ABO Approach for Successful Meta Ads Campaigns

[00:00:00] We saved, um, money here, instead of, you know, spending essentially 10% less. So, we saved seven thousand. Our cost per, you know, lead, as you call it, went down, but our sales went up from 168 to 200. Our customer base went up from 169 to 200, and our ENCAC went down 11.4%. So

[00:00:17] 18% more new customers for 10% less spending, and also decreased by 11.5%. I want to share a new metastrategy with you all. Let me bring all this up. Good. So, I'm going to bring up Wicked 2. Okay. And good. This is loading. Let me raise the vision of the sales funnel. Click on the date.

[00:01:01] Good, good, good. And then it loads. Please wait. We started a new way of scaling. Hmm, that's actually pretty cool. So, I think I told you how we launched instead of our L1 for creative testing. Last week we decided to withdraw 22k from our L1 and invest it in our. And the way

[00:01:28] we added that 118% is because we found that you can scale very effectively by cloning campaigns or, sorry, cloning ad sets on. So, we found that, you know, if you look at, you know, Michelle, oh, sorry, Barbara, Barbara, Barbara, Barbara, Barbara, these two-new Barbaras. We

[00:01:52] created Barbara in the app. She was getting 125, and she was getting a good first click in Wicked. So we clone it again. Barbara gets 155. We're like: great. So we clone it again. Give him 141. We're like: great. Then these last two clonings happened, um, forget it, it was around the 12th. They are

[00:02:11] actually starting to show points of slight return, but still look decent from a new customer perspective. It's interesting that we said, "Okay, let's clone Barbara," and every time they achieve their goal. Since it's on, it still costs $225 per day. So every time we clone Barbara for

[00:02:28] $225 a day, we are effectively adding $6,700 a month to the campaign. So we just clone it a few times. So we said, "Okay, let's try again." We did Joan. We had Joan, Joan, and more Joan. And in the appendix it was something like: “ Okay, 321—that’s terrible ; 384—this is terrible; and 390 received nothing— it's terrible." Okay, we made one example, it worked fine, and the same thing happened when we cloned it. So

[00:02:50] it's pretty much like this: if an ad worked well, we clone it, it still works well; or if it doesn't work well and we clone it, it still doesn't work well. So then we thought, “Okay, how do we do this on a mass scale? How can we actually scale this very quickly?” So, we chose this new method of creative testing and scaling, which is to take one new ad, let's call it "Pam," clone them five times at $100 per day each. So now I get

[00:03:16] $500 a day to test the creative "Pam." And then we see the results in about a week. And what's nice about that is we look at it and say, "Okay, Pam actually made $500." "There's nothing here." Okay, this is a failure. Now let's try Suzanne. Suzanne cloned five times. Two out of five are still

[00:03:37] operating at a cost per conversion of less than $100. Perfectly. Continue in the same spirit. Let's wait a little longer. And then we'll move on to Lori. Lori has metrics from here to here. So you have these five, two out of five. Good. She shows results just as well as Suzanne, but also outperforms Pam. And then we

[00:03:56] look at Claudia's next commercial. And Claudia also has five ads, nothing. Okay, that's terrible too. The last one, Diana. So we just remove these. Diana actually has three out of five ads working, one of which has already generated two purchases at a cost of $50 per new customer. Okay,

[00:04:19] great. So it's funny that once we have one ad, instead of adding it and hoping it succeeds, or trying to scale it, we've created five different versions of the same ad, which means they all have to perform roughly the same. Or we compare the average of each five with the average of the other five. And if the

[00:04:40] other five don't work, these, for example, Claudias, are suspended. So these five are suspended. Then $500 a day goes to cloning Diana five more times. Diana will go from five to ten. So we take a loser out of five and a winner out of five. We take the advertising costs of the loser and

[00:04:59] give them to the winner, cloning him. And then we add more ads. And so we incur additional costs. The most interesting thing is that since we do this, we extract 50% of our costs inside our L1. And this 50% cost obviously gives us economies of scale by reducing ENCAC. Our L0, where

[00:05:22] we calculate roughly like this, scaling is 34%, and ENCAC is falling. But what's interesting is that our new option that we just launched has a $2,400 price tag. The new version where they have to clone five times is growing, and this one starts with a final CAC of 273. It's better than

[00:05:40] remarketing, and this version, if it doesn't continue to grow, will actually be the one that we just keep cloning more and more and more and more and more. So it's actually really interesting to watch how this is, when we just take these five and clone them, we're going to have a whopping $500 a day testing five ads that are also much cooler than our L1, and we're just going to disable this one.

[00:06:07] Basically, this one will just be erased and we'll only be using this new one here for the foreseeable future. This is where we will start working hard and scaling. But we see that since we've been doing this, we've spent 10% less because we've saved money here, instead of, you know, essentially spending 10% less. So we

[00:06:25] saved 7 thousand. Our cost per, you know, lead, as you call it, went down, but our sales went up from 168 to 200. Our customer base went up from 169 to 200, and our ENCAC went down 11.4%. So, 18% more new customers for 10% less costs, as well as an 11.5% reduction in ENCAC. We had Diane when

[00:06:45] she started, we added her to L1, got twice as much in-app and twice as much CPA in CAC, as if that didn't work. We tried scaling Diane in L0, and our creative testing got worse when we cloned Diane, because that's when we thought, what if we just did the old way... when we cloned Diane,

[00:07:04] it worked. We cloned Diane three more times after cloning her twice, and she kept working. So I said, instead of making clones, starting with one and seeing if we can clone it, seeing if we can clone it, seeing if we can clone it, we want to scale. So, instead of taking one 250 and

[00:07:21] then trying to clone it, what if we took five 100s? So now we're spending twice as much, but it works towards Meta's economies of scale in reverse. So if I have five, if I have these five shares of Diana, if I get one sale, I'm already a third of the way through my other campaigns. So,

[00:07:42] since it's very easy to spend $100 for three days and get nothing, I buy myself the week-long time slot I want. If I put it at $1,000 a day and gave it one day, it would be like, "Oh, well, that did n't work." Like, " Well, did you give this enough time?" I don't know. So I can have

[00:07:56] multiple cases and then compare the average of those multiple cases with each other. If I make one sale, let alone if I make two sales, like right now, I'm six times better than my average right now. So economies of scale, for example, it's much easier to get sales at low costs. It's

[00:08:11] much easier to get cheap sales at low costs. What if I just kept doing it sideways? As you said, the old strategy, and now I have five cold traffic data points, um, um, each, over seven days. And I can compare it to four others. And we're just going to keep cloning it until it breaks. I'm not

[00:08:29] saying it's perfect, but so far we're having record days on Amazon. And because the cost of cold traffic has gone up so much, um, it's something, because since we've been doing this for about a month now, we've had something else that's been fantastic. So, regardless of the record days, which are

[00:08:48] about 20% higher than we've ever seen, our continuity, um, is like this. Our continuity since the beginning is such that we have gained 96% more subscribers. This is similar to stagnation. We just had a whole bunch of people who ended up canceling their subscriptions, but it was still going to happen, like on this day we had plus 11 + 11 + 15 + 14 + 15.

[00:09:14] As if there's no stagnation here. We just had cancellations dragging on. But this thing is on a really good rise, and these are all our monthly subscribers. So okay, go ahead, because it's much more valuable than the app. Yes. Yes. So great with this. We wanted to make it so that

[00:09:28] if we cloned it five times, we didn't want the AI ​​to essentially change it to five different versions and then say, "Oh, now it doesn't matter." So we just said, "What if this was a perfect duplication, and we just scaled sideways?" Yes. So our sales that we should measure is when we make

[00:09:43] changes here in advertising, how did our Amazon, Shopify, and subscribers move? And when we added another 20k this month, did it consistently improve? And so far so good. L0. The one you see here is our creative testing. In fact, it's now been carried over into our main campaign from the start. And this

[00:10:14] L0 campaign is essentially a CBO opentarget, customers excluded with no cost cap for bidders, eye maximization.