Latest / Scalability School / Inside The Ad Account: How To Do Creative Testing In 2026

Transcript

Inside The Ad Account: How To Do Creative Testing In 2026 · Scalability School

• 0:00
  I've been reading it verbatim, but it's
• 0:01
  part of a broader discussion.
• 0:03
  But he said more
• 0:03
  substantial concept differentiation,
• 0:05
  assets with a high degree of visual
• 0:06
  conceptual contrast, distinct messages.
• 0:08
  This is creative diversity.
• 0:09
  What he is talking about is creative
• 0:10
  diversity back in 2023.
• 0:12
  And I know it feels like it's
• 0:13
  been a long time since 2023,
• 0:14
  but we've been talking
• 0:14
  about creative, not me and we,
• 0:16
  but like the
• 0:16
  internet, Twitter, e-commerce,
• 0:17
  Twitter has been talking
• 0:18
  about this for years now.
• 0:19
  And I think that's really easy to forget.
• 0:20
  The bet that we're making is that meta is
• 0:22
  going to be right on the allocation
• 0:23
  more often than we are.
• 0:25
  And I know there's conversations from
• 0:26
  people I respect very heavily
• 0:27
  on Twitter and in this space that say,
• 0:30
  well, if you have a high hit
• 0:31
  rate, then it doesn't matter.
• 0:32
  And I just, I'm not
• 0:33
  willing to make that bet.
• 0:34
  And I'm not just, we're not just throwing
• 0:35
  creative unintentionally up
• 0:37
  and saying, well, if
• 0:38
  it spends, it spends.
• 0:39
  But it is freeing in a
• 0:41
  way to know that you can,
• 0:42
  you can push the volume and know that if
• 0:45
  it's not going to deliver,
• 0:46
  it just won't deliver.
• 0:47
  Now, it doesn't mean be lazy on your ads
• 0:48
  and your creative strategy.
• 0:49
  But what it does mean
• 0:50
  is you can know that
• 0:51
  you don't have to try to
• 0:52
  force a spend into something.
• 0:53
  And that's frustrating.
• 0:54
  It can be a very frustrating thing.
• 0:55
  I've got one campaign
• 0:56
  and I've got two ad sets.
• 0:58
  And if you look at the top campaign,
• 0:59
  there's one thing that stands out.
• 1:00
  It is spent, what is that,
• 1:01
  like four times more budget
• 1:03
  than the campaign below it.
• 1:04
  And it has a ROAS of a whole point lower.
• 1:06
  So top campaign, if you're
• 1:07
  listening, spent $197,000.
• 1:09
  Bottom campaign spent $55,000.
• 1:11
  Top campaign performing at a 2.63.
• 1:13
  Bottom campaign performing at a 3.4.
• 1:16
  Oftentimes people see that and they say,
• 1:17
  Fred, what are you doing?
• 1:19
  Like why are you not scaling more into
• 1:21
  that specific ad set?
• 1:21
  Like you're leaving money on the table.
• 1:23
  It's performing 50% more
• 1:24
  efficient than the other one.
• 1:26
  And my answer is, well,
• 1:27
  Meta knows something that we don't
• 1:29
  about why that can't spend more.
• 1:30
  And what we've found is
• 1:31
  that there are usually
• 1:32
  some signals that give us
• 1:33
  context into what that is.
• 1:34
  Now, we can't say for sure
• 1:35
  because Meta is a
• 1:36
  little bit of a black box,
• 1:37
  but we can look at some
• 1:38
  other metrics to tell us,
• 1:38
  well, why is Meta making this decision
• 1:40
  that they're making?
• 1:40
  And now, let's take a listen to the
• 1:49
  Scalability School podcast.
• 1:51
  Hey, everybody.
• 1:51
  Welcome back to another episode of the
• 1:53
  Scalability School podcast.
• 1:55
  I am flying solo today to talk about all
• 1:57
  things creative
• 1:58
  testing in 2026 through Meta.
• 2:01
  I'll move my face out of the way and jump
• 2:02
  into the presentation in a second.
• 2:05
  Just want to say, glad you're here and
• 2:07
  excited to go through this.
• 2:08
  I know Andrew is off in, I believe,
• 2:10
  Lisbon hosting a Foxwell Founders event.
• 2:14
  He texted me this morning saying it's
• 2:15
  going amazingly well.
• 2:16
  And so really excited for him.
• 2:18
  And I'm sure Zach is off somewhere in the
• 2:19
  world slaying in socks.
• 2:21
  So I'm going to hold down
• 2:22
  the floor today by myself
• 2:23
  and jump into this
• 2:25
  creative testing deep dive.
• 2:27
  So we did a podcast episode a couple of
• 2:30
  weeks back with Phil.
• 2:32
  It was fantastic.
• 2:33
  I've got it open here.
• 2:33
  If you're on YouTube, you can see this.
• 2:35
  If you haven't watched it, I would
• 2:36
  recommend pausing this,
• 2:37
  going back and watching that
• 2:38
  because that's kind of the setup to this.
• 2:41
  This episode is going to
• 2:42
  be a bit more tactical.
• 2:43
  I'm actually going to show you
• 2:43
  screenshots from ad accounts
• 2:44
  and show you how we do the setup
• 2:46
  and just be a little bit more specific
• 2:47
  with here's actually
• 2:48
  exactly how we do this.
• 2:50
  Whereas in that episode, we talked a
• 2:52
  little bit more and
• 2:52
  kind of Phil and I agree,
• 2:55
  but we debated a bit more about what the
• 2:56
  options are and what
• 2:57
  people do for creative testing.
• 2:58
  And in a follow-up to that episode, this
• 3:00
  episode you're listening
• 3:02
  to right now is a follow-up
• 3:03
  to a follow-up because I made
• 3:05
  a follow-up post on X Twitter.
• 3:07
  I don't call it X.
• 3:08
  I don't like saying that.
• 3:10
  But I made a follow-up post and an
• 3:12
  article on Twitter talking
• 3:14
  about exactly this and kind
• 3:15
  of how we do this.
• 3:16
  So I'm taking this, but I'm adding a
• 3:17
  little bit more sauce
• 3:18
  to the podcast episode.
• 3:20
  And I'm going to go into this.
• 3:21
  This was both that episode as well as my
• 3:24
  article were some of the most popular
• 3:26
  episodes that we've,
• 3:28
  or things that we've put out that we kind
• 3:29
  of have experienced.
• 3:30
  And so people are very interested in this
• 3:32
  information, it seems.
• 3:33
  And so I wanted to
• 3:34
  make a follow-up to that.
• 3:37
  I will say before we jump in, and then
• 3:38
  I'm going to get into this,
• 3:39
  and I'm going to rip through and try to
• 3:40
  make this the most tactical
• 3:42
  second-by-second piece possible.
• 3:44
  But I do notice that a lot of people kind
• 3:46
  of get really stuck up on
• 3:47
  like small tactic things.
• 3:50
  I wouldn't say this is small necessarily.
• 3:52
  And I know that there are different
• 3:53
  opinions about how to do this.
• 3:55
  I think the most important thing is not
• 3:57
  to be dogmatic about
• 3:59
  what the right option is.
• 4:01
  And at the same time, you should have an
• 4:03
  opinion about how
• 4:03
  creative testing should be done,
• 4:05
  because then you can focus on the things
• 4:07
  that matter as
• 4:07
  opposed to being stuck into,
• 4:09
  you know, constantly trying to change up
• 4:12
  your ad account structure to see if
• 4:14
  you're missing out on
• 4:15
  what's next for creative testing.
• 4:17
  So pick a lane and stick with it and
• 4:19
  reevaluate every maybe couple of months
• 4:21
  or quarterly or once a year.
• 4:24
  This is not something that you probably
• 4:25
  need to be changing,
• 4:26
  certainly not weekly or even monthly.
• 4:28
  So this is how we do meta creative
• 4:31
  testing in 2026 across over
• 4:33
  eight figures a year in meta ads.
• 4:35
  So let's go ahead and
• 4:35
  jump into this thing.
• 4:36
  So the two camps that people fall in, and
• 4:38
  we talked about this
• 4:39
  in the previous episode.
• 4:40
  Also, if you're not watching this on
• 4:41
  YouTube, I'm going to be showing some
• 4:42
  visuals and kind of
• 4:43
  giving a presentation.
• 4:44
  So I would suggest popping over to that.
• 4:46
  I'll do my best to explain things that
• 4:48
  are on screen, but
• 4:49
  just for clarity's sake.
• 4:51
  There's two camps that people fall into.
• 4:52
  People generally fall into the camps of
• 4:54
  either you want to force spend into
• 4:57
  creative tests because you invested in
• 4:58
  creative and you want to know, is it
• 5:00
  going to work or not?
• 5:00
  And the other camp is letting meta
• 5:03
  allocate budget between the creatives
• 5:05
  that you're launching and the existing
• 5:06
  ads that you already
• 5:07
  have in the ad account.
• 5:08
  That's kind of
• 5:08
  fundamentally what it boils down to.
• 5:10
  Now, there are different ways that that
• 5:11
  shows up in the ad account.
• 5:12
  And the way that we like to do this is a
• 5:14
  single CBO campaign per product or offer.
• 5:18
  Right.
• 5:18
  And so let's say we're selling pants.
• 5:20
  We will set up a single campaign for
• 5:22
  those pair of pants.
• 5:23
  Let's say we were selling a pant bundle.
• 5:25
  We would make a new
• 5:25
  campaign for that pant bundle.
• 5:27
  Right.
• 5:27
  And the reason that we do that is because
• 5:29
  there's a different value assigned to
• 5:31
  that purchase event.
• 5:33
  And so we want to hold those things
• 5:34
  separate and give them space to to
• 5:36
  breathe and test and set targets specific
• 5:39
  to the offers and the product.
• 5:41
  And so there are definitely cases for
• 5:42
  expanding out from one campaign.
• 5:44
  But the way that you can think about this
• 5:45
  is if you are a single SKU brand, you can
• 5:48
  get away most times with a single
• 5:50
  campaign with some caveats.
• 5:51
  The other caveats, and there's more, but
• 5:53
  one of the more common caveats is if
• 5:55
  you're running a cost based bidding.
• 5:57
  So whether that's highest volume cost
• 6:00
  cap, bid cap or a value based bidding, a
• 6:02
  min ROAS or highest value, you can have
• 6:05
  two campaigns that are basically the
• 6:06
  exact same structure.
• 6:07
  The bidding type
• 6:08
  descends it being different.
• 6:09
  So that's another way that shows up.
• 6:10
  So here's an example of what it looks
• 6:12
  like in one of our ad accounts.
• 6:13
  This is a little bit small, but this is
• 6:14
  basically the entirety of the ad account.
• 6:16
  Right.
• 6:16
  It's a single CBO that spent 400, a
• 6:19
  little over 400 grand
• 6:20
  in this 30 day stretch.
• 6:21
  And it's a, it's a,
• 6:22
  it's a single campaign.
• 6:23
  Right.
• 6:24
  And so it's running on broad targeting.
• 6:26
  We are, it's a, it's a
• 6:28
  single, single SKU or offer.
• 6:30
  Right.
• 6:30
  So the offer is the same throughout this,
• 6:32
  this entire campaign on
• 6:33
  that specific product.
• 6:34
  We're not running, you know, 50, you
• 6:36
  know, $50 off and, you know, 20% off or
• 6:39
  something, you know,
• 6:40
  like it's consistent.
• 6:41
  We do exclude lifetime customers.
• 6:44
  So there's a couple of
• 6:44
  ways that we do that.
• 6:45
  I'll go from, this
• 6:46
  isn't on my screenshot here.
• 6:47
  It's not, it's not listed from worst to
• 6:50
  best, but I'll kind of explain
• 6:52
  them from, from worst to best.
• 6:54
  Pixel purchases 180 days.
• 6:55
  So that's using Meta's
• 6:56
  pixel based audience.
• 6:57
  It's fine, but it's not great.
• 6:59
  Using Klaviyo lists are also nice because
• 7:02
  they can sync up and be automated.
• 7:03
  And it's also included
• 7:04
  in your Klaviyo plan.
• 7:05
  If you have a Klaviyo plan
• 7:06
  and then Shopify audiences.
• 7:07
  So Shopify audiences and Klaviyo
• 7:09
  audiences are similar in that they're
• 7:10
  both first party data.
• 7:12
  Shopify audiences just tend to be better.
• 7:14
  You know, in my experience from an
• 7:15
  exclusion perspective.
• 7:16
  There's a time and a place to target
• 7:17
  repeat customers, but we're not going to
• 7:19
  talk about that in this video.
• 7:21
  Probably.
• 7:22
  I may come back to it at the end.
• 7:23
  But for the most part, we are trying to
• 7:25
  reach net new people to the brand and net
• 7:28
  new customers more, more, more precisely.
• 7:30
  I don't know that I'll talk about
• 7:31
  retargeting directly in this video, but I
• 7:33
  may, I may touch on it a bit.
• 7:35
  And so that's, that's how we set that up.
• 7:37
  And the bet that we're making, right?
• 7:38
  The bet that we're making with this
• 7:39
  campaign setup is that Meta is going to
• 7:41
  more often, that's an important point,
• 7:43
  make a better allocation of the dollars
• 7:45
  than, than the human
• 7:46
  deciding to run creative tests.
• 7:48
  And again, that's, that's the bet that
• 7:49
  we're making here, as opposed to running
• 7:51
  an ABO setup where we're forcing ad spend
• 7:53
  or even a CBO with mid budgets or even
• 7:55
  separate campaigns for
• 7:56
  separate types of creative.
• 7:58
  The bet that we're making is that Meta is
• 7:59
  just going to be right more often than we
• 8:01
  are, assuming that we give them clear
• 8:03
  expectations of what our goals are.
• 8:04
  So that's, that's the premise.
• 8:06
  So what that means then is that ad sets
• 8:08
  are just functionally become folders and
• 8:09
  folders is actually even generous in the
• 8:11
  way that we set this up, which you'll see
• 8:12
  in a second in my screenshot.
• 8:14
  But unless you're using ad sets to define
• 8:16
  audiences, like maybe you are using
• 8:18
  lookalike audiences or you are using
• 8:20
  interest-based audiences, or you're
• 8:21
  trying to reach a specific cohort or, or
• 8:24
  age demographic or even
• 8:25
  geographic demo, right?
• 8:27
  That's, that, that's a caveat that I
• 8:28
  would, I would set
• 8:29
  aside for a different point.
• 8:30
  But if, if you're not using that, then ad
• 8:32
  sets functionally just become folders.
• 8:34
  And they're not, although you can dictate
• 8:37
  the targeting on the ad set level, if you
• 8:39
  just have two broad ad sets that are set
• 8:41
  up with the exact same, with the same
• 8:42
  exclusions, they
• 8:43
  functionally just become folders.
• 8:45
  They're just containers for that.
• 8:47
  Now, some people will launch new creative
• 8:49
  batches into their own ad sets or unique
• 8:52
  customer personas into their own ad sets.
• 8:55
  The only pushback that, and the reason I
• 8:56
  pushed back on the idea of doing that is
• 8:58
  because, again, if, assuming you're going
• 9:00
  broad, which we would suggest doing,
• 9:02
  assuming you're going broad, launching a
• 9:05
  new ad set doesn't tell meta to go target
• 9:07
  anyone in particular on, on broad.
• 9:10
  That is done on the creative level.
• 9:11
  So you could have two ads that speak to
• 9:13
  completely different customers in the
• 9:14
  same ad set, and they will reach the
• 9:16
  person that they're supposed to reach
• 9:17
  after a little bit of time, right?
• 9:18
  Like, I'll take a little bit of time for
• 9:19
  meta to pick up and get some conversions
• 9:20
  and start to go down the right route.
• 9:22
  And a lot of people hesitate, and they
• 9:24
  say, well, learning is stored at the ad
• 9:26
  set level, and that's how
• 9:26
  meta is going to make decisions.
• 9:28
  That is not true.
• 9:29
  So you can take solace.
• 9:31
  I don't know if that's the right word
• 9:32
  here, but you can take comfort in the
• 9:33
  fact that if you have an ad targeting a
• 9:36
  65-plus-year-old female and an
• 9:39
  18-year-old dude in the same ad set, they
• 9:41
  will reach the person that they are
• 9:42
  supposed to reach, especially true if you
• 9:44
  have, like, landing pages that are super
• 9:46
  dialed for them, which is beyond the
• 9:48
  topic of this video.
• 9:49
  So ad sets functionally become folders.
• 9:50
  And so my face is going to be in the way
• 9:52
  of some of this, but that's okay.
• 9:54
  So why do we have
• 9:54
  multiple ad sets then, right?
• 9:56
  So the simplest version of this, if you
• 9:57
  were setting this up today from scratch,
• 9:59
  you would have a single campaign with a
• 10:00
  single ad set, and you would continue to
• 10:02
  add ads in there until
• 10:03
  you hit the max limit.
• 10:05
  Sometimes that's 50 ads.
• 10:06
  Sometimes it's 150.
• 10:08
  This seems like it differs by ad
• 10:09
  accounts, and they slowly have been
• 10:11
  rolling this out since the introduction
• 10:12
  of ASC back in the day.
• 10:13
  But once you hit the max
• 10:16
  limit, then you add a new ad set.
• 10:18
  And, you know, people, I think people get
• 10:20
  a little bit stressed out by this because
• 10:21
  what that means is, if you're looking at
• 10:23
  my screenshot here, I have five active ad
• 10:25
  sets with spend to varying degrees and
• 10:28
  different numbers of orders.
• 10:29
  The numbers in this case is just the
• 10:31
  order in which we launched them.
• 10:32
  And so you see that ad set four has the
• 10:34
  most ad spend, and that's because the ads
• 10:35
  in there are the best.
• 10:36
  We don't have a min spend
• 10:37
  forcing budget in that direction.
• 10:39
  Meta has allocated a budget between these
• 10:41
  in the way that you see
• 10:42
  them showing up in here.
• 10:43
  And so I think that that
• 10:44
  is very helpful context.
• 10:45
  And so the question that I had before we
• 10:48
  committed to doing this, and we've been
• 10:49
  running this setup for a little over
• 10:51
  three years now, we used to segment the
• 10:53
  ad sets by creative batches or by
• 10:55
  persona, which I talked about previously.
• 10:57
  But I had a conversation with a meta rep
• 10:59
  on Twitter and then
• 11:00
  separately about does it matter?
• 11:02
  Like, do we need to be splitting creative
• 11:04
  by these different themes, especially if
• 11:05
  we're just running broad targeting?
• 11:07
  And they said, no, these are just
• 11:09
  functioning folders.
• 11:09
  And so what I want to do is show you that
• 11:11
  this is from 2023, and it hasn't changed
• 11:13
  since then because I had a conversation
• 11:14
  more recently with our direct meta rep.
• 11:16
  And so this tweet from Yoni, who I
• 11:18
  believe is still a meta, he said, more
• 11:20
  substantial concept differentiation is
• 11:22
  one of the ways to get this.
• 11:23
  I'm not reading it verbatim, but it's
• 11:25
  part of a broader discussion.
• 11:26
  But he said, more substantial concept
• 11:28
  differentiation, assets with a high
• 11:29
  degree of visual conceptual
• 11:31
  contrast, distinct messages.
• 11:33
  This is creative diversity.
• 11:34
  What he is talking about is creative
• 11:35
  diversity back in 2023.
• 11:36
  And I know it feels like it's been a long
• 11:38
  time since 2023, but we've been talking
• 11:39
  about creative, not me and we, but like
• 11:41
  the internet, Twitter, e-com Twitter has
• 11:43
  been talking about this for years now.
• 11:45
  And I think it's really easy to forget.
• 11:46
  And I asked the follow up.
• 11:47
  I said, so different messages to
• 11:49
  different people causes no issues with
• 11:50
  the delivery of the ad set, even if those
• 11:52
  people are widely different demos.
• 11:54
  So what I'm asking is, can I put my two
• 11:56
  different ads speaking to two different
• 11:57
  people in the same ad set?
• 11:58
  And he said, yes, that's the point.
• 12:00
  Fuel the system with more ad variety to
• 12:01
  reach a broader and more diverse audience
• 12:03
  of prospects with a
• 12:04
  propensity to convert.
• 12:05
  Give the platform more flexibility to
• 12:07
  decide where the next best dollar is
• 12:08
  spent across a wider range of ads.
• 12:10
  Right.
• 12:10
  And so we lean into that.
• 12:11
  And that's how we decided
• 12:12
  to set up the ad accounts.
• 12:14
  Now, there are some
• 12:15
  objections that come with this.
• 12:17
  And one of the ones that we hear all the
• 12:18
  time is if I don't force spend with ABO,
• 12:20
  how do I know that my
• 12:21
  ads will get a fair shot?
• 12:22
  Well, we know is that spend
• 12:24
  is signal as a result of this.
• 12:26
  And again, as I started the video by
• 12:27
  saying, I started this episode by saying
• 12:29
  the bet that we're making is not that we
• 12:31
  are always it's not even that the meta is
• 12:34
  always going to be right.
• 12:35
  The bet that we're making is that meta is
• 12:37
  going to be right on the
• 12:38
  allocation more often than we are.
• 12:40
  And I know there's conversations from
• 12:42
  people I respect very heavily on Twitter
• 12:45
  and in the space that say, well, if you
• 12:48
  have a high hit rate,
• 12:49
  then it doesn't matter.
• 12:49
  And I just I'm not
• 12:50
  willing to make that bet.
• 12:52
  And I'm not just we're not just throwing
• 12:54
  creative unintentionally up and saying,
• 12:56
  well, if it spends, it spends.
• 12:57
  But it is freeing in a way to know that
• 13:00
  you can you can push the push the volume
• 13:03
  and know that if it's not going to
• 13:05
  deliver, it just won't deliver.
• 13:07
  Now, it doesn't mean to be lazy on your
• 13:08
  ads and your creative strategy.
• 13:09
  But what it does mean is you can know
• 13:11
  that you don't have to try
• 13:13
  to force spend into something.
• 13:14
  And that's frustrating.
• 13:15
  It can be a very frustrating thing to
• 13:16
  spend a bunch of time on a creative
• 13:17
  concept or a bunch of
• 13:18
  money on a creative concept.
• 13:20
  I again, my two cents
• 13:21
  is that it's a sunk.
• 13:22
  It's you don't want to fall victim to
• 13:25
  sunk cost fallacy and just spend it deep,
• 13:28
  spend yourself deeper into the hole
• 13:29
  because meta doesn't
• 13:30
  believe in the concept.
• 13:31
  Because even if you even if meta is right
• 13:33
  or even if you're right one time for
• 13:38
  every time that meta is right 10 times,
• 13:41
  then you could still be in a massive, a
• 13:43
  massive hole of creative testing.
• 13:45
  And so that's kind of why we lean into
• 13:47
  the spend is signal philosophy.
• 13:49
  And what's great about that is you can
• 13:50
  actually see this
• 13:51
  playing out in real time.
• 13:52
  So this is a screenshot from Motion.
• 13:53
  Motion has this awesome
• 13:55
  on their creative reports.
• 13:57
  No, no.
• 13:57
  Yeah, this is not sponsored by Motion.
• 13:59
  But in Motion's creative reports, they
• 14:02
  have a little like graph, a line graph
• 14:04
  view where you can see how
• 14:05
  spend goes and trends over time.
• 14:08
  And you can do this in a
• 14:09
  bunch of different ways.
• 14:09
  You can do this on an ad level.
• 14:10
  You can do it on a landing page level.
• 14:12
  You can do this on the AI tag
• 14:14
  level now, which is amazing.
• 14:15
  But I love I love this graph.
• 14:17
  I think it's one of the most valuable
• 14:17
  graphs that people just don't think about
• 14:19
  because spend is signal.
• 14:21
  And you can actually see in real time how
• 14:23
  meta is making decisions
• 14:24
  between all of your different ads.
• 14:25
  And so I think Shireen, Shireen Albert
• 14:28
  coined the term tornado.
• 14:30
  Like she basically said, like, hopefully
• 14:33
  I'm phrasing this correctly, but you
• 14:35
  don't want to make funnels.
• 14:36
  You want to make tornadoes because what
• 14:37
  you don't know is at any particular time,
• 14:40
  what ad somebody is going to respond to
• 14:42
  that's going to get them to convert or is
• 14:44
  going to get them interested and at least
• 14:47
  down the right path.
• 14:48
  So what you want to do, and this is I
• 14:49
  think is the real point
• 14:50
  of creative diversity,
• 14:51
  is try to have ads that speak to all of
• 14:55
  your potential customers at all the
• 14:58
  different stages of
• 14:58
  buying that they might have
• 14:59
  and answer all the questions they might
• 15:01
  possibly need answered in
• 15:02
  order to make a purchase.
• 15:03
  And that's both from a creative
• 15:04
  perspective, but also from
• 15:05
  a landing page perspective,
• 15:06
  which is probably time for a different
• 15:08
  podcast, because I think landing page
• 15:10
  diversity is widely slept on.
• 15:11
  And I think that's going to be one of the
• 15:13
  one of the key themes for the back half
• 15:14
  of this year that people
• 15:15
  will finally catch on to,
• 15:16
  is that you might have you might have an
• 15:18
  ad that is very top of funnel that runs
• 15:19
  to, you know, maybe a PDP.
• 15:21
  But maybe you have one that runs to a
• 15:22
  quiz page, and that's actually driving
• 15:24
  more net new traffic to the website.
• 15:26
  It lowers your cost per 1000 reach, which
• 15:28
  I know is a big topic of discussion.
• 15:30
  But then you have a bunch of images that
• 15:31
  talk about HSA or promos or
• 15:35
  all of these different pieces,
• 15:37
  which we would consider, right, more
• 15:38
  bottom of funnel messaging.
• 15:40
  But who is to say that, you know, you
• 15:42
  know exactly what the customer needs to
• 15:45
  see at this moment in
• 15:46
  time in order to convert.
• 15:47
  And so the point is, and the way that we
• 15:49
  think about setting up ad accounts more
• 15:50
  and more is give meta all of the
• 15:52
  messaging and all the resources.
• 15:54
  Right. And I think this is why you see
• 15:56
  with Andromeda and their AI usage.
• 15:58
  They are trying to push towards letting
• 16:00
  you let them run more AI features because
• 16:04
  they are trying to say,
• 16:05
  we actually know better than you to a
• 16:07
  degree that you don't
• 16:07
  even understand yet.
• 16:08
  And that's why we want to customize this.
• 16:10
  Now, is it there to be determined?
• 16:12
  And it's hard to say, but more and more
• 16:14
  we're thinking about how do we just like
• 16:16
  set up at the ad
• 16:16
  accounts to give it all the ads,
• 16:18
  give it a give it a governor with cost
• 16:19
  controls and our goals, which is the way
• 16:21
  that we prefer to do things.
• 16:22
  And just give meta the ecosystem of ads
• 16:25
  that somebody might need to see at
• 16:27
  varying points of the purchase journey
• 16:28
  and let them figure it out.
• 16:30
  And you can see that very clearly.
• 16:31
  Right. We've had ads that pop off really
• 16:33
  quickly and then fall off.
• 16:34
  Right. And this did really well for a
• 16:36
  short moment of time, but then it kind of
• 16:37
  fell by the wayside.
• 16:38
  But it still makes up a
• 16:39
  decent chunk of spend.
• 16:40
  And so that's a little bit.
• 16:42
  I'm kind of going
• 16:43
  overkill on my on the this graph.
• 16:45
  But I just think spend velocity is is a
• 16:48
  is a really important
• 16:49
  signal to lean into.
• 16:50
  Usually when winners hit, they hit fast.
• 16:53
  Now, there can be times where
• 16:54
  seasonality, you know, you've had an ad
• 16:56
  live for months and all of a sudden
• 16:58
  something happens
• 16:58
  seasonally and it shifts.
• 16:59
  And all of a sudden that
• 17:00
  becomes a winning ad again.
• 17:02
  Spend velocity in the short term is
• 17:03
  something that we pay a lot of attention
• 17:04
  to because spend is signal.
• 17:06
  OK, friends, quick break.
• 17:07
  This episode of Scalability School is
• 17:09
  brought to you by North Beam
• 17:11
  Incrementality, which
• 17:12
  honestly is so awesome.
• 17:14
  Let me just tell you
• 17:15
  a little bit about it.
• 17:16
  If you're a marketer, you probably agree.
• 17:18
  Incrementality testing is broken.
• 17:20
  It's insane.
• 17:21
  You got to do all this crazy stuff.
• 17:22
  But with North Beam, you
• 17:23
  do it right in the platform.
• 17:25
  It builds tests using your own MTA data,
• 17:27
  launches them directly in the ad platform
• 17:30
  and automatically monitors them
• 17:32
  throughout the test.
• 17:33
  And if something threatens validity,
• 17:35
  you're alerted to it within North Beam.
• 17:38
  Before it ruins the results.
• 17:40
  The really nice thing is it solves all
• 17:42
  these problems you hear about
• 17:43
  incrementality testing all the time, like
• 17:45
  labor, spreadsheets,
• 17:47
  statistical reference.
• 17:48
  Designing tests is
• 17:49
  normally so slow, right?
• 17:50
  And error prone.
• 17:51
  So with North Beam Incrementality, it's
• 17:53
  automated and accurate and your team is
• 17:56
  able to spend with confidence.
• 17:58
  Give it a shot with
• 17:59
  North Beam Incrementality.
• 18:01
  Really, the future of incrementality is
• 18:03
  here and North Beam has
• 18:04
  figured out a way to make it easy.
• 18:06
  OK.
• 18:07
  Well, what we don't like to do is turn
• 18:09
  off a ton of ads very frequently, right?
• 18:12
  Just because an ad is spending today
• 18:13
  doesn't mean you need to kill it today.
• 18:15
  And just because it doesn't spend it for
• 18:17
  the next three days or the next seven
• 18:19
  days doesn't mean you need to kill it
• 18:20
  over the course of that time.
• 18:21
  Now, is there a time where it's worth
• 18:23
  going through and cleaning up ads that
• 18:25
  are getting no spend?
• 18:25
  Sure.
• 18:27
  Sure.
• 18:27
  I think that that's a fair, fair thing to
• 18:28
  say, especially if you're running on, you
• 18:30
  know, running up on page limits or
• 18:32
  something like that.
• 18:33
  If an ad has been live for two to three
• 18:35
  weeks, it's spending a
• 18:36
  dollar or less than sure.
• 18:38
  You know, you can probably feel mostly
• 18:40
  confident in turning that off.
• 18:41
  But there's just
• 18:41
  because it's not spending.
• 18:43
  There's not a lot of
• 18:43
  downside in leaving it on either.
• 18:45
  I think a lot of people get worried that
• 18:47
  they have too many active ads and that
• 18:49
  meta can't distinguish between them.
• 18:51
  And I would just say that we've not ever
• 18:53
  really seen that to be the issue.
• 18:55
  There are some caveats to this, right?
• 18:56
  I've got some concerns of
• 18:57
  like when we turn off ads.
• 18:58
  If you've got an ad that all of a sudden
• 19:00
  it's super clickbaity and your
• 19:02
  click-through rate is three times the
• 19:04
  average of your ad account,
• 19:05
  meta is going to overspend on it because
• 19:06
  it has built up an understanding of your
• 19:09
  historical action rates,
• 19:10
  which is basically if you have a 1%
• 19:12
  click-through rate consistently and your
• 19:13
  conversion rate is 3%,
• 19:15
  all of a sudden you have a 3%
• 19:16
  click-through rate, meta still thinks
• 19:18
  your conversion rate is going to be 3%,
• 19:19
  at least in the short term on that ad.
• 19:21
  It's using historicals to kind of make a
• 19:23
  bet a little bit about what
• 19:24
  will happen in the future.
• 19:25
  So over the course of two to three days,
• 19:27
  your spammy ad is going to rip on
• 19:28
  delivery because meta
• 19:30
  still hasn't caught up.
• 19:31
  They will catch up.
• 19:32
  But if you watch that ad and you
• 19:35
  understand that it's very clickbaity,
• 19:37
  then totally fine to go
• 19:38
  ahead and turn that off.
• 19:40
  Okay.
• 19:40
  So let's keep going.
• 19:42
  So the real system, creative cohorts, and
• 19:44
  this is where you can actually find some
• 19:46
  rhythm and some cadence to it.
• 19:47
  And so we've kind of gone through, okay,
• 19:49
  what does the actual ad
• 19:50
  account structure look like?
• 19:51
  It's single campaign ad sets.
• 19:54
  When you max the ad sets out, you add in
• 19:56
  more ads and you just keep stacking ad
• 19:57
  sets until you find your winners, right?
• 20:00
  But it's important to
• 20:01
  still look at your batches.
• 20:04
  So you can launch them on a weekly basis.
• 20:06
  You can launch them
• 20:07
  more frequently than that.
• 20:08
  I would just make it a goal to launch
• 20:09
  something every single week.
• 20:11
  And if all you can do for
• 20:12
  now is two ads a week, fine.
• 20:13
  Just do two ads a week.
• 20:14
  But it's important to figure out the
• 20:16
  system to get scale in the first place.
• 20:18
  And then you can refine
• 20:19
  and get better over time.
• 20:20
  But creative strategy can really, it
• 20:22
  thrives when it has more to work off of.
• 20:24
  So if you've got a bunch of cohorts of
• 20:26
  weekly launches, right, like you can see
• 20:28
  on the screenshot, and this is another,
• 20:30
  this is the new creative report, I
• 20:32
  believe, from Motion.
• 20:33
  It's fantastic.
• 20:34
  You can see how much creative was
• 20:35
  launched over the course of a week.
• 20:36
  You can click the little dropdown.
• 20:37
  And you can actually start to see how
• 20:39
  much did this batch of
• 20:40
  creative get in overtime.
• 20:42
  And what you'll see most times is that
• 20:45
  new batches aren't
• 20:46
  getting a ton of spend.
• 20:47
  Every once in a while,
• 20:48
  you're going to hit a banger.
• 20:48
  And like that is just, that is the nature
• 20:50
  of outlier ads is they're outliers.
• 20:52
  And it takes a lot to get those.
• 20:54
  But you can still stack up
• 20:55
  small wins on a weekly basis.
• 20:57
  I've seen ad accounts that they have ads
• 21:01
  that make up 90% of their spend.
• 21:02
  And I've seen ad accounts where they have
• 21:03
  ads that make up 10% of their spend.
• 21:05
  And nothing is actually over that.
• 21:07
  But what you're doing is you
• 21:08
  are stacking up the baseline.
• 21:10
  If you can just squeeze out $100 a day
• 21:12
  per spend across a couple of new ads and
• 21:14
  a couple of new ads, you just
• 21:15
  continue to stack the volume.
• 21:17
  And that's the goal.
• 21:19
  Like I said, the goal is accumulation
• 21:21
  over time and not fixating on just one
• 21:23
  week of performance.
• 21:24
  Because sometimes it just takes time for
• 21:26
  things to actually pop off and spend.
• 21:28
  And you just want to consistently build a
• 21:30
  portfolio of winners,
• 21:31
  which is what you can see here.
• 21:32
  I'm recording this just a couple weeks
• 21:33
  after I took this screenshot now.
• 21:35
  I was recording this a bit after.
• 21:37
  But what you can see is we had a lot of
• 21:39
  winners over this creative cohort.
• 21:42
  This is ad set number four, like the
• 21:43
  screenshot you were seeing earlier.
• 21:45
  We have a ton of winners in this ad set.
• 21:47
  But you're not always going to hit a
• 21:48
  bunch of winners like this.
• 21:49
  The goal is across a bunch of weeks, you
• 21:51
  are hitting winners
• 21:52
  every once in a while.
• 21:53
  That just adds some amount of volume and
• 21:55
  scale to your ad account.
• 21:56
  And if you keep making those bets,
• 21:58
  eventually you will crack the one that
• 21:59
  can spend 90% of the ad accounts.
• 22:01
  You still have a bunch below.
• 22:02
  Now, what I wanted to do was provide some
• 22:05
  screenshots and some
• 22:06
  context to different ad accounts.
• 22:08
  So you can see this in real time.
• 22:09
  And when we start to deviate from the
• 22:11
  single campaign
• 22:12
  structure, you can see why.
• 22:13
  And I can provide some context as to why.
• 22:15
  So in this case, pretty straightforward.
• 22:17
  Single campaign spent
• 22:19
  $300,000 over the course of 30 days.
• 22:22
  Here's a different one.
• 22:23
  What you'll notice is there's a couple of
• 22:24
  differences in here.
• 22:25
  The top campaign is just run on bid cap.
• 22:28
  The one below that is a min ROAS bidding.
• 22:32
  So the same content in both of those,
• 22:33
  just bid cap versus min ROAS.
• 22:35
  Then we've got a New Year's, which was a
• 22:37
  promo specific campaign.
• 22:38
  And then we've got an additional one
• 22:40
  below that, which was
• 22:41
  a different product.
• 22:42
  So the bid cap on top and the bid cap on
• 22:43
  bottom were just two different products.
• 22:45
  Now, the sourcing, interestingly, is
• 22:47
  actually combining the products.
• 22:48
  And because they have, they're different
• 22:51
  price points, but they operate on a
• 22:53
  similar target ROAS,
• 22:55
  we could combine those into one min ROAS
• 22:57
  campaign and we feel
• 22:58
  more confident about that.
• 22:59
  But the top and the bottom campaign, two
• 23:01
  different products, sourcing combined,
• 23:03
  but is running on min ROAS.
• 23:04
  We feel comfortable combining them.
• 23:05
  And then the New Year's campaign, which
• 23:07
  was a promo campaign,
• 23:08
  which we also split out
• 23:09
  as a separate campaign.
• 23:10
  In this case, we've got
• 23:11
  three different campaigns.
• 23:12
  This is for three different products.
• 23:14
  For this one, two different campaigns,
• 23:16
  two different products.
• 23:17
  This one, three different campaigns,
• 23:19
  three different products.
• 23:19
  Two campaigns, two
• 23:20
  different products, right?
• 23:21
  Pretty straightforward on
• 23:22
  how it's actually set up.
• 23:23
  Now, there are some tools that are
• 23:25
  helpful in the process of doing this.
• 23:27
  Here's kind of how we
• 23:28
  think about creative.
• 23:28
  This is just like some
• 23:29
  nice stuff at the end.
• 23:30
  Again, no proper
• 23:32
  affiliation with any of these tools,
• 23:33
  just things that we really like that help
• 23:35
  with the creative testing stack.
• 23:36
  Obviously, you've got Meta
• 23:37
  Ads Manager, appreciate it for.
• 23:38
  But then we use Ad Managed AI.
• 23:40
  There's a ton of great ad
• 23:41
  launching tools out there.
• 23:42
  There's new ones launching every single
• 23:43
  day that everybody's vibe coding.
• 23:44
  But we use a tool like Ad Managed AI to
• 23:48
  launch all of our ads
• 23:49
  because we're constantly launching a ton.
• 23:51
  We use Motion for performance review,
• 23:53
  cohort tracking, like I showed you.
• 23:55
  It's baked in.
• 23:56
  And then they have AI tags, right?
• 23:57
  So when I was showing the graph over time
• 23:59
  of how things spend and trend over time,
• 24:02
  that's baked into Motion.
• 24:03
  And it's, again, one
• 24:04
  of my favorite tools.
• 24:05
  So you can see not only how individual
• 24:07
  ads are trending over time,
• 24:09
  but different types of messaging, right?
• 24:10
  We had a client last year, I did a
• 24:13
  six-month performance analysis,
• 24:16
  halfway through the year.
• 24:17
  And what we looked at was, how did
• 24:18
  messaging trend by month?
• 24:20
  And they sell a product
• 24:21
  that can be used year-round,
• 24:22
  and they have some
• 24:23
  seasonality, but for the most part,
• 24:24
  like they do really well year-round.
• 24:26
  But every two months, the
• 24:28
  top-performing messaging changed.
• 24:30
  And so it's really important that we
• 24:31
  looked at the AI tags that Motion has.
• 24:33
  And you can see when you
• 24:34
  talk about how this works for,
• 24:37
  I'm going to have to
• 24:37
  anonymize this a bit,
• 24:38
  but when you talk about how it works for
• 24:39
  winter versus how it works in spring
• 24:41
  versus how it works in
• 24:42
  summer versus how it works,
• 24:43
  you know, the messaging
• 24:45
  changed substantially.
• 24:46
  Now, it was more specific than that,
• 24:48
  but there was always a new theme of
• 24:50
  something that worked
• 24:51
  better every two months.
• 24:53
  And if you don't track
• 24:54
  that, you don't know that
• 24:55
  because seasonality can impact messaging
• 24:57
  very easily, as you can imagine.
• 25:00
  And therefore, you might have winners
• 25:02
  that look like losers all of a sudden
• 25:03
  because you're forcing adspend into them,
• 25:05
  but it just might not be
• 25:06
  right for the time of year.
• 25:07
  But wait two months and it could pop off.
• 25:09
  So that's why I love
• 25:10
  those cohort views for Motion.
• 25:14
  Okay, I had a bunch of different
• 25:16
  questions that came up as I posted on X
• 25:18
  and some of the replies that we got
• 25:20
  following the YouTube video
• 25:23
  or the previous podcast
• 25:24
  that we did with Phil.
• 25:25
  And so I kind of chunked up the most
• 25:26
  popular questions that I got
• 25:28
  and just wanted to address
• 25:29
  those here as quickly as I can.
• 25:31
  So first was, how are you adjusting
• 25:33
  budgets when you launch these?
• 25:34
  Well, the good news is that the campaign
• 25:36
  is set at what I'm willing
• 25:37
  to spend on a daily basis.
• 25:39
  And because we run cost controls, it's
• 25:42
  usually inflated by 30 to 50%,
• 25:44
  which means if we are hitting our budget,
• 25:46
  it's because we're hitting our budget,
• 25:48
  we're on target, which means
• 25:49
  we should open up our budget.
• 25:50
  The degree to which you open up your
• 25:51
  budget is relative to
• 25:52
  how efficient you are.
• 25:54
  You know, if you're super efficient,
• 25:55
  you're 2x your efficiency goal,
• 25:56
  then you can be more aggressive with
• 25:57
  100%, 200% increases.
• 25:59
  If you're right on your target, then
• 26:01
  maybe you want to be a
• 26:02
  little bit more cautious
• 26:03
  and keep it no more
• 26:04
  than 50% of an increase.
• 26:06
  And if you're not running on cost
• 26:07
  controls, you know, how do
• 26:08
  you think about scaling it?
• 26:09
  I think you can feel, again, it depends
• 26:11
  on the degree to which
• 26:12
  you are above your targets.
• 26:14
  But if you're above your
• 26:14
  targets, 20 to 30% is safe.
• 26:17
  But I think you can feel confident
• 26:18
  increasing by 50 to 100%
• 26:20
  if you are smashing and you're on target.
• 26:23
  I personally don't love running lowest
• 26:25
  cost or highest volume,
• 26:27
  whatever they're calling it these days.
• 26:28
  But that's how I would think about budget
• 26:31
  adjustments in that case.
• 26:32
  Are you running ads to your exclusions
• 26:34
  list in a different campaign?
• 26:35
  Yes.
• 26:35
  So I said I wasn't going
• 26:36
  to talk about this earlier,
• 26:37
  but there was a question about it.
• 26:38
  I got this a few times.
• 26:39
  We run specific retention campaigns.
• 26:41
  Simply, our retention campaign setup is
• 26:43
  we target lifetime customers,
• 26:45
  but we exclude people who
• 26:46
  are kind of active, right?
• 26:47
  Like they have a chance of coming back.
• 26:49
  There's like 80% or less
• 26:52
  chance that they do come back.
• 26:55
  And so we're targeting
• 26:56
  basically lapsed people,
• 26:57
  people who without seeing an ad would be
• 26:59
  unlikely to come back.
• 27:01
  Now, if you have solid volume, you know,
• 27:03
  if you have hundreds of thousands or
• 27:06
  millions of customers,
• 27:07
  you can run holdout
• 27:08
  tests on that fairly easily
• 27:09
  to see if meta is actually incremental
• 27:12
  when you target those people.
• 27:13
  We ran this recently and we found that
• 27:14
  meta was incremental
• 27:16
  when doing this for one of our clients
• 27:17
  who's got hundreds of
• 27:18
  thousands of customers.
• 27:19
  They run at like a 5X in platform.
• 27:21
  And it was, it came out to be like a 4X
• 27:23
  in the meta conversion
• 27:26
  of study that we ran.
• 27:27
  So it was definitely incremental.
• 27:29
  A lot of concerns that people have is
• 27:31
  that new ads don't get spent.
• 27:32
  So I have a little
• 27:33
  bit of a side anecdote.
• 27:34
  I actually saw a tweet from, I believe,
• 27:37
  I might butcher his last name.
• 27:38
  So apologies, Nick, if you hear this,
• 27:39
  but Nick Therio, he's on Twitter.
• 27:43
  I think he also makes
• 27:44
  some great YouTube videos
• 27:45
  and reminded me of an
• 27:47
  experience that I had
• 27:48
  with one of our clients
• 27:48
  where we've had multiple clients
• 27:50
  who are spending 10 to 20K a day
• 27:52
  and they are spending time and attention
• 27:54
  on making new creative assets.
• 27:55
  And we're also helping them
• 27:57
  with creative assets, right?
• 27:58
  And we were having issues getting spend
• 28:01
  into some new creative assets, right?
• 28:03
  We would have winners and it was long.
• 28:05
  It's allowed the account to scale,
• 28:07
  but we were making so much new creative
• 28:08
  and there's obviously tons of batches
• 28:10
  that just aren't getting spent.
• 28:11
  And so the hesitancy becomes,
• 28:13
  well, we're spending all this money
• 28:14
  and all this time on this creative.
• 28:15
  Like we should try to
• 28:16
  force some spend into it.
• 28:18
  And so if the ad
• 28:19
  didn't get spent in the CBO,
• 28:20
  we would put it into an
• 28:21
  ABO and force spend into it.
• 28:23
  So it was a loser in the CBO
• 28:24
  and we put it into the
• 28:26
  ABO to give it a chance.
• 28:27
  Every single one of them fail.
• 28:28
  100% of them.
• 28:29
  And this has happened
• 28:30
  every time I've done this.
• 28:31
  And I would love,
• 28:32
  if anybody has had an
• 28:33
  opposite experience,
• 28:34
  I would love to see that, honestly,
• 28:35
  because I think it'd be
• 28:36
  very interesting to see.
• 28:37
  But my experience has
• 28:38
  been when we've met,
• 28:39
  meta is right way more often
• 28:41
  than I think people give them credit for
• 28:43
  when for the people
• 28:44
  that have the ABO opinion.
• 28:46
  And again, there's plenty of people
• 28:47
  who have scaled the ABO.
• 28:48
  I'm not saying it doesn't work.
• 28:49
  I'm just saying this is my preference
• 28:51
  of what I've seen to work well.
• 28:52
  A lot of people ask,
• 28:53
  do you just keep adding creative
• 28:54
  to the latest ad set until it's full?
• 28:56
  Yes, very simply.
• 28:57
  That answers that.
• 28:59
  The one clarification that kept coming up
• 29:01
  is your testing campaign
• 29:03
  than your scaling campaign.
• 29:04
  Yes, we do not split
• 29:05
  testing versus scaling
• 29:06
  just for quick context on that.
• 29:08
  And then we have this,
• 29:09
  I've had this thing
• 29:09
  come up several times,
• 29:11
  which is about having
• 29:13
  your top spending ads,
• 29:14
  having a lower ROAS and other ads.
• 29:15
  And so what I wanted to do
• 29:16
  was pull an example
• 29:18
  from an account recently
• 29:19
  and just kind of talk through this.
• 29:20
  So these are on an ad set level,
• 29:22
  but it still tells the same story.
• 29:24
  And what you can see,
• 29:25
  there's a couple of things going on here.
• 29:26
  So I've got one campaign
• 29:28
  and I've got two ad sets.
• 29:29
  And if you look at the top campaign,
• 29:31
  there's one thing that stands out.
• 29:32
  It is spent, what is that,
• 29:33
  like four times more budget
• 29:35
  than the campaign below it.
• 29:37
  And it has a ROAS of a whole point lower.
• 29:39
  So top campaign, if you're listening,
• 29:41
  spends $197,000.
• 29:43
  Bottom campaign spends $55,000.
• 29:45
  Top campaign performing at a 2.63.
• 29:48
  Bottom campaign performing at a 3.4.
• 29:50
  So oftentimes people see that
• 29:53
  and they say,
• 29:53
  Brad, what are you doing?
• 29:55
  Like, why are you not scaling more
• 29:57
  into that specific ad set?
• 29:58
  Like you're leaving money on the table.
• 29:59
  It's performing,
• 30:00
  it's performing 50% more efficient
• 30:02
  than the other one.
• 30:03
  Oh, and my answer is,
• 30:04
  well, Meta knows something that we don't
• 30:06
  about why that can't spend more.
• 30:08
  And what we've found
• 30:09
  is that there are usually some signals
• 30:11
  that give us context into what that is.
• 30:13
  Now, we can't say for sure
• 30:14
  because Meta is a
• 30:15
  little bit of a black box,
• 30:16
  but we can look at some other metrics
• 30:17
  to tell us, well,
• 30:18
  why is Meta making this decision
• 30:19
  that they're making?
• 30:20
  And so things that I
• 30:20
  always like to look at,
• 30:22
  you can find this in North Beam,
• 30:23
  Triple Whale, Google Analytics.
• 30:25
  Any tool will give you this insight.
• 30:28
  Any one of those analytics tools
• 30:29
  will give you this insight.
• 30:30
  It's new visitors as a
• 30:31
  percentage of traffic
• 30:32
  from that campaign.
• 30:33
  And so on the campaign number one,
• 30:34
  which is the one that
• 30:35
  has worse efficiency,
• 30:36
  but way more spend,
• 30:37
  it has a 61% new visitor percentage.
• 30:40
  So out of the people that click the ads
• 30:42
  over this time period,
• 30:43
  61% of them have never
• 30:44
  been to the website before.
• 30:45
  40% of them have been
• 30:46
  to the website before.
• 30:47
  Now that campaign as a whole
• 30:48
  is delivering a much
• 30:49
  better new visitor percentage,
• 30:50
  but this is just one example.
• 30:52
  The ad set below that,
• 30:53
  the one that is way more efficient
• 30:55
  is sending half as many new people
• 30:57
  to the website at only 33%.
• 30:58
  So that's, you can tell,
• 31:01
  kind of what I'm alluding to here
• 31:02
  is the top campaign is fueling
• 31:04
  the rest of the funnel.
• 31:05
  And that starts to become evident
• 31:06
  as you look at new visitor percentage.
• 31:08
  If you don't have
• 31:09
  access to one of those tools
• 31:10
  or you're not
• 31:10
  comfortable in Google Analytics
• 31:11
  since they torched it,
• 31:13
  you can look inside of Meta
• 31:15
  and there are some signals
• 31:15
  that can lead you down this direction.
• 31:17
  And so the example that I'm showing
• 31:18
  in this screenshot is,
• 31:20
  again, top spending campaign
• 31:21
  performing less or worse
• 31:24
  on an efficiency level,
• 31:25
  but the frequency is a 2.59,
• 31:27
  whereas the lower spending one,
• 31:29
  it has a higher,
• 31:29
  even a higher frequency at a 3.64.
• 31:31
  So it's a higher frequency
• 31:32
  on substantially less spend.
• 31:34
  And what you can see also next to that
• 31:35
  is the cost per 1,000 accounts reach.
• 31:38
  So that's not CPM,
• 31:39
  that's reach,
• 31:39
  it's CPMR,
• 31:41
  which is what people refer to.
• 31:42
  It's half.
• 31:43
  It is half of what the top campaign is.
• 31:45
  And so the top campaign
• 31:46
  is just fueling the rest of the funnel.
• 31:49
  And that's why we generally try to,
• 31:51
  again,
• 31:51
  it's one of the other reasons
• 31:52
  why we try to not tell Meta
• 31:54
  how much they should spend
• 31:55
  into specific things across the board,
• 31:58
  right?
• 31:58
  Specific promo campaigns.
• 31:59
  This happens on Black Friday every year.
• 32:01
  You know,
• 32:01
  people have this debate on,
• 32:03
  are my Black Friday assets
• 32:04
  going to spend more
• 32:05
  than my Evergreen assets?
• 32:06
  Well,
• 32:06
  if you run them in campaigns
• 32:08
  that are on cost controls,
• 32:09
  you don't have to make that decision.
• 32:11
  You just leave room for them
• 32:12
  to spend into it
• 32:13
  and Meta will make that allocation.
• 32:14
  Now, again,
• 32:15
  there are super talented media buyers
• 32:16
  who maybe watch CPMR like a hawk.
• 32:20
  Like a hawk,
• 32:21
  they watch new visitor percentage
• 32:22
  like a hawk.
• 32:22
  They watch frequency by a hawk,
• 32:24
  like a hawk rather.
• 32:26
  And they can analyze these decisions.
• 32:29
  I just personally think
• 32:30
  it's not a great use of time
• 32:32
  when Meta can do it very well.
• 32:34
  So I think that helps
• 32:35
  kind of answer some of that question.
• 32:36
  So here's like a simple version recap
• 32:38
  of everything.
• 32:39
  So if you want to pause and watch this,
• 32:40
  long story short,
• 32:41
  one CBO campaign per product or offer.
• 32:44
  Ad sets are functionally just folders.
• 32:45
  Make new ones when you get the limit.
• 32:47
  Build the system
• 32:48
  for making consistent weekly content
• 32:50
  and just focus on getting volume first
• 32:54
  and then you can
• 32:55
  optimize for quality later
• 32:56
  because I think
• 32:56
  quality is a bit subjective.
• 32:58
  Let Meta allocate the spend.
• 33:00
  Don't force it.
• 33:00
  And then track your cohorts over time.
• 33:02
  If you go weeks and weeks
• 33:03
  without hitting any winners,
• 33:05
  then it's time to think about quality
• 33:06
  a little bit more.
• 33:06
  Like what are you missing
• 33:07
  in your creative strategy
• 33:08
  that is leading you
• 33:09
  to launching a bunch of losers?
• 33:11
  And let spend dictate that.
• 33:13
  But the good news is
• 33:14
  as you review those losers,
• 33:15
  you can feel a lot better
• 33:16
  knowing you haven't just blown
• 33:17
  thousands of dollars
• 33:18
  in testing creative.
• 33:19
  That wasn't going to work.
• 33:20
  So if you have any questions
• 33:21
  about any of this,
• 33:22
  feel free to leave it in the comments.
• 33:23
  You can hit me up on Twitter.
• 33:25
  You can probably email Andrew.
• 33:26
  His email will probably be
• 33:26
  somewhere in the description.
• 33:28
  Happy to help,
• 33:29
  but wanted to show
• 33:29
  some more tactical follow-ups
• 33:31
  to how we actually do creative testing.
• 33:33
  Thanks so much for tuning in this week
• 33:34
  and we'll see you next time.
• 33:36
  And yeah, sorry I missed beefs.
• 33:38
  I don't have any beefs this week.
• 33:39
  Just trying to make the world
• 33:40
  a better place one CBO at a time.
• 33:42
  This episode is brought to you
• 33:48
  by Brad's company, Work Marketing.
• 33:52
  If you need a D2C marketing agency,
• 33:54
  let me tell you,
• 33:55
  Homestead is great,
• 33:56
  but Work Marketing is also fantastic.
• 33:59
  And let me tell you,
• 34:00
  you aren't going to find
• 34:02
  friendlier people out there
• 34:03
  in the e-commerce space.
• 34:06
  So we decided to do these little ads
• 34:08
  for each other's companies.
• 34:09
  So hopefully you find it interesting.
• 34:11
  But seriously, great
• 34:11
  team at Work Marketing.
• 34:13
  Very smart.
• 34:14
  Brad and Jordan are incredibly dialed in.
• 34:16
  I just gave them a lead.
• 34:18
  Already made this brand.
• 34:19
  I gave them like,
• 34:20
  I don't even know,
• 34:21
  double the revenue that they had
• 34:23
  the previous month or something.
• 34:25
  So, you know,
• 34:26
  it's very exciting to
• 34:27
  be connected with Brad.
• 34:28
  And if you need a great agency,
• 34:30
  there's really no one better.
• 34:32
  Zach, anything to
• 34:33
  comment on Work Marketing?
• 34:34
  Yeah, I mean,
• 34:35
  if you want an agency
• 34:36
  that cares about your business
• 34:37
  much more than they care
• 34:38
  about their own website,
• 34:39
  I just tried to load workmarketing.com
• 34:41
  and it was broken.
• 34:42
  So they're definitely going to give
• 34:43
  more of a shit about your business
• 34:45
  than their own.
• 34:46
  So I highly recommend Brad
• 34:47
  and the team over at work.
• 34:48
  They've been incredible.
• 34:49
  We've referred a lot of business
• 34:51
  over to them as well.
• 34:52
  Really, really good
• 34:52
  as far as like cracking funnels
• 34:54
  and figuring out like
• 34:55
  rapid growth for brands.
• 34:56
  So I highly recommend these guys.
• 35:03
  Yeah, please go check us out on YouTube.
• 35:06
  Rack up those views for us.
• 35:07
  We'd love to see it.
• 35:08
  And then subscribe.
• 35:09
  Make sure to subscribe
• 35:09
  on YouTube as well.
• 35:10
  And I relentlessly refresh
• 35:11
  the YouTube comments
• 35:12
  because it dictates
• 35:13
  my mental health for the day.
• 35:14
  So please say something nice
• 35:15
  about all of us.
• 35:17
  Thank you, everyone.
• 35:18
  Thanks for listening.
• 35:19
  Honestly.