Latest / Scalability School / How To Test Creative On Meta In 2026

Transcript

How To Test Creative On Meta In 2026 · Scalability School

• 0:00
  And then you're launching these ads into
• 0:02
  those ad sets and you're allowing Meta to either
• 0:04
  spend or not spend.
• 0:06
  And there is a whole host of pros and a whole host of cons.
• 0:10
  A whole host of pros mainly falls under,
• 0:13
  you're probably going to lose less often, lose less
• 0:16
  tests, also lose less money, cash, profit.
• 0:20
  The con is you're probably going to make a
• 0:22
  lot of ads that don't ever receive any spend.
• 0:24
  A bit more like how I like to think of apparel
• 0:26
  and fashion, which is you walk into a department
• 0:28
  store for a brand.
• 0:30
  Each section of the shop is a different campaign.
• 0:33
  So you've got like jackets and you've got
• 0:34
  men's wear and you've got women's wear rather
• 0:36
  than testing area of the shop and a scaling area of the shop.
• 0:41
  I like that.
• 0:42
  I'm not a Zuck apologist here.
• 0:44
  So like I'm not dying on any hills, but I
• 0:46
  think for most of the time it's like, no, you don't
• 0:48
  need to make room.
• 0:48
  If your budget is set on the campaign, an ad
• 0:51
  set level where you can get out of learning,
• 0:52
  then you should trust that Meta is 90% of the
• 0:55
  time or more is going to make the right allocation.
• 0:56
  Brand owners or founders over the last couple
• 0:59
  of weeks where they felt really guilty because
• 1:01
  they're not hitting this volume of ads that's going live.
• 1:04
  Because I've got to work on something creative.
• 1:06
  I've got to work on something creative.
• 1:07
  And instead of launching any, they're just launching none.
• 1:10
  And my recommendation was always the same.
• 1:11
  It's just like, forget about trying to launch a certain number.
• 1:14
  Create one piece of content as if it was going to be posted organically.
• 1:17
  And instead of posting it to organic, just upload it to the ad account.
• 1:31
  All right.
• 1:31
  Welcome to scalability school episode about
• 1:34
  how to do creative testing in 2026 and beyond
• 1:39
  really, hopefully, but you know how to do creative testing today.
• 1:42
  Insane guest for this.
• 1:45
  We have Phil, who is an absolute legend in the DTC space.
• 1:51
  Is Phil Kiel, of course, from Tycoon Digital.
• 1:54
  Has been somebody that people follow all over the world and is an absolute legend.
• 1:58
  And so Phil is here to help talk about this at 9 p.m. his time.
• 2:02
  So thank you, Phil, for joining.
• 2:04
  I would for sure not have said yes to this.
• 2:05
  So thank you.
• 2:06
  Well, I've slept all day and I've awoken like Dracula.
• 2:09
  And of course, we have the always handsome Brad here with us to my co-host.
• 2:15
  Rumor is that could be coming back soon.
• 2:17
  We'll have to see.
• 2:19
  He is currently, you know, doing a million things and raising a child.
• 2:23
  Still on on dad leave a little bit.
• 2:25
  So we look forward to that.
• 2:27
  One thing I want to mention at the top of
• 2:28
  the episode is we do drop in the newsletters
• 2:32
  on scalability school dot com.
• 2:33
  You can sign up for our email newsletter worksheets and extra perks and things.
• 2:38
  So you may want to go over there and sign up
• 2:40
  for the email at scalabilityschool dot com.
• 2:43
  Make sure that you have, you know, kind of
• 2:45
  all the resources that you're getting or that
• 2:47
  we talk about on here.
• 2:48
  Those sometimes get sent out via email.
• 2:50
  So, you know, a little plug there at the beginning.
• 2:53
  So going into the episode, talking about how to do creative testing.
• 2:56
  Number one thing to say is this is the number
• 2:58
  one question in the Foxwell Founders community.
• 3:00
  So I got a report on our bot, which is within the Foxwell Founders community.
• 3:05
  You can basically ask a question to the AI
• 3:08
  bot, which sources answers from, you know,
• 3:13
  from Slack, basically from our community.
• 3:15
  And it this is interesting because it came
• 3:18
  back and said that the number one question that
• 3:20
  people are asking it over the last six
• 3:23
  months was it was asked one hundred and oh, yeah,
• 3:27
  forty five days.
• 3:28
  It was not six months asked over one hundred and sixty times in forty five days.
• 3:31
  How do I do creative testing?
• 3:33
  And this is something I asked Brad and he's like, yeah, you just like rip it.
• 3:38
  I'm like, but what does that mean?
• 3:40
  So we need to get into talking about this.
• 3:42
  And and Phil is a perfect guest to get into it.
• 3:46
  So number one on creative testing, two camps, I
• 3:49
  think there's the for spend under ABO creative
• 3:53
  testing.
• 3:54
  And then there's the second camp, which is let ads earn spend.
• 3:58
  Brad, you want to give a little more color on this?
• 4:01
  Yeah, I can.
• 4:01
  Before I say that, Phil's Phil, I go to I go to Phil for two reasons.
• 4:06
  One, incredible Twitter content, amazing learnings.
• 4:09
  I really enjoy what you put out on the Internet.
• 4:11
  Two, when I'm trying to convince my wife to
• 4:13
  let me get my hands tattooed, I'm just like,
• 4:16
  look, look at how sick this guy looks.
• 4:19
  You're making the first mistake where you're asking the wife.
• 4:22
  That's true.
• 4:24
  Yeah, that is a mistake.
• 4:25
  I just need to show up.
• 4:27
  I just need to show up.
• 4:28
  OK, let's talk about the two camps.
• 4:31
  Phil, do you do you would you prefer to take the CBO side or the ABO side?
• 4:35
  Not because we're picking one better, but we can explain it.
• 4:37
  I'll let you explain whichever one you prefer.
• 4:39
  I mean, it's a tough it's a tough one.
• 4:41
  So I'm more of a CBO fan.
• 4:43
  I am.
• 4:44
  We are running both.
• 4:46
  Definitely.
• 4:46
  We are definitely running both.
• 4:47
  And I'm much more on the mindset of like what's right for the business right now.
• 4:52
  Yeah.
• 4:53
  Yeah.
• 4:53
  Makes sense.
• 4:54
  OK, so let me I'll tee up.
• 4:55
  I'll tee up the ABO camp.
• 4:57
  You tee up the CBO camp.
• 4:58
  And we'll talk about pros and cons.
• 4:59
  Just kind of go back and forth.
• 5:00
  So on the ABO side, so to Andrew's point,
• 5:03
  there's there's there's really two camps that
• 5:04
  people kind of fall into.
• 5:05
  And there's like mini versions of both of
• 5:08
  them that kind of exists across ad accounts.
• 5:09
  But it's basically we force spend.
• 5:11
  We let meta decide.
• 5:12
  And there's a little bit of like a middle
• 5:14
  ground in the middle there where you can do
• 5:15
  the mid budget, things like that.
• 5:16
  But I'll talk about the forced spend, which
• 5:18
  generally looks like ABO creative testing side.
• 5:21
  And I'll just present a couple of ways that you can do it.
• 5:24
  We'll talk.
• 5:24
  We can talk about pros and cons after we present the both.
• 5:26
  So on the forced spend side, basically what
• 5:29
  this looks like is you set up when I say ABO,
• 5:32
  if you're not familiar with that, ad set
• 5:33
  budget optimization and each new creative test that
• 5:36
  you're launching gets its own ad set with its own dedicated budget budget.
• 5:40
  And you were telling meta, I want to spend
• 5:42
  these dollars no matter what the efficiency
• 5:44
  is, whatever, like I want to spend because
• 5:45
  I'm going to learn against this creative.
• 5:47
  That's what I'm trying to do.
• 5:48
  There's a couple of ways that people split their ad sets then on top of that.
• 5:51
  So you can either you can batch ad sets based on like creative batches.
• 5:55
  So maybe you have like a weekly cadence for
• 5:57
  producing creative and your team submits a
• 5:59
  batch and you get the batch and you can launch like that.
• 6:02
  There are people who I laugh, but like I've
• 6:05
  seen it and I'm sure it's fine in many cases,
• 6:07
  but people will set up a single ad set per ad,
• 6:11
  even to a degree where they're running literally
• 6:13
  one at a time.
• 6:13
  And they're just like trying to say like, okay, does this, does this work?
• 6:16
  And it's got single ad with its own budget.
• 6:18
  There's a couple of like mini variations of
• 6:19
  that where it's like similar to batch where
• 6:21
  it's like maybe you have a video concept,
• 6:23
  you have three hooks for the video concept,
• 6:25
  you launch that into an ad set.
• 6:27
  The last way that I generally see that these,
• 6:29
  these are run is kind of some kind of grouping.
• 6:32
  So it's an ad set budget, but I have a theme
• 6:36
  behind the way that I'm grouping my ad sets.
• 6:38
  So maybe it's handsome men with glasses and a beautiful beard.
• 6:43
  Like that's who I'm targeting.
• 6:44
  And then the mustache, the mustache group, like those are the two groups.
• 6:47
  And you're like, put your creative tests go into, into those ad sets.
• 6:50
  But the, the, the big takeaway from, from
• 6:52
  camp number one is you are setting up mainly
• 6:55
  ad sets where you were forcing spend into
• 6:57
  creative tests and saying meta, I don't really
• 6:59
  care what the results are.
• 7:01
  I'm framing that maybe a little bit poorly,
• 7:02
  but like I'm going to spend these dollars no
• 7:04
  matter what, because I'm going to use the
• 7:05
  learnings of that to dictate what I want to
• 7:07
  do as opposed to this other camp, which
• 7:10
  Phil, as you jump into that, which is the letting
• 7:13
  ads earn spend.
• 7:15
  And I wonder if you agree with the framing
• 7:17
  of that camp, which is a forcing versus not
• 7:19
  forcing.
• 7:21
  Yeah.
• 7:21
  I think that's, I think you've, I think
• 7:23
  you've explained it beautifully there, Brad.
• 7:24
  Um, and yeah, the CBO approach is where you
• 7:28
  are, you're controlling more of the activity
• 7:31
  outside of the ad account.
• 7:33
  And then when it goes into the ad account,
• 7:34
  you're essentially letting the babies leave the
• 7:38
  nest and, and I have a fly or fall.
• 7:41
  And we can keep going on that metaphor or we can leave it there.
• 7:44
  But yeah, you're essentially having a CBO campaign, campaign budget optimization ad
• 7:49
  sets per whatever.
• 7:53
  There's lots of options there and we can, and we can talk through that.
• 7:57
  And then you're launching these ads into
• 7:58
  those ad sets and you're allowing meta to either
• 8:01
  spend or not spend.
• 8:02
  And there is a whole host of pros and a whole
• 8:07
  host of cons, a whole host of pros mainly falls
• 8:10
  under.
• 8:11
  You're probably going to lose less often,
• 8:14
  lose less tests, also lose less money, cash,
• 8:18
  profit.
• 8:18
  The con is you're probably going to make a
• 8:22
  lot of ads that don't ever receive any spend.
• 8:24
  The counter argument to that is if they don't
• 8:25
  receive spend, how often when you then force
• 8:28
  spend into those ads, do they, are you just,
• 8:32
  you know, meta's decision is confirmed as a
• 8:37
  whole, you know, the right thing to have happened.
• 8:41
  And it's quite often that that is the case.
• 8:44
  We have, you know, zombie campaigns, which
• 8:45
  we can talk to, which is sort of like round
• 8:47
  two and which, which is always a good thing to do.
• 8:50
  I think the main thing that I think about is
• 8:52
  what's right for your ad account right now.
• 8:54
  And there's so much content out there on the
• 8:58
  internet around brands that are doing amazing
• 9:01
  things and the structures that they're using.
• 9:04
  And so often I order ad accounts and we're
• 9:07
  getting into the why rather than maybe we should
• 9:09
  just talking about how first, but, and so
• 9:11
  often I order ad accounts and I'm like, when
• 9:13
  was the last time you tested an ad that won and you're still doing ABO testing?
• 9:18
  It's like, you keep banging your head against
• 9:20
  this brick wall and admit, and yeah, I think
• 9:23
  that's, that's usually where we then make the suggestion to move to CBL.
• 9:28
  I mean, it's so much of it is like, I think a
• 9:30
  lot of the, the ABO method was built and from
• 9:33
  what people thought was the right thing to do
• 9:35
  and where a lot of the education was focused
• 9:37
  for a long time, which was the graduation method.
• 9:40
  You put it in there, it's separate, it's clean, right?
• 9:43
  And then you don't do that.
• 9:44
  And I think this reason that CB, well, let me, let me wrap that up.
• 9:48
  So you, you know, take an ad, if it works well, you put it into a scaling campaign.
• 9:53
  Well, this is not something that I think the majority of us are doing anymore.
• 9:56
  You're, you're, you know, it's just, it kills the performance of it.
• 10:00
  And I think if you're still doing this, you're putting yourself at a disadvantage.
• 10:03
  And so, you know, now what we're talking
• 10:06
  about is putting it into a CBO campaign, which is
• 10:10
  one goes on the, the idea of, you know, consolidation.
• 10:13
  But also the second option of the CBO campaign
• 10:17
  in that camp, which I think you'll learn through
• 10:19
  this episode, we're, we're certainly in more
• 10:21
  is you have to have a certain amount of budget
• 10:24
  in that CBO campaign to really like be able to do this.
• 10:27
  Is there a way that you calculate that, right?
• 10:31
  Like, let's say that you have a CBO that's
• 10:34
  running and you're going to launch a new creative
• 10:37
  into it.
• 10:39
  And the budget is hypothetical.
• 10:42
  It's a thousand dollars a day.
• 10:44
  Okay.
• 10:44
  If you're going to add 10 new creatives into
• 10:46
  it, are you going to increase that to $1,500
• 10:48
  a day as you're testing it?
• 10:51
  Or are you going to keep it the same and just
• 10:52
  say, we're just going to let it compete against
• 10:54
  these other ones because spend is the metric, right?
• 10:57
  And I was talking to Harry, Harry D shout out Harry Demelish.
• 11:01
  He was just an incredible advertiser, creative strategist about this.
• 11:03
  And he really is like, look, spends like the only thing.
• 11:06
  And we've talked about that in this podcast as well, right?
• 11:09
  You can talk about creative metrics.
• 11:11
  Somebody I noticed this morning asked you
• 11:12
  about your, some of your social content about
• 11:14
  asking how you track, fill your win rate,
• 11:17
  which I'm interested in hearing about that
• 11:19
  too.
• 11:20
  But I think, is there, is there a budget there that you have to think about?
• 11:23
  You can just keep tossing stuff in the same CBO?
• 11:26
  In the agency, we scale based on performance, not based on launching new ads.
• 11:32
  And if the performance isn't there, we're
• 11:35
  launching new ads to try to improve the performance
• 11:38
  rather than we're in a hole.
• 11:41
  Let's get a bigger shovel.
• 11:41
  So we keep, so we, so we dig ourselves into an
• 11:44
  even bigger, we're going to improve performance
• 11:46
  by launching different ads.
• 11:48
  And there's a very big chance that,
• 11:50
  especially if the performance isn't there, there's a
• 11:51
  very big chance that new ads will get spent.
• 11:55
  And that's a good place to be in.
• 11:57
  The bad place to be in is when your highest
• 12:00
  spending ads aren't good enough because they
• 12:02
  can become a bit of a crutch where, you know,
• 12:06
  got to keep them live to drive that performance.
• 12:07
  But when you launch anything new, it doesn't get spent.
• 12:11
  And then you have to ask, okay, why aren't those new ads getting spent?
• 12:14
  That then I think is a question around
• 12:16
  everything else other than testing method, offer, message,
• 12:21
  angle, you know, the creator, product market fit, all of those sort of things.
• 12:26
  And I think that's a much bigger lever than ABO versus CBO.
• 12:30
  I think if performance isn't there, moving
• 12:32
  from CBO to ABO will rarely improve performance
• 12:35
  in the short term.
• 12:36
  So we should probably end the podcast there.
• 12:38
  Just like that's, all right, lesson learned.
• 12:42
  I did write it down as well.
• 12:43
  If you're asking the question, ABO or CBO,
• 12:47
  and this isn't to put anyone down, but I think
• 12:48
  if you're asking the question, ABO or CBO, you should probably choose CBO.
• 12:53
  And that goes back to, you know, a lot of
• 12:54
  the content that we see online and, you know,
• 12:55
  how all these huge brands that are growing
• 12:57
  massively year on year, they're so aggressive.
• 13:00
  Their margins are so aggressive.
• 13:01
  Their performance targets are like 0.8x
• 13:04
  ROAS, for example, from an acquisition point.
• 13:06
  If you, they're in a very good place to do ABO.
• 13:10
  Yeah.
• 13:11
  The budget, the budget point is an
• 13:12
  interesting one, Andrew, because it's like, Phil, you
• 13:14
  answered it exactly how I would have.
• 13:16
  So it's like, I'm going to say it's brilliant, but it's because I totally agree.
• 13:20
  It's like, yeah, like you shouldn't need to, you shouldn't need to make room.
• 13:25
  And like, we have, we, we get, we will have
• 13:27
  people, you know, I'm sure, Phil, you've experienced
• 13:31
  this, or it's like every once in a while, a client will send you an audit.
• 13:34
  And it's like, hey, did you see that they said
• 13:35
  you have too many active ads in your campaigns?
• 13:38
  And it's like, well, what does too many ads mean and how you define it?
• 13:41
  Because it makes these, it make you, what
• 13:43
  assumption you're making when you say that
• 13:45
  is meta is treating all these ads equally and spending against them equally.
• 13:50
  And I think it really like, should, should you increase budget?
• 13:53
  I mean, the, the, the 80% answer I think is
• 13:56
  like, no, for most people where it can get
• 13:59
  weird is if you have a really high AOE product
• 14:02
  and you launch a ton of creative all the time.
• 14:04
  And when, to your point, Phil, it's like,
• 14:07
  when you launch something, it's going to get some
• 14:09
  spend.
• 14:09
  It might not be a ton, but if you launch a
• 14:11
  hundred ads and you have a $500 AOV meta is going
• 14:14
  to spend against all of that.
• 14:15
  And so it's going to try to use a good chunk
• 14:17
  of your, your existing budget to go towards
• 14:19
  stuff because it wants to give it a chance at least at the beginning.
• 14:21
  And so that can screw up efficiency in the
• 14:23
  short term, but otherwise it's like, no, you
• 14:26
  don't need to increase the budget because meta is, meta is allocating.
• 14:29
  And I'm not, uh, I'm not, uh, I'm not a Zuck apologist here.
• 14:33
  So like, I'm not, uh, I'm not dying on any
• 14:34
  Hills, but, um, I think most of the time it's
• 14:37
  like, no, you don't, you don't need to make room.
• 14:38
  If your budget is set on the campaign, uh, an
• 14:41
  ad set level where you can get out of learning,
• 14:43
  then you should trust that meta is 90% of
• 14:46
  the time more and more is going to make the
• 14:47
  right allocation decision.
• 14:48
  You know, let's talk about bidding as it
• 14:49
  relates to this and do create, doing creative
• 14:51
  testing.
• 14:51
  So, you know, sometimes I've heard people
• 14:53
  saying, look, we're going to, we do all of our
• 14:56
  testing, uh, or, you know, we run it all on lowest cost or we run it all on min
• 15:00
  ROAS or, um, I've even heard things about,
• 15:03
  um, you know, minimum spend with new creatives.
• 15:07
  Um, do you, number one, do you put in minimum spend in your opinion?
• 15:11
  Is that a good thing to do?
• 15:12
  And number two, what do you think about the
• 15:14
  bids as it relates to testing within CBOs?
• 15:18
  So I want to just real quick on it, like say
• 15:20
  like, okay, so if we're, if, if, if, if you're
• 15:22
  coming to, to, to Brad and Phil to make a
• 15:24
  recommendation on like what you should do in your ad account,
• 15:26
  we're, we're both collectively leaning
• 15:28
  towards like, Hey, the CBO approach is probably
• 15:29
  the right one.
• 15:30
  So the reason to stay away from ABO to your
• 15:32
  points is try to avoid this because you're
• 15:34
  going to be forcing spend.
• 15:35
  And if you don't have a good hit rate or
• 15:37
  success rate with your creative, or you've been, you
• 15:39
  know, bad at making ads for a while or your
• 15:41
  offer sucks or whatever it is, you're going
• 15:42
  to, you're going to, you're picking up a
• 15:44
  bigger shovel, shovel and digging yourself a bigger
• 15:45
  hole.
• 15:46
  So if we're leaning towards the AB or the
• 15:48
  CBO camp, what does the CBO camp look like?
• 15:50
  Which I think is what Andrew is starting to, to, to the path that we're going down.
• 15:54
  Um, and like, what does that actually look like?
• 15:55
  So we're saying you should probably go CBO.
• 15:58
  So like within CBO, there's now a bunch of different ways to set that up.
• 16:01
  Um, the way that we generally do it, and I
• 16:03
  think this might be a little bit different
• 16:04
  from how you guys do it.
• 16:05
  Um, and I think you'll be able to better
• 16:06
  answer Andrew's actual question as a result of that
• 16:08
  is we generally, like we have a single campaign.
• 16:12
  Like let's say we, we'll use, we'll use Zach cause he's not here.
• 16:14
  We're selling socks and we have one specific sock product.
• 16:17
  We're setting up a single campaign for that skew of socks.
• 16:20
  And with one ad set in there for the socks, we're not launching new ad sets.
• 16:23
  In most cases, when we have new creative, we're
• 16:25
  launching everything into that existing campaign,
• 16:27
  regardless of who it's speaking to.
• 16:29
  Because on an ad level, and I think I have
• 16:31
  a point saved on this at some point, like on
• 16:33
  an ad level, meta is still making a decision
• 16:35
  around how much to spend into each individual
• 16:37
  ad, um, based on your total budget.
• 16:39
  It's not looking at who it's delivering.
• 16:41
  Like the ad set, this ad set learning thing, I think is-
• 16:43
  Yeah, you're giving it the freedom and you're, and you're segmenting it by product.
• 16:48
  Yes, exactly.
• 16:49
  And, and by product is not because we're trying
• 16:51
  to force creative learning to a specific product.
• 16:53
  It's because those products might have a certain margin of the profile.
• 16:55
  That's because that's where you're making those money.
• 16:55
  Yeah, right.
• 16:56
  Yeah, exactly.
• 16:56
  I'm just printing socks, you know?
• 16:59
  Yeah, exactly.
• 16:59
  But yeah.
• 17:00
  Yeah.
• 17:00
  It's like, I'm trying to set, I'm trying to
• 17:01
  build a funnel around socks because they have
• 17:03
  a certain margin or a certain offer profile or whatever it is.
• 17:05
  And so, so what we're doing, I guess, is to
• 17:08
  wrap that up is single, single ad set in most
• 17:11
  cases.
• 17:11
  And we're just launching in new ads until we max it out.
• 17:14
  Very rarely are we going through and pausing a ton of ads.
• 17:18
  Again, not that we don't do it.
• 17:19
  It's just like, it's just rarely we do it.
• 17:21
  So that leads to the mid, the min spends and like those variations.
• 17:24
  I guess I'm curious if you are doing the budget
• 17:27
  and bid adjustments or like how you're looking
• 17:28
  at that.
• 17:29
  We don't use minimum spend very often, but
• 17:31
  we have done in the past and it's, I like the
• 17:33
  idea of it.
• 17:34
  And they did bring out, so the first iteration of that product or feature was cash.
• 17:42
  So we've, we've got the, the campaign budget is a thousand dollars.
• 17:45
  We can say, I'd set one at a hundred dollars per day minimum.
• 17:49
  I'd set two at $200 per day minimum.
• 17:52
  And then, you know, that's $700 remaining.
• 17:54
  That will be spread wherever Meta wants.
• 17:58
  The latest development of that, which I
• 18:00
  think is a couple of months ago was percent.
• 18:02
  So you could essentially say 10% of spend.
• 18:04
  And then as you scaled the total campaign
• 18:06
  budget, that would then mean, you know, that 10% stays
• 18:11
  relative.
• 18:12
  And I think there's a, you know, there's a,
• 18:13
  there's a logic to why that is attractive, but
• 18:16
  we don't, we don't, we don't use it whatsoever.
• 18:18
  And the main, yeah, we don't use it.
• 18:21
  The main reason is because quite often we'll
• 18:23
  launch ads and they'll get spend in some format.
• 18:25
  If we launch five ads, two of them will get spent.
• 18:28
  And if they don't, then that's a pretty strong indication.
• 18:33
  The bids, I think is a, is an interesting topic and a good, and a good argument.
• 18:37
  And this moves away from sort of how to test
• 18:40
  creative and more into sort of lowest cost and
• 18:42
  bid caps and stuff.
• 18:44
  We don't do any, we don't do any manual bidding on creative testing.
• 18:47
  And we do do some of it on sort of scaling.
• 18:51
  I think it's just like, I think, you know, you've got, you've got all these.
• 18:53
  Yeah.
• 18:55
  I mean, what's the difference for you from a creative testing versus a scaling?
• 19:00
  Like the reason I asked this question is,
• 19:02
  by the way, Taylor holiday, a hundred percent
• 19:04
  listens to this.
• 19:05
  So if we talk about cost caps the wrong way, he's going to flame us all on X.
• 19:09
  Okay.
• 19:10
  So actually I'm sure he doesn't listen to this.
• 19:13
  All right, friends, quick break.
• 19:15
  This episode is brought to you by North
• 19:16
  beam, the marketing attribution platform that we
• 19:18
  love over here at the pod and good news.
• 19:20
  North beam is launching North beam incrementality.
• 19:24
  North beam incrementality gives you automated
• 19:26
  easy self-service incrementality tests while
• 19:28
  protecting you from the major mistakes.
• 19:31
  So many people make while running these incrementality tests.
• 19:33
  Cause honestly, most of these incrementality tests are kind of a mess, right?
• 19:37
  So what I love about it is you, it gives you a casual truth.
• 19:41
  You can act on every day, not one-off siloed reports.
• 19:44
  And it does it end to end.
• 19:46
  So it automates lift testing does.
• 19:48
  It connects you with your MTA and MMM.
• 19:50
  So you can scale.
• 19:52
  It works and cut what doesn't all in one dashboard.
• 19:55
  That's the real game changer.
• 19:57
  Now the North beam has incrementality.
• 19:58
  It means you can run the absolute trifecta of
• 20:00
  marketing measurement in one single platform,
• 20:03
  all using the same data.
• 20:05
  North beam incrementality launches this quarter.
• 20:08
  Get over to north beam.io forward slash
• 20:10
  incrementality and be the first on the list.
• 20:13
  And they're offering 50% off unlimited
• 20:14
  tests for a year to advertisers who join now.
• 20:18
  Tell them we sent you at the scalability school.
• 20:20
  Can't wait.
• 20:22
  Actually, I'm sure he doesn't listen to this by the way.
• 20:24
  But you do, listener, and you are freaking awesome.
• 20:29
  So good job because you're smarter than most
• 20:31
  people on the internet by listening to this
• 20:33
  podcast.
• 20:34
  So on that sense, like you have, what's the difference between those two?
• 20:41
  Is it scaling versus creative testing?
• 20:43
  The only place where that will be really
• 20:46
  dramatically different is when you're using something like
• 20:48
  inflated budgets, where you're setting the
• 20:51
  budget at like $20,000 per day, but it's only spending
• 20:54
  2K, 3K.
• 20:57
  And structurally, this would be you taking
• 21:00
  banger ads that have worked and keep working.
• 21:04
  And they are in a new, they're in a CBO
• 21:08
  that's been running for a while or something.
• 21:10
  Yeah.
• 21:10
  And you're going to jack it up to let it spend when it does, when it's really good.
• 21:17
  And to that, and that's like a scaling thing.
• 21:20
  So these are a group of ads that have started
• 21:21
  as babies that have grown up that are now in
• 21:24
  their own scaling thing.
• 21:25
  And you have a new CBO that then you're testing other stuff in.
• 21:30
  Is that correct?
• 21:31
  Yeah, correct.
• 21:32
  We don't do that very often.
• 21:34
  And it's kind of, and I don't know how you
• 21:36
  would test creative if you were doing using
• 21:38
  an inflated budget structure, because you're
• 21:41
  essentially saying you can spend as much as
• 21:42
  you want here when the performance is good,
• 21:44
  but I'm going to force spend into this, these
• 21:46
  assets over here.
• 21:47
  And it feels like you're trying to do two things at the same time.
• 21:51
  And there's also going to be a hell of a lot of overlap.
• 21:54
  So you've been asking, you just asked, how would we do that?
• 21:57
  And I've answered the question, whether or not we would do that.
• 22:00
  No, no, no.
• 22:01
  I mean, I'm not talking about that.
• 22:02
  Like, I'm not talking about you inflating on a testing one.
• 22:06
  I'm talking about on a scaling one.
• 22:08
  So like, how many ads have to get in?
• 22:12
  Like, one of the things I've heard within
• 22:14
  these, within the challenge of, okay, it's
• 22:17
  a CBO's, you know, testing structure is, won't I have too many ads that are live?
• 22:22
  Like I have too many ads that are live, you know?
• 22:26
  To answer that question, if you've got too
• 22:27
  many ads live and your performance is good,
• 22:29
  then you would increase the budget.
• 22:31
  And that increased budget would therefore
• 22:33
  go back to your original question, you know,
• 22:35
  earlier on, which was when you launch more ads, do you increase the budget?
• 22:39
  So if performance is good, you can double the budget.
• 22:41
  In theory, that gives more room for assets to go live.
• 22:46
  For assets.
• 22:46
  Yeah.
• 22:49
  The way I'm sort of thinking about it now is
• 22:51
  more from like a, an organic marketing structure.
• 22:55
  And it's less about, it's less about
• 22:58
  scientific test and scale and duplicate, which I think
• 23:02
  is where we were sort of five, six years ago.
• 23:05
  And it's more around make the best possible
• 23:07
  content that we can for the brand, try and learn
• 23:11
  along the way and have a piece of strategy and then scale it most efficiently.
• 23:16
  Yeah.
• 23:16
  And so that's kind of your mentality with it,
• 23:20
  which is not trying to do much manual stuff,
• 23:22
  but letting meta decide.
• 23:23
  I think that, I mean, I completely agree.
• 23:25
  It's focused on the, it's content centric
• 23:27
  versus, you know, logistic centric and, and
• 23:31
  lever pulling centric.
• 23:32
  So what I wonder is, let's say that you are
• 23:35
  scaling, have, you know, you're scaling in this,
• 23:38
  you're doing, it's doing well and you want
• 23:41
  to launch a, well, in what instance would
• 23:44
  you launch a brand new CBO then?
• 23:47
  Brad, do you want to answer that?
• 23:48
  Or do you never launch a brand new CBO?
• 23:50
  I mean, like I've heard both.
• 23:53
  Like some people will say, look, I'm going to do a new CBO every quarter.
• 23:56
  Cause like it's becomes too much.
• 23:58
  I don't want to have everything dependent on this one.
• 24:01
  You know, some of it's obviously product or offer or funnel centric, right?
• 24:05
  They're going to launch a new thing.
• 24:07
  And it's like, that's what we're going to do.
• 24:09
  What's your opinion on this, Brad?
• 24:11
  So we will generally speaking, we have campaigns by product or by offer.
• 24:17
  So like, let's say we sell the same product,
• 24:19
  but we're splitting up into, we have different
• 24:21
  offers.
• 24:22
  We'll generally make a new campaign for the
• 24:23
  new offer because we, Taylor is listening.
• 24:26
  We do run a lot of bids, so please don't yell at me.
• 24:28
  Um, but no, we, because, because we're,
• 24:31
  because we're setting bids differently based on the
• 24:33
  goals of that campaign, then we want them to be separate.
• 24:36
  You could do that within, you could do it within a CBO, but so, okay.
• 24:40
  So let's go back to our socks examples.
• 24:41
  Um, if I have a sock campaign and it is,
• 24:44
  let's say it's, you know, lowest cost or cost cap,
• 24:46
  cause it's bidding on the same thing.
• 24:47
  Well, functionally, um, depending on if you're putting a limit at it or not.
• 24:51
  So you've got a sock campaign on lowest cost
• 24:54
  or cost cap, and you've got your single ad set
• 24:57
  in there.
• 24:57
  Um, we will just keep launching ads in there.
• 24:59
  And the only time we'll make a new ad set is if we need, we ran out of room.
• 25:02
  They, thankfully in most cases, now they've
• 25:04
  increased the limit on that to 150 from 50,
• 25:06
  but there's still, there's still ads that are limited at 50.
• 25:09
  Um, so what it might look like is we have
• 25:11
  a, we have a campaign with a bunch of ad sets
• 25:13
  in it that, um, are all in purpose of the same thing.
• 25:17
  They have the same bids.
• 25:18
  It's, but it's all for that same product.
• 25:20
  So when do we introduce a new campaign?
• 25:21
  There's a couple of, couple of cases, new product or new offer, um, or promo.
• 25:25
  Um, we want to switch to a different type of bidding.
• 25:29
  So maybe we're going to go from a cost-based
• 25:31
  bidding to a value-based bidding, um, which,
• 25:33
  in which case we'll do that.
• 25:34
  Um, and those are really like, that's
• 25:36
  really like the only scenario that we're adding a
• 25:37
  new campaign.
• 25:38
  Like we're not doing any like crazy
• 25:40
  horizontal scaling thing or like, you know, trying to
• 25:42
  do it that way or we're forcing budgets.
• 25:44
  Um, and the main reason from that, like
• 25:46
  I've, the consolidation point is like, it'd been
• 25:48
  an interesting topic on my mind lately.
• 25:49
  I saw, I saw a thread recently where, um, I
• 25:53
  can't remember who it was, but he was talking about
• 25:54
  his bid caps and his lowest costs.
• 25:55
  And he's like, you know, I split out interest-based testing on top of that.
• 25:59
  And all of a sudden the interest-based testing
• 26:00
  looked, looked good, but it made my broad not
• 26:04
  perform as well.
• 26:05
  And then it's like, I think what we're
• 26:07
  trying to do is like give meta the meta, the, the,
• 26:09
  the parameters on what success looks like,
• 26:11
  which is purchase optimized campaigns for a
• 26:13
  lot of people for a very long time.
• 26:14
  Um, maybe use cost controls, maybe you don't,
• 26:16
  but then as much as I can allow them to make
• 26:18
  decisions within those guardrails of, of that,
• 26:21
  like they're going to do the best job allocating
• 26:23
  things.
• 26:24
  Uh, so long winded way of answering your question.
• 26:26
  It's like they really, the only time we
• 26:27
  change and add a new campaign is we've added a, uh,
• 26:30
  a different way we're doing value-based
• 26:31
  bidding instead of cost-based bidding or alongside
• 26:33
  or running both.
• 26:34
  We, um, we've run an ad account, which is
• 26:36
  for a, a single supplement and we have a CBO
• 26:40
  campaign per, uh, segment and that's, and it's a liver product.
• 26:46
  And that segment is, is the types of people that would buy this liver product.
• 26:51
  So it's, uh, people who are overweight,
• 26:54
  people who drink too much, people who have fat
• 26:57
  liver disease, you know, X, Y, Z, and you can continue down that route.
• 27:00
  And there's logic that those campaigns could work together.
• 27:05
  Or there's a, you know, there's a sort of
• 27:06
  romantic idea that each one of those different
• 27:08
  campaigns are going to be targeted in different audience.
• 27:11
  There will always be overlap, but it supports our input, which is creative.
• 27:17
  So if we're going to 10 X the number of assets
• 27:20
  for over drinkers, where are they going to go?
• 27:24
  We have to have somewhere to put them.
• 27:26
  Um, and that's the main reason is for, it's just folders on a computer essentially.
• 27:30
  And to ensure that we are acquiring customers
• 27:34
  across all of those different segments, because
• 27:37
  obviously we can control the budget at the campaign level.
• 27:40
  Whereas if we didn't, and we had them all
• 27:41
  grouped together, maybe one CBO with all the ad sets
• 27:45
  being those different segments, then in
• 27:47
  theory, we could, that could actually lead to risk.
• 27:49
  We've moved away from CBO to, from APO to CBO.
• 27:53
  We've moved away to, into CBO because we want to limit risk.
• 27:57
  But actually if we do, if the example I had where all those different segments were
• 28:01
  different ad set in the CBO, we would over focus on one audience.
• 28:04
  And that would, um, you know, put us in a potentially a tough situation after a
• 28:09
  couple of years, you know, we're over focused on
• 28:11
  one product, one audience that then disappears.
• 28:14
  That's a business risk.
• 28:15
  Our competitors are over focusing somewhere else,
• 28:17
  you know, when we're not taking market share there.
• 28:20
  So, yeah, but I think there's a good, if you've got a single product, I think
• 28:25
  there's, there's, there will be reasons why you could launch more CBO campaigns.
• 28:28
  I would probably recommend focusing on
• 28:31
  persona, angle, uh, creative type, potentially.
• 28:37
  And we could talk about that in partnership ads going into CBOs.
• 28:39
  Um, Brad's example of like a CBO per product is a bit more like how I like to think
• 28:46
  of apparel and fashion, which is you walk into a department store for a brand.
• 28:51
  Each section of the shop is a different campaign.
• 28:54
  So you've got like jackets and you've got
• 28:56
  menswear and you've got women's wear rather than test.
• 29:00
  You don't have a testing area of the shop and a scaling area of the shop.
• 29:04
  I'll set up like that.
• 29:05
  I like that.
• 29:06
  But like in the, in the event, so like, uh, on the, on the example that you gave,
• 29:09
  the supplement example you gave, um, those personas, like the, the functional
• 29:13
  benefit inside of meta to you guys is it's folders, it's organization.
• 29:17
  We can see how performance looks on each of these things.
• 29:20
  And like, maybe it's not perfectly delivering to
• 29:22
  that exact segment, but like the intent behind it.
• 29:25
  And that drives what you guys do from a decision making basis and how your creative
• 29:28
  strategy comes together, which I think is really valuable.
• 29:30
  The argument that people make in favor of ABO testing, which I think is a totally
• 29:33
  fair one is like, it's, I think it's, I think it's fair in theory, but I think it
• 29:37
  makes a wrong assumption about people's intent, which is like, we're forcing it.
• 29:41
  So it means we need to care more about the, the input here.
• 29:44
  Um, and we're going to pay more attention to it.
• 29:45
  Cause we know if we need to turn it off, we know if it's
• 29:47
  not working, we know we can learn from it as a result of that.
• 29:49
  Um, but so anyways, the, the functional benefit of the CBO splitting that you're
• 29:53
  talking about is we can talk to these different personas.
• 29:56
  We're going to be accountable to making creative for
• 29:59
  these different personas actually in meta's delivery.
• 30:02
  There's maybe not a massive like functional difference because of they're really
• 30:06
  just folders to your point, but like it drives how you guys take action.
• 30:11
  Is that fair?
• 30:12
  Yeah, exactly.
• 30:13
  And then, and then there's benefits from a business level.
• 30:15
  We can see which customers come from those campaigns.
• 30:17
  We can track them in Clavio.
• 30:18
  We can track their LTV.
• 30:20
  We can, we can, we can pause one.
• 30:22
  We can double down on the other.
• 30:23
  We can raise credit to support one campaign cause the LTV is so strong.
• 30:27
  But you wouldn't do that if you only had a hundred bucks a day or if you had 500
• 30:30
  bucks a day, like you're, you're, you're getting new campaign when you feel
• 30:32
  confident that this is going to be efficient.
• 30:35
  I'm using the term learning now cause I think there's something magical about
• 30:37
  learning, but like, it's just a good practice to have, which is like, we shouldn't
• 30:40
  be splitting out new campaigns if we don't
• 30:42
  think that we're going to be efficient enough.
• 30:43
  One really quick example.
• 30:45
  We have a apparel client who like ran out of inventory through the end of the year,
• 30:49
  basically through most of January, cause things just went really well in November.
• 30:53
  And so, yes, they have a bunch of different products across a bunch of different
• 30:55
  price points, but it's like we, if we start to split those campaigns out, it's like
• 30:58
  we have seven campaigns and there's like no budget
• 31:00
  going through here where that could possibly be efficient.
• 31:02
  So then we, we did consolidate everything
• 31:04
  back down and we made, here's all products.
• 31:06
  It's value-based bidding to control for the ROAS.
• 31:09
  We're just going to let it rip because otherwise there's no way where this is going
• 31:12
  to be efficient enough cause we just can't spend enough.
• 31:15
  The second inventory is back and we can push spend back up.
• 31:17
  Then we're going to split those back out.
• 31:18
  But like, it all kind of comes back to the whole point.
• 31:21
  Andrew, you brought up earlier consolidation.
• 31:22
  It's like, it's context dependent on what the program is spending.
• 31:25
  So Phil, I guess what I'm saying is like you, you
• 31:27
  wouldn't be doing that if you had a hundred, 500 bucks a day.
• 31:31
  It's like you're doing that cause you feel confident
• 31:32
  that each of those are going to be efficient on their own.
• 31:35
  Yeah.
• 31:36
  Yeah.
• 31:36
  And they, they take, they take months, courses, years to build up to.
• 31:40
  Yeah.
• 31:41
  Yeah.
• 31:41
  I mean, spend is a huge function of this, I think.
• 31:43
  Right.
• 31:44
  In terms of the testing.
• 31:45
  Let, let me talk about the, you know, metrics that we look at when we're testing.
• 31:50
  So, so I'm going to do create, I think, you know, I'm going to do creative testing.
• 31:55
  I'm a brand.
• 31:55
  I haven't done it very efficiently.
• 31:57
  I've spent all this money on creative.
• 31:59
  I'm hesitant to put it all into the same CBO.
• 32:03
  Right.
• 32:03
  Because what if I don't get clean metrics?
• 32:07
  And I think that there's two things to say here is like, one, I think to a degree
• 32:10
  advertisers and brands have been sold the idea that the metric is the whole, like
• 32:17
  metrics, creative metrics specifically are the holy grail.
• 32:21
  Right.
• 32:22
  And I think that actually what we're learning about really scaling brands is that
• 32:27
  spend within an efficient CBO is like the most
• 32:33
  efficient thing and is the best metric to look at.
• 32:37
  Because, you know, many times these, there's not necessarily a correlation between
• 32:43
  an ad having a good creative metric or a good
• 32:46
  creative hit rate and actually selling something.
• 32:50
  So how do you feel about creative metrics now, Phil?
• 32:55
  And how are you judging this if something isn't getting the spend?
• 32:59
  I think because like I know this is a common thing that we hear from clients like,
• 33:03
  you know, I spent all this money in creative.
• 33:05
  You're not even going to look at the, you know, how are you going to say like,
• 33:09
  well, actually, we have some early indications here
• 33:11
  based on this rate that this is doing pretty well.
• 33:14
  What is that and what are you measuring and how are you explaining that to clients?
• 33:18
  We use the leading indicators other than spend.
• 33:21
  So, and we can, and we can talk through those, but we use those other metrics,
• 33:25
  everything that's available, CPM, CTR, cost per click, thumb stop, everything up to
• 33:33
  when someone buys the product, add to cart rate, cost per add to cart.
• 33:38
  We use those to justify why something got spent or why it didn't get spent.
• 33:44
  So 90% of the spend went to add A and we had five ads.
• 33:48
  Why do we think that happened?
• 33:51
  The click through rate was 3% and all the
• 33:53
  other ads had a click through rate of 0.5%.
• 33:57
  Okay.
• 33:57
  That tells us something.
• 33:59
  What was this?
• 34:00
  What was the, what were the other metrics?
• 34:02
  Thumb stop, add to cart rate, CPA and ROAS.
• 34:05
  Were they good?
• 34:07
  Yes.
• 34:07
  Okay.
• 34:08
  That's good.
• 34:08
  That 3% click through rate was logically relevant.
• 34:12
  People were clicking on it for the right reason or the same thing again.
• 34:16
  All the spend went into add A.
• 34:18
  It had a 3% click through rate and the performance was awful.
• 34:21
  Okay.
• 34:21
  Why did it have that click through rate?
• 34:22
  Oh, the first three seconds, something happened.
• 34:27
  You know, there was a dog.
• 34:29
  There was something in it that looked phallic.
• 34:30
  I've seen that before.
• 34:32
  Okay.
• 34:33
  We should probably pause that, pause that ad now.
• 34:36
  And the other ads, two to five, will then get spent.
• 34:39
  So we'll use those to justify why something got
• 34:43
  spent or it didn't because clients do not want to hear.
• 34:48
  You spend five grand on these assets.
• 34:50
  This one ad, it had a click through rate of 5% and it spent $40.
• 34:56
  Clients do not.
• 34:56
  That doesn't, that doesn't communicate trust from us as an agency.
• 35:00
  Right.
• 35:00
  Yeah.
• 35:01
  Yeah.
• 35:01
  That doesn't, that's not going to go well.
• 35:03
  I have a slightly different take, which is I used to like, we can force spend into
• 35:09
  it if you want, but we're not going to, we wouldn't recommend doing that.
• 35:12
  And it's just like, yes, you spend five grand on the assets.
• 35:14
  Would you like to spend an additional five grand finding out they still don't work?
• 35:17
  So it's like, I'm, I'm the arbiter of your money.
• 35:21
  I'm not making an opinion about whether I like or don't like the creative.
• 35:25
  I'm making an opinion about, will this make you more money?
• 35:27
  And I, if Meta doesn't think it will, I'm
• 35:29
  inclined to think that they're right most of the time.
• 35:32
  But I, but what you said there is really important is like, you can't just get us
• 35:35
  to have all this money spent on the creative and then do nothing with it.
• 35:37
  I think back to the point about what people who prefer ABO, their argument is,
• 35:42
  well, it forces us to look at it and reconcile why something did or didn't work.
• 35:46
  Fine.
• 35:46
  Um, what, what, what I'm not saying, what you, what you said is like, we're going
• 35:51
  to look at the, we're going to look at the creative and we're going to look at the
• 35:53
  metrics and we're going to try to come up with
• 35:55
  reasons why it didn't spend and justification for it.
• 35:58
  Cause that's super valuable.
• 35:59
  Cause you can take, you can, you can take the
• 36:01
  analysis and do something with that information.
• 36:03
  So what I'm not saying is, nope, sorry, it didn't work.
• 36:06
  See ya.
• 36:06
  Let me know what you got to do creative.
• 36:07
  But we're saying is like, okay, it didn't get spent.
• 36:10
  If you want to force spend, sure, we can do it.
• 36:12
  We probably wouldn't recommend doing that, but we can do it.
• 36:14
  But here's some of the reasons why it might not work.
• 36:16
  If you compare it against other ads in the ad account,
• 36:18
  um, we, I will say we shy away from the soft metrics.
• 36:22
  We try not to lean into soft metrics too much.
• 36:23
  Cause I think that I personally just think that they cause more confusion than
• 36:26
  anything, um, in the way that we communicate at least.
• 36:29
  Um, but I think it's like the broader point is like you
• 36:31
  were trying to give reasons to why something didn't work.
• 36:34
  So there's actionable takeaways with that, um,
• 36:36
  whether or not it's, it's spent or it didn't spend.
• 36:38
  So I think we're, I think we're getting to the
• 36:39
  same conclusion just in that slightly different paths.
• 36:42
  I really liked the idea of benchmarks when I've been training paid social execs
• 36:47
  over the past couple of years, what I've been trying to train them on is like, get
• 36:50
  a, get a daily sense of what the benchmarks are in your ad account.
• 36:53
  And that just takes, that just takes looking at the ad account every day, even
• 36:57
  though you're not potentially doing anything
• 36:58
  every day, you're looking at the ad account every day.
• 37:01
  So you understand the benchmark of every metric.
• 37:04
  And then if every video that goes live has a 20% thumb stop rate and you launch
• 37:08
  some new ads with a different creator and some different
• 37:11
  hooks and it's, it's high production versus low production.
• 37:15
  You've never done that before.
• 37:16
  And then all of a sudden the thumb stop becomes 40% rather than 20%.
• 37:21
  It's like, hold on, let's see what happens here.
• 37:24
  Because, you know, we, we, it's not very
• 37:27
  often that performance dramatically improves.
• 37:29
  It's not a very often that you go from like a $200 CPA down to a $50 CPA, you know,
• 37:35
  doubling your thumb stop could be a sign of
• 37:39
  something that's really, that's really big to happen.
• 37:41
  Especially if you're making a big jump when it
• 37:44
  comes to the type of assets that are going live.
• 37:48
  We're not, we're not, we're not looking at thumb stop every day, but again, it's
• 37:50
  those benchmarks that, okay, you know, what's a
• 37:52
  benchmark click through rate in this ad account?
• 37:54
  Okay.
• 37:54
  It's about, it's about 1% normally.
• 37:56
  Okay.
• 37:57
  Let's just make sure it stays around there.
• 37:59
  I mean, so let's talk about this.
• 38:00
  So we're talking about, you know, to do creative testing properly, recapping your,
• 38:05
  you know, your, you're putting assets in a CBO, potentially by product, if you're
• 38:10
  able to do that or by funnel or by offer, maybe if you've expanded that way.
• 38:14
  There also, I think is this, you know, thought about, look, what, but, but I love
• 38:21
  an ABO because, you know, it's a container that it keeps it in.
• 38:27
  And, you know, I think it was Brad, your meta rep, right?
• 38:32
  Said something like outside of different bids, targeting or exclusions.
• 38:37
  It's a, it's just organization.
• 38:39
  That's all that an ad set is, right?
• 38:41
  We had two clients over a short window of time,
• 38:44
  basically say like, we want to try forced spend into this.
• 38:47
  And I said, I wouldn't recommend it.
• 38:49
  And I gave, gave the logic, but it's like, look, I'm comfortable doing it.
• 38:53
  But I'm, I'm going to tell you that, like, I don't, I don't
• 38:56
  think it's the move, but if you want to do it, let's do it.
• 38:59
  So we did it.
• 38:59
  And all we found was losers.
• 39:01
  It's like, and like, we were all making creative.
• 39:04
  Like, it wasn't just like, it wasn't like we were making shit ads.
• 39:06
  Like they weren't, it's like, we were all making a ton of ads.
• 39:09
  And what, what happened is we just ended up spending more into, into the worst ads
• 39:13
  when we, and then you look in the CBO and they
• 39:16
  just, they weren't getting, they weren't getting spent.
• 39:18
  And that happened, that happened twice
• 39:19
  where I tried the forced, forced spend method.
• 39:22
  And all I found was just force efficiency across the board.
• 39:25
  Or when we had something that looked like a winner and
• 39:27
  we dropped it in the CBO, it just still didn't get spent.
• 39:29
  Cause like, it was so contained in that small, that small window.
• 39:32
  And so, um, I messaged our rep cause I was
• 39:34
  just like, I need to know what's going on here.
• 39:37
  Cause I think the, the, the thought for a long time was if I use, because targeting
• 39:41
  is, is, is tinkered on, on an ad set level, because I can put an interest, because
• 39:45
  I can exclude people, because I can do this on an ad set level.
• 39:48
  And that has implications for learning.
• 39:49
  The ad set is where the learning is stored.
• 39:51
  I think that's what the thought was for a long time.
• 39:52
  I thought this for a long time until I saw a tweet from,
• 39:55
  from Yoni, who I don't think he's even on Twitter anymore.
• 39:57
  I don't even know where that guy went.
• 39:58
  He's, he's, he's in, he's in Canada and he's awesome.
• 40:00
  He's super helpful for meta.
• 40:01
  Um, what happened is he probably got a bunch of
• 40:03
  DMs asking for people to get their accounts unbanned.
• 40:06
  And he's like, I'm out.
• 40:07
  But I saw him say like, no, the, the actual ads, it doesn't matter.
• 40:10
  So I messaged our rep.
• 40:11
  I said, does the ad set do anything?
• 40:12
  Is there actual functional learning?
• 40:13
  If I put two ads that are speaking to
• 40:16
  completely different people in the same ad set,
• 40:17
  is that going to mess up my delivery on an ad set level?
• 40:20
  And his, his answer was no.
• 40:21
  Like it makes no difference.
• 40:22
  If you exclude people in a different way, or
• 40:24
  you, you set value-based rules on that ad set level,
• 40:26
  you can influence the way that delivery happens.
• 40:29
  But the ads themselves are delivering
• 40:31
  independent of each other within that to a degree,
• 40:34
  like independence and extreme works.
• 40:36
  Like if you have a video and an image, the video
• 40:38
  might drive more new, new, um, visitors to the website.
• 40:41
  But if you have completely different personas
• 40:43
  and an extreme example of this, we have a client,
• 40:45
  we had a client that sold this like body wash
• 40:47
  product and they had a, they had a collab and we had, um,
• 40:51
  creative for the girlfriend and the creative for the mom.
• 40:55
  And they delivered properly to those segments within the same ad set.
• 40:58
  And so I guess like the confirmation is like the ad set is functionally a folder
• 41:02
  outside of what you can do on the ad set level to
• 41:05
  influence targeting and bids and things like that.
• 41:08
  Now there's this feeling, you know, look, you look at an account and every brand
• 41:12
  owners here, you need to be testing more creative.
• 41:15
  You need to be trying more creative.
• 41:16
  That's, you know, huge narrative that we're, we're pushing as advertising agencies
• 41:20
  and, you know, because it's true and it's, it's needed in, in most cases.
• 41:27
  So how do you think about type Phil?
• 41:30
  Um, you've mentioned it previously a little
• 41:32
  bit, but like, you know, is it creative diversity?
• 41:37
  Uh, that you're just like, kind of like the way that you're doing, like, look, you
• 41:40
  need to, you know, develop personas and then, you know, you need to be testing
• 41:44
  video statics, need to be testing partnership,
• 41:47
  whitelisting, like, cause it's a diverse set of ideas.
• 41:51
  And, you know, within the same CBO, is it kind of that mentality or like, how do
• 41:55
  you determine a number or how much you should test creatively in terms of type?
• 41:59
  In terms of creative diversity across type of asset, so formats.
• 42:04
  Yeah.
• 42:05
  I mean, like, how does it, how do you determine that?
• 42:07
  Or is it just like, I fricking don't know.
• 42:09
  We should have five videos, five statics and five partnership ads.
• 42:14
  You know, like, what do you, like, what's the number and
• 42:17
  like, how do you determine how much you should be testing by type?
• 42:20
  Because I think we've all seen that we're testing down persona, but like what, what
• 42:25
  by type is the, is the number that makes sense?
• 42:28
  Yeah.
• 42:29
  Yeah. I mean, if you were starting, if you were starting at scratch, so you didn't have
• 42:33
  any data to start off with to tell you what type of asset you should be launching,
• 42:36
  I would be trying to be as equal as possible within reason.
• 42:41
  Probably gonna, you're probably gonna lean more into images just because they're,
• 42:44
  they're easier to, and cheaper to deliver and quite, and quicker to deliver.
• 42:47
  So, um, it could go, it could, it could go 50, 50, for example, 70, 70% statics,
• 42:54
  30% video is probably a good starting point.
• 42:57
  That's not going to cause too many issues for most brands.
• 43:00
  So, um, where we try to strategically do that
• 43:06
  is when we look at percent of spend per format.
• 43:10
  And the only time we'll ever do that really is when there's other, there's a reason
• 43:14
  to invest into a format that we're not currently spending on.
• 43:18
  Like, uh, okay, video shows some promise.
• 43:22
  We should be doing more.
• 43:23
  We should be creating more videos.
• 43:24
  If the data is showing this ad account never works with videos and statics
• 43:28
  absolutely crushing it, then we're not really going to be revisiting that.
• 43:34
  I'd say for a good year, you know, circle back at Q4, for example.
• 43:38
  Uh, so that's sort of like diversifying across formats.
• 43:42
  The, the second piece is, okay, so when do you go into it?
• 43:46
  Well, how do you go into videos?
• 43:48
  And a, and a logical way is you test an asset in statics, find some copy.
• 43:52
  Okay.
• 43:53
  How do we make that into a video?
• 43:55
  You know, we've got a headline, we've got a
• 43:57
  couple of bullet points on a static image ad.
• 43:59
  Let's just get the founder ripping a video,
• 44:01
  face the camera, talking through those points.
• 44:03
  And that would be a very straightforward way to start that.
• 44:06
  Uh, you've got justification from the static ad.
• 44:10
  Got spend and it performed the static ad.
• 44:13
  Let's turn it into a video.
• 44:14
  Does it get spend again?
• 44:16
  Um, and then we can get, we can get busy.
• 44:19
  We can get nerdy.
• 44:21
  Those different formats spend in different placements.
• 44:24
  That would be great.
• 44:25
  You know, the static goes to more Facebook feed, desktop.
• 44:28
  The video goes to more Instagram reels and stories.
• 44:31
  Great.
• 44:32
  We're, we're active across all placements where audiences are.
• 44:34
  The partnership ad is really, is really just like all chips on black, to be honest,
• 44:38
  because every time we test them, it's all anyone,
• 44:41
  I mean, it's all anyone's ever banging on about.
• 44:43
  Very rarely.
• 44:45
  I don't think we've ever actually seen a brand who's really invested in partnership
• 44:50
  ads from a volume and time point of view,
• 44:53
  where it hasn't turned out to be a fruitful.
• 44:56
  So it's worth, it's worth it to you at this point in time.
• 44:59
  As part of how you do creative testing.
• 45:01
  Yeah.
• 45:02
  But it does fail if you just try it once and it doesn't work.
• 45:05
  And then you stop.
• 45:06
  I think that that's what you're saying is like the volume.
• 45:07
  It's like the time that they put in.
• 45:09
  So they, they figured out who is worth partnering
• 45:11
  with, take the time to make those relationships,
• 45:13
  get the content, brief them thoughtfully, and
• 45:15
  then get it in the ad account as a result of that.
• 45:17
  All right.
• 45:17
  Yeah.
• 45:18
  Yeah.
• 45:18
  Yeah.
• 45:19
  So we're like where we want to get to the sort of the, the, the destination that we
• 45:24
  want to get to is where we've got new
• 45:25
  partner content that we can pull live every week.
• 45:28
  And that could just be free for ads.
• 45:30
  It doesn't need to be any more than that, but we're doing it every week.
• 45:33
  Like clockwork.
• 45:35
  And it'll really compound over time.
• 45:36
  I think it's good to bring us to this segment
• 45:39
  in the show where we call, we call this beefs.
• 45:42
  And this is a time to sort of take out the trash.
• 45:44
  If you have any, you know, I think my beef with creative testing, you know, really
• 45:49
  is that there's this idea that it has to be scientifically done.
• 45:54
  And you kind of mentioned this previously, right?
• 45:56
  And I think that's how we as marketers were taught to think
• 46:00
  and how we were brought up of, okay, it has to make sense.
• 46:04
  It has to be a clean test.
• 46:05
  It has to be statistically significant, et cetera, et cetera.
• 46:09
  And the reality of it is meta is now serving, whether it's an ABO or a CBO to
• 46:14
  Brad's earlier point, they're serving it to the proper audience, looking at the
• 46:18
  content of the ad and the landing page that it's being sent to.
• 46:21
  So the more that you can consolidate and test alongside things that are getting a
• 46:28
  lot of spend, it doesn't feel like that's the right thing to do.
• 46:32
  But that ultimately is actually the right thing to do in 2026, I think, for
• 46:36
  creative testing and also diversifying on persona
• 46:39
  and the things we've talked about, creative type.
• 46:41
  But, you know, that's my beef is that it
• 46:44
  has to be this like clean scientific thing.
• 46:47
  And it's like it doesn't.
• 46:48
  It's actually like a lot simpler.
• 46:49
  It's what Brad told me via DM last week.
• 46:52
  Like, just fucking rip it.
• 46:54
  That's like, you know, so like, I don't know.
• 46:57
  What do you guys any beefs on creative
• 46:59
  testing or anything you guys want to get into?
• 47:01
  I would just say it feels like when people start to
• 47:03
  talk about creative testing, they are in these camps.
• 47:06
  And like, I'm cool with you being in either camp.
• 47:08
  But like most of the people end up talking past each other.
• 47:11
  Because like what we're talking about is being thoughtful about the types of
• 47:14
  creative that's going into the ad account.
• 47:15
  And then Phil, to your earlier points, like spending more time on the things that
• 47:19
  actually influence the ad account performance, which is like
• 47:22
  the analysis of the creative, the partnership focus and ads.
• 47:26
  And how do we get more of that?
• 47:27
  How do we get more content?
• 47:28
  How do we set up these systems?
• 47:29
  How are we thinking about the landing pages?
• 47:31
  Is the ad to landing page to offer consistent with the person?
• 47:34
  It's like, if you set up creative testing in a way where it's like you have a
• 47:37
  system for getting it live and paying attention to it and analyzing it, reviewing
• 47:41
  it and doing something with it, then it shouldn't be some like complex, crazy
• 47:45
  system that, you know, is like so confusing that you just don't do it.
• 47:48
  But I guess my beef is like most people are just talking past
• 47:51
  each other when they argue about creative testing structures.
• 47:54
  Because like what we're all saying is, and this is true of the, this is true of
• 47:57
  cost caps, bid caps, lowest costs argument as well.
• 48:00
  It's like, well, all of all, most of us are saying that generally know what they're
• 48:05
  talking about is like, you should focus on the things that actually move the
• 48:07
  business forward, which is probably not the nuance between a cost cap, a bid cap, a
• 48:11
  lowest cost and all that dumb shit that we argue about.
• 48:14
  I've got tons of beefs. I'm quite a, I'm quite a, I'm quite a mind mannered, happy
• 48:19
  person, but yeah, I've got tons of beefs.
• 48:22
  Specifically on creative testing, I think my beef is on, we've moved away from how
• 48:28
  much spend you've measured as a size of your, you know, sort of, that's the wrong
• 48:34
  thing to say, but you know, like your talent level is how much spend you manage.
• 48:39
  And now the thing to measure on is how many assets you
• 48:42
  launch. And there's always going to be someone with a faster car.
• 48:46
  There's always going to be someone who's launching more ads than you. You know,
• 48:49
  it's like, I launched 5,000 ads last month.
• 48:52
  It's like, someone's going to say a bigger number,
• 48:54
  but we've, we've got, we've got a brand at the moment.
• 48:57
  I don't think we've launched an ad this month, this year. I don't think we've
• 49:01
  already launched any in January, sorry, in December.
• 49:04
  They're, they're growing massively, like 50% year over year. They're like, I think
• 49:09
  they're like, they did 20 million last year.
• 49:11
  They're going to try and double again this year.
• 49:12
  They're spending like six, 7k a day in January.
• 49:16
  They're a massive Q4 tax season brand, which is obviously coming up. And it's
• 49:21
  nothing, it's nothing that, that gets the
• 49:25
  tactics and nothing that gets engagement on Twitter.
• 49:29
  It's everything but that.
• 49:30
  That's not fun.
• 49:31
  I mean, obviously the model is built, of course, as we've talked about that, it's
• 49:34
  like shock and awe, try to make people, you know, feel like an idiot and feel like
• 49:40
  they have FOMO and then like sell your product and that you have all the answers.
• 49:44
  You know, I think, yeah, it, it doesn't give enough credit to the fact that most of
• 49:49
  us are so busy on a daily basis running around trying to just like keep the ship
• 49:54
  running on the brand or the agency or with
• 49:57
  clients and bringing them like good creative.
• 50:01
  That's going to scale on a regular basis that
• 50:04
  like this other stuff on X is like kind of a joke.
• 50:07
  It's like, here's my seven step process of how I, you
• 50:10
  know, and you're like, all right, like nobody's, it's not real.
• 50:14
  It's just fake.
• 50:16
  I have spoken to a couple of brand owners or founders over the last couple of weeks
• 50:21
  where they felt really guilty because they're
• 50:24
  not hitting this volume of ads that's going live.
• 50:26
  Because, you know, I've got to work on some new creative.
• 50:29
  I've got to work on some new creative and
• 50:30
  instead of launching any, they're just launching none.
• 50:34
  And my recommendation was always the same.
• 50:35
  It's just like, forget about trying to
• 50:37
  launch a certain number and just work on one ad.
• 50:40
  Yeah.
• 50:41
  Don't, don't try and do four with different hooks.
• 50:44
  You know, we're going to test the hook.
• 50:45
  And there's like Andrew was saying, you know, there's this scientific test.
• 50:48
  Just create one piece of content as if it was going to be posted organically.
• 50:53
  And instead of posting it to organic, just upload it to the outcome.
• 50:58
  Yeah.
• 50:59
  I think it's great advice.
• 51:00
  You just got to set up the system that
• 51:01
  you'll be, it's like a, it's like diet advice.
• 51:03
  It's like pick the diet that you'll be, you'll adhere to.
• 51:06
  At least at first you can optimize from there, but it's like, you got to at least,
• 51:09
  you got to at least build them to build the muscle.
• 51:11
  Yeah.
• 51:12
  Yeah.
• 51:12
  The system is, is interesting.
• 51:14
  With most.
• 51:15
  Okay.
• 51:16
  One more question makes me think about this.
• 51:18
  With most of the clients that you guys have on creative testing, is there, like you
• 51:24
  said in the system, is there an amount or is there, is it, or is it just about
• 51:28
  like, we need to get net new ideas into the pipeline.
• 51:32
  And if we're doing that on a weekly basis, as
• 51:36
  long as it's not more than zero, we are okay.
• 51:39
  How do you approach telling clients about that?
• 51:42
  So we've, we have a default starting point, which is eight to 10 concepts a month.
• 51:47
  And we certainly focus on the number of concepts.
• 51:49
  It's like the briefs.
• 51:50
  Start with the number of briefs that we've got to write.
• 51:52
  Not to the clients then say, well, how many, how many assets is that going to be?
• 51:56
  And I try not to answer that, but it's usually around five to six assets, five to
• 52:00
  six assets, a brief to call that 40 to 50 assets a month.
• 52:08
  And I think if you're being really intentional about the briefs and the variation
• 52:12
  between those assets in each brief, then that
• 52:17
  should be, that should be enough for most brands.
• 52:20
  We could, we could talk about spend figure
• 52:22
  where you might need to start doing more.
• 52:25
  I think in most cases, if you're not doing that, then that's a good starting point.
• 52:29
  Yeah.
• 52:29
  Was that, is that like the situation you find yourself in?
• 52:31
  It's like when you're doing these audits and you're talking to these brands, it's
• 52:34
  like, okay, you've launched three things in the last month.
• 52:36
  Like you don't, you just haven't, you don't have the baseline.
• 52:38
  So like our baseline is taking you away above where you are today.
• 52:42
  No, for most of them, they're launching too many.
• 52:44
  Really?
• 52:45
  Yeah.
• 52:45
  Like I ordered, ordered an account the
• 52:47
  other day where, and I've got the transcript.
• 52:49
  So maybe I can like turn the transcript into
• 52:51
  like an AI piece and then drop it in Foxwell.
• 52:53
  Well, that's, that's quite helpful.
• 52:55
  Usually it's use.
• 52:56
  That's quite helpful usually to get the sort of points
• 52:59
  across, but they were launching a new ad set every day.
• 53:02
  Sure.
• 53:04
  With new assets in.
• 53:05
  And I'm like, when was the last time you launched an ad that would?
• 53:08
  I'm like, I don't know.
• 53:09
  It's like, well, why are you still doing it?
• 53:11
  So.
• 53:11
  Yeah.
• 53:12
  They're trying to, I mean, I think to some degree, right?
• 53:14
  Like people in their brain want a system.
• 53:17
  And I think that their ad sets are a clean way to do that.
• 53:19
  And that's what's also was the predominant
• 53:21
  knowledge and the prevailing knowledge for a long time.
• 53:23
  So you can't really like blame people for, for thinking this way.
• 53:27
  Um, because it's what makes sense to them and it's what they're trying to do.
• 53:32
  We're all trying to build a system.
• 53:33
  The system is, yeah, I think more of what you said.
• 53:35
  It's around different concepts with briefs and then, or, you know, different sort
• 53:40
  of pitch towards different personas and whatever.
• 53:43
  And then we develop the ads around that.
• 53:46
  And that makes sense.
• 53:46
  So look, great episode, uh, almost an hour.
• 53:49
  Thank you to those of you that listened all the way through.
• 53:53
  And Phil, thank you for joining us late night.
• 53:55
  Hopefully you can get some good rest tonight after now you're stoked after this.
• 53:59
  You're just going to have to go do a workout.
• 54:01
  Brad, as always a pleasure.
• 54:03
  And, uh, thank you for listening.
• 54:05
  If you like what you hear, please rate us on all of the
• 54:09
  platforms because we are lonely and need to know that.
• 54:13
  We are validated by your votes.
• 54:18
  So like, comment, share, whatever, all that shit.
• 54:21
  We appreciate it to the tens of listeners out there.
• 54:24
  And, uh, yeah, look forward to next time.
• 54:27
  Adios.
• 54:31
  This episode is brought to you by Brad's company, Work Marketing.
• 54:36
  If you need a D2C marketing agency, let me tell you, Homestead is great,
• 54:40
  but Work Marketing is also fantastic.
• 54:43
  And let me tell you, you aren't going to find
• 54:46
  friendlier people out there, uh, in the e-commerce space.
• 54:50
  So we decided to do these little ads for each other's companies.
• 54:53
  So hopefully you find it interesting, but seriously, great team at Work Marketing.
• 54:57
  Very smart.
• 54:58
  Brad and Jordan are incredibly dialed in.
• 55:00
  Um, I just gave them a lead, already made this
• 55:03
  brand that I gave them, like, I don't even know,
• 55:05
  double, double revenue that they had in the previous month or something.
• 55:09
  So, you know, it's very exciting to, to be connected with Brad.
• 55:13
  And if you need a great agency, there's really no one better.
• 55:16
  Uh, Zach, anything to comment on Work Marketing?
• 55:18
  Yeah.
• 55:19
  I mean, if you want an agency that cares about your
• 55:21
  business much more than they care about their own website,
• 55:24
  I just tried to load workmarketing.com and it was broken.
• 55:27
  So they're definitely going to give more of
• 55:28
  a shit about your business than their own.
• 55:30
  So I highly recommend Brad and the team over at work.
• 55:33
  They've been incredible.
• 55:34
  We've referred a lot of business over to them as well.
• 55:36
  Really, really good as far as like cracking
• 55:38
  funnels and figuring out like rapid growth for brands.
• 55:41
  So I highly recommend these guys.
• 55:42
  The only way that we grow this podcast is by you sharing it with your friends.
• 55:51
  Honestly, like reviews kind of don't really mean anything too much anymore.
• 55:54
  They're really meaningful, but they don't do a lot for the growth of the podcast.
• 55:58
  Um, and so sharing YouTube links, sharing Spotify links,
• 56:01
  sharing Apple, whatever we call it under the podcast app.
• 56:05
  Now, anything you can share, the better we're going to be.
• 56:09
  Guys, anything else you want to say on this?
• 56:11
  Yeah, please go check us out on YouTube.
• 56:13
  Rack up those views for us.
• 56:14
  We'd love to see it.
• 56:15
  And then subscribe.
• 56:16
  Make sure to subscribe on YouTube as well.
• 56:17
  And I relentlessly refresh the YouTube comments
• 56:19
  because it dictates my mental health for the day.
• 56:21
  So please say something nice about all of us.
• 56:24
  Thank you, everyone.
• 56:25
  Thanks for listening.
• 56:26
  Honestly.