:40.670 — 00:47:41.910 · Tommy
True. No notes.
00:47:42.070 — 00:47:45.030 · Speaker 5
Your ad account is a research tool.
00:47:45.430 — 00:48:07.030 · Tommy
Uh, true. It's a lot of other things. But importantly, it is a research tool. Uh, it has a lot of experimentation tools and analyzes tools built into it. So use them, uh, because they're there for for that, for research and learning and getting more and more efficient and getting more sales and doing better over time.
00:48:07.750 — 00:48:08.110 · Speaker 5
I love.
00:48:08.110 — 00:48:08.670 · Tommy
It.
00:48:08.950 — 00:48:13.110 · David
I like it. If you had one thing you want to say to all that, all the people out there, all the brands out there.
00:48:13.670 — 00:48:17.310 · Tommy
Mhm. Uh, creative is the new targeting. Uh,
00:48:18.590 — 00:48:28.709 · Tommy
it's not just the thing we say. It's really true. Uh, it's a really important lever. Uh, spend more time focusing on that than you do on minute audience changes or
00:48:29.750 — 00:48:31.230 · Tommy
things like that. So.
00:48:35.790 — 00:48:36.710 · David
Hey, thanks.
00:48:37.110 — 00:48:44.750 · Tommy
Welcome, guys. This is fun. Yeah, that was good. That was great. Like. And subscribe. Smash the smash. Smash the like button. Don't say it. Don't say like that.
00:48:44.790 — 00:48:49.870 · Speaker 5
Let the data guide you to a comment and a like. And subscribe. Thanks.
00:48:51.270 — 00:48:52.110 · Tommy
You're beautiful.