7 Meta Ads Mistakes Costing You Insane Amounts Of Money transcript

So many people are making mistakes in their ad account that is costing them hundreds of thousands of dollars. I'm just amazed how often I see these mistakes in meta ads accounts that I audit coming across my desk. I'm going to show you seven of them telltale signs that you could be making more money in your business right now. All right. Number one, if you are still running interest and lookalike targeting, like it is not 2016 anymore.

Stop doing it. Meta Ads actually tells you that interest and lookalike targeting on your ads is essentially a suggestion. At this point, you can look in your ad account setup, at your adset level setup, you'll see a couple little notes that say meta may go outside of the interest setting that you are applying or the lookalike setting and expand our audiences to beyond this. What they're telling you is that they're going to ignore your interest selection and your lookalike selection if they think they can make you more money by going outside of that.

And the thing is they can. The idea of interest and lookalike targeting was definitely how I built every ad account for a long time. Like in the old days, it was the first ads you run or to 1% lookalike and then you go out from there and then you do interest targeting and whatever. That worked. But what Meta has done is basically moved that kind of thinking into the creative itself. And so now according to Meta themselves, according to every ad account I run, instead what you do is you set your targeting broad.

I mean you literally can go like 18 to 65 plus men and women. Set it to everybody. There's sometimes where there is a reason to set an age or gender exclusion based on the product. Um, but not really on a performance level. Don't do it because you think you're going to perform better there. Do it because your product literally doesn't apply to somebody in a certain age or gender. In any case, um, besides that kind of possibility, what you do is you just set the targeting broad, excluding past customers probably, and I'm going to talk more about that in a minute.

Um, but at the level of interest and lookike trying to narrow that prospecting audience, stop doing it. Mostly because it is a waste of your time. It is it is muddying up your ad account. It's probably not necessarily hurting you that bad. Except that usually what it means is that if you're doing that in one ad set, you're doing it in lots of adsets. And so you're building your ad account with the same ad in a bunch of different adsets with different targeting.

Here's the thing. Meta is ignoring your targeting. So all you're doing is just taking one ad and taking the learnings from that ad and spreading them out across a whole bunch of adsets and adding volatility volatility to your spend. Instead, target broad and let Meta do what it does, which is rebuild its own lookalike targeting based on how people respond to the creative. That is how it works now. So you run an ad. Meta gets engagement signal from its users about how people are responding to your ad and then Meta builds a ad level uh lookike audience and distributes your ad according to that basically right it's not going to show you that but that's what it's basically doing right so if if it serves your ad to a thousand people and some subset of those thousand people consistently respond to your ad differently than another subset and maybe they're defined demographically or by gender or whatever meta is going to take that ad and then scale it out, spend it out to other people who are like that audience.

So the creative is the targeting. And that's one of the arguments in favor of creative variation, right? You can reach more people by generating ads that appeal to different kinds of people. Sometimes this is at the level of format. Sometimes this is the level of message. Sometimes it's the level of style of of u who is in your ad, male versus female actors, age, you know, all kinds of different things can do that. There's a bunch of different uh details from there that I'm not going to get into in this episode, but that's the basic idea.

Target broad. Let the creative do the targeting. Stop fighting it by trying to test this. It is a worthless test. Worthless. I mean it. Worthless test that is not helping your ad account at all. Number two, if you are still running adset budget optimization, holy cow, you are costing yourself huge amounts of money. This is probably the biggest mistake I still see people doing. often it is paired with interest and lookalike targeting. people are taking the same set of ads and what they're doing with those ads is they're um they're setting different adset budgets for each ad set um and and you know trying to expand and trying to say you know like you know look alike 1% here's my adset budget interest uh you know crossfitters whatever right uh same same ads both adsets and setting budgets separately no no first of all we've already talked about getting rid of your interest in lookalike targeting instead the thing is what I really want you to notice is when you set up adsets like that, what you are doing is you're making a problem of budget management drastically more complex.

I mean, I've seen ad accounts still where like one ad adset budget optimized campaign, campaign has like hundreds of individual adsets in it, right? And even if you get to, I don't know, 25 adsets that you are managing yourself where you are trying to select how much money to put in the level of each adset, like all you're doing is making a simpler problem or uh the problem of budget management, budget allocation much more complex by splitting it out into a bunch of different places.

Stop doing that. And here's the telltale uh evidence that you're losing money doing this. Okay? Um, it's that you if you go look at if you're running an ad account this way and you have a bunch of different adsets with a bunch of different budgets, look at the difference in the performance of those ads across even large sample samples uh to and see how widely the performance is varying. That should not be the case. assuming that your products have basically the same margin profile and and if uh and this is a big assumption.

Some products certainly have different margin and LTV profiles. But if they do, you should expect that your um performance is is within a relatively narrow band as the sample size builds. Okay? If you're in a small sample size, if you don't have that many purchases per adset, per campaign, whatever, then there's going to be a more a wider variability in outcomes because that's the way statistical noise works. Otherwise, outside of that scenario, right, if you actually build these up, what you'll see all the time is a bunch of adsets with widely ranging outcomes.

And that's because human media buyers cannot efficiently allocate spend across all of those. It is too much of a problem. Look, if I have literally one budget, one budget, like let's say my ad account is one campaign, and I have to make decisions about how to scale or not with that budget myself every day, like I think I'm going to make a ton of mistakes because I'm going to bring my human bias to this problem. I'm going to over interterpret some signals and under interinterpret some other signals.

I'm going to overreact to uh you know someday when I get five extra purchases and so my row looks awesome. I'll be like scale it scale it go. Like it's an impossible problem for me. Maybe I didn't sleep well last night, right? The the example is maybe some maybe your media buyer's um girlfriend broke up with him, right? Or it's who knows? You know, it just could be all kinds of things that lead to human bias and bad decision-making.

And of course, when you take that and then multiply it out by tons and tons of adsets, all you're doing is making an impossible problem for humans, probabilistic forecasting of the future, which is what budget decisions are. You're taking that and you're making it a much much more complex problem. Okay? And so, uh, and so that that's the challenge. Okay? The challenge is not to do that because you you don't want them to do that.

If you go look at those adsets at the different uh uh different adset budgets in in a campaign where you've got a bunch of them, if you're if you've fallen prey to this trap, right? Again, this is how we used to run ads, but those days are behind us. Look at the percentage range based on a 28 day click outcome, 28 day click revenue outcome, okay? Look at the percentage range or really any any attribution you want of different performances there and you'll see like bands of like 100 even 200% variation in the performance of your ads.

That is what you do not want. What you want is to build a system where meta, and I'm going to tell you how to do this in a minute, um, what you should do instead, where meta makes decisions for you about how to allocate your ads. And what you'll see is a narrowing of results. Okay? Let the machine do it for you. Um, this is the value of campaign budget optimization of ASC. Both of these tools can work for this really, really well.

We're assuming you're running ads for the same product, the same LTV profile. Put all those ads across a bunch of ad sets with CBO or in one ASC campaign with 150 ads, whatever. Give it a bunch of variation and say to Meta, "You go figure out how best to distribute those ads." Okay? You go figure out how best to distribute those ads and let it do it for you. Set the budget and go. Okay? You may or may not know that my agency is really boutique.

I keep about four clients at a given time that are true full clients of mine. And uh and so I'm really selective about what tools I use across those clients to make sure that they can be successful. I don't want to get too crazy with all the different things. And Vermont has become a crucial part of my process. I uh I saw Vermont increase the spend of one of my clients by about 50% right away when using it. And I have recently now added Vermont to a second client we're onboarding now and probably will be a third too because Vermont is the best landing page software of which I am aware.

It is just becoming a crucial part of my toolkit. And there's really a couple reasons for that. One of them is pages perform incredibly well. Fastest loading landing pages by far I've ever seen. But most importantly, what it really allows you to do is to build landing pages, including offers and like even like cart upselles and really specific funnels, like full funnels, way faster and way easier than any other tool out there.

And it allows me to do that without touching the main site. And that is crucial because if you know anything about working with brands, you know that a lot of times like you can't go just like mess with the PDP whenever you want to because you want to change an offer. uh you know, you can't go um just like mess with the cart whenever you want to because you want to change the upsells. Vermont creates this self-contained experience where a performance marketer can really do all kinds of offer and message and angle testing, concept testing, and connect those tests all the way through the funnel um to any individual ad they want.

The the ability to uh duplicate and generate additional iterations of pages is incredibly easy. So, you can go really fast. What we do with uh our client who's been using it for a while is basically any ad that starts performing, we start giving it a custom Vermont lander and try to really narrow that funnel specifically and see if we can make that funnel go even further. It just allows me as a performance marketer to move much faster and much easier than any other landing page software out there and it performs great.

I've seen it in the data. I'm doing it with more clients. I'm a huge fan. So go to vermontCommerce.com to get started with Vermont today. It's really awesome. verontcommerce.comf. Number three. Oh my gosh, this one makes me crazy when I see ads turned off in an ad account. Ads turned off. Stop turning off high-spending ads. I am begging you to stop doing this. This is maybe of the things on my list, the thing that is costing you the most money.

There are exceptions to this rule, and let me name them really quickly. Occasionally, an ad will have a delivery error where it seems to go to a really, really weird audience. and you know all of the elements of the ad like recently I have a brand that gets like three four $5 CPCs like really high expensive CPMs works for them it's fine um they were getting suddenly like 10 cent CPCs okay extremely CPMs and like a 40th of of their of the rest of their account CPMs something was clearly broken there but they were getting a bunch of comments in different languages on the ads like we went and assumed that Meta had made a bug so yeah we turned off the ads in that case right I've also seen situations where like extremely clickbaity ads where you do something crazy with the hook that is like sort of outside of the normal product.

I've seen those sometimes get really really high engagement which signals to Meta that there's performance and in fact you don't convert off of that. Meta will realize that eventually but you may get out in front of that turn off the ad if you can really measure that it's that kind of approach to an ad. Besides those scenarios and there are extreme ones like I would say 99% of the ads I turn off are um not for those reasons.

I turn them off because they're not spending at all and I have to open up more space on my Facebook pages to run more ads. Okay, that's the only reason I turn off ads. Otherwise, don't turn them off. Particularly the mistake you want to avoid is you see a high spending ad where the performance is anywhere within the range of right for you and you turn it off because it's underperforming other ads quote unquote so far as you can tell.

The thing about an turning on turning on or off an ad is turning off an ad is actually a prediction of the future. Okay, that's what you're doing. You're you are you are when you turn off an ad, you are predicting that this ad will perform under your target. Okay. And the reason that that's a problem is that Meta is also trying to do that. You and Meta have different opinions of the future. In that case, you are predicting the ad is not going to work.

Meta is predicting the ad is going to work. Okay. So, what is the basis of your prediction versus metas? I'll tell you what meta is. Metas is a sophisticated basian probabilistic machine learning uh forecast of what is most likely to happen next based off of all kinds of data points including some you probably don't have access to. Your decision is probably based on recent behavior of purchases relative to the outcome honestly is probably what it is.

You may look at hook rate or click rate or clickthrough rate or CPC or whatever also and and have some other decisions about that. But like that is like crazy. Okay. Um you are looking at so much less data at so much of a smaller time window comparing it to a much smaller data set of probabilistic performance indicators. Okay. Which is what probabilistic forecasting does. It it looks at the data on one thing and then forecasts the future based off of its comparison to the past.

Okay. you are looking at you're comparing it to a worse and smaller data set in every way you are massively outgunned in that decision and so when Meta is for is spending on an ad that you think is not performing what's very likely to happen is that um you are basically flipping a coin or you know Meta is flipping a coin and it's getting heads five times in a row and that and and for you you might say oh my gosh this coin is tilted towards heads right uh and turn off the ad okay if you need tails uh right um but What Meta is doing is going, "No, it's a coin.

If I just keep flipping it, it's going to come up 50% tails in the future, no matter what happened the last five times." Because it knows that there's been billions of coin flips before that, and it knows that on average, it's going to get 50/50. It just was a weird run of ads. That's essentially what is happening there. Don't turn off your ads. Here's the way to uh evaluate your ads. Spend is performance. Spend equals performance.

If Meta is spending on the ad, assuming that your unit economics are in place in a way that makes sense, assuming you have some sense of financially what needs to happen in your business. If Meta is spending on the ad, the ad is working. Okay, that's number three. Okay, so that is crucial. People waste they lose so much money by turning off ads. It's just cra it's amazing how often I see it that they especially if you're turning off a high-spending ad, very likely what's happening is that you're costing yourself like a ton of money.

That is actually probably like the biggest weapon you have in your ad account is a high spending ad. As long as your account setup is anywhere reasonable at all. And if you turn that off, I mean, you're just you're just really making the game a lot harder and costing yourself a ton of money on the upside. Number four, exclusions. This is another thing I see. Um, so the first thing you want to do is go into your advertising settings and meta, define your audiences.

I would do that with three different sources, okay? Um, use it with your pixel audiences and your custom audiences. use your um email list so that you're sending data, integrating, you know, Clavio with Meta using that information. Have an engaged list. Have a list of past customers. And then that third list is um using Shopify's audiences. If you go into your Shopify account and you see um you know, just search for audiences among your apps and you'll see they have like a retargeting boost audience.

Now they have a like customers plus where they have like enhanced information to give a bunch of IDs back to Meta so they can get the best customer matching possible. Use all of those sources, your pixel, your email list, and Shopify's audiences to define your audiences according to the, you know, according to engaged past customers and then true prospecting. All right? So, do that first and then exclude accordingly.

Go take those audiences at those levels and in your ad accounts, go exclude past customers from your prospecting audiences. Go exclude um well, that's probably the only one I would exclude actually. Exclude your past customers from your from your prospecting audiences. You don't need to go separate out prospecting and remarketing. these days. Um, I'm playing with some things here. The the problem with that for a little while, uh, for the last couple years, post iOS 14.5 has been that those categories are really leaky.

It's really hard for for Meta to actually identify enough of those customers for it to make a difference. So, um, so that used to be the way we would buy ads. I don't think that's probably viable now. Have a couple little ideas around that that might change that, but but right now I wouldn't do that. And if and if there's a win here, I think it's probably on the margins. But excluding past customers is really important.

And the reason for that is actually um is actually sort of simple, which is that I I do think you should be running ads to past customers for most brands. It's just that they should be running at a really different rowass target than new customers. Like let's say you need a 2 to1 for new customers, that means you probably need at least a six to one for returning customers for it to be incremental. If you run it at a 2:1, you're probably just cannibalizing your email list, your SMS, etc.

But if you don't run it at all, you're probably leaving past customers on the table. Like customers are some customers are not going to open your emails. they're not going to be on your SMS list. They're going to forget about your brand. But you drop a new product and you want to get in front of them. Yeah, go run go run an ad to them. Just run it at a higher margin than others. There's more sophisticated ways of um measuring this sort of thing and what that target should actually be.

Incrementality testing could actually be an option here. But but don't even worry about it too much. Like uh you know, if you're if you're at a lower spend, you're not doing incrementality testing. Again, 3x your prospecting number is probably a reasonable place to start. You should be fine. Measure it on click basis. I'm going to talk more about that in a second, but this is the key. If you if you are running ads to past customers, measuring on a click basis.

And I'll say one way brands are losing money is they don't have a separate retention campaign set up. Especially if you're in a category like apparel or if you're making a lot of sort of one-time use goods that people might come back to, right? Like if I was running uh if I was running Ridge, okay, Ridge has a bunch of different products across a bunch of different categories and I've got somebody who's bought a wallet from me.

I would definitely run my bags and my rings and things like that as ads to those customers um and run them actually as ads and and try to sort of reacquire them and particularly if I had new product drops in some of those categories, I'd be putting those into that account. There's there's definitely a reason to do that. You know, I've talked to the guys at Simple Modern about this and they took of course you got to run ads to past customers.

Like they're not going to stay up with everything that you're doing all the time. Again, just run them at a different target. Get those exclusions set up and once you do that, you should be in really good shape. And the reason you should be in good shape for that is now you can actually separate the measurement much more easily. The mistake I see people make is they'll look at a blended campaign. I actually had this happen for an account of mine recently.

I had accidentally left out um an exclusion for one adset for a bunch of reasons. It doesn't really matter why. And when I went and analyzed that adset, I was like, why is this adset performing so much better than my others? And I was looking at rorowass on a click basis and I couldn't figure it out. And eventually figured out it's because that adset had existing customers in it. And so the existing customers were weighting the performance of the total adset because it was just blended into it.

Um they're waiting the performance of that uh in a way that was making it so that um so that it sort of made the whole uh adset look like it was performing better than it was. And it wasn't a disaster or anything like this. It was running on manual bids, so nothing could get really out of hand, but it was still um it was still sort of unhelpful on the whole separating it out really mattered. This matters all the more if your brand has different like AOVs for past for new customers and returning customers.

This happens definitely for brands. So um so because especially if you're running like a lowest cost ad right which is what most people are running. I'll talk more about manual bids in a second. But if you're running lowest cost, like highest volume ads as your campaign optimization in Meta, um then that means Meta is going to go try and find the the um the lowest CAC possible on your customers. Well, it's going to be really um if you're let's say your new customers are worth $100 and your past customers are worth $50 on average, that's the AOV. it's going to disproportionately get past customers in a campaign that's targeting lowest CAC for the simple reason that not only are they more likely to buy because they already have deeper experience with your brand, but the AOV is lower and so it just costs less to convert somebody at a lower AOV than it does at a higher AOV.

And so separating that out can be a really big deal for brands where they have that particular dynamic that that actually can be a major money um a major cost center or or value center by getting that right. Okay, so go do that. If this kind of financially driven, machine learning driven thinking that I am laying out in this episode is something that you're interested in learning more about, there is no better place to do that than through admission from common thread collective.

Admission is the very best learning center. Like think about it as like e-commerce university basically um for any brand that I can think of. If you're in the seven figures, if you're trying to run your ad account with manual bids, machine machine learning basis, think about financial targets, unit economics, etc., and you're trying to build your account that way, admission is unparalleled. So, there's a couple things, a couple reasons why.

First of all, they have guided courses to take you through all of the things that I'm talking about here, like what's the right step-by-step account setup for your ad account. Secondly, there's a lot more courses beyond that around things like cohort forecasting, Google ads, um you know, all all kinds of different elements of thinking about how to run a profitable e-commerce business. If you're an operator, uh and you're building your business, I just can't think of anything that will save you more money on wasted dollars uh put into software and agencies and things like that that are just going to not be useful for you as well as make you more money by directing you towards the best possible outcomes.

There's also um private webinars including with Taylor Holiday once a month where you can do Q&As's with in my view the best of mind in e-commerce. Um so there's there's just so much in admission that if you like this content in this video, you want to go deeper, you want to apply it seriously to your brand, you should go to admission right now and join. On top of that, when you join admission through my link or just tell them that I sent you if you, you know, can't figure that out, uh you will get your a free coaching call and consulting call for your ad account.

So, what you could do is actually rebuild your ad account the way I'm talking about in this video, and then use that free coaching call with a media buyer who's really, really experienced running manual bids in their ad account, doing this kind of media buying, and they will help you take an hour, look over the shoulder of your account, and help you see where any mistakes are or not, and then optimize along the way. That coaching call alone, if more people would take that, I just they would just save so much money on wasted ad dollars.

It's easily easily worth the cost of admission. Go join it today. And on top of that, there's a 30-day guarantee. So, if you join admission, you start doing the content, you're like, "This this is not for me." Then, no problem. You jump out of there, you ask for your money back, it's no risk to you. So, go join admission right now. It's an incredible um program. Go just follow the link in the show notes to this episode.

Whether you're uh listening in a podcast or or watching on YouTube, either way, go follow the link, join admission, tell them I sent you, get that free coaching call, make more money in your business by being smarter with the information from admission. Get the exclusions right. Know who you're running your ads to. Um, and then target both and look for incrementality wherever you can. Don't forget, by the way, that incrementality, the ultimate judge of incrementality is your bank account.

Your bank account is ultimately what's going to um what's going to show the difference between good and bad performance. Uh, if if you are making money, at the end of the day, it's working. And so there you go. All right. Number five, attribution. This can be a quick one, but get rid of view attribution. Let me make this case for you really simply. Meta has recently introduced the idea of engaged views. Okay, if you think about that, the idea of a view conversion, which is somebody who has viewed your ad and purchased within one day of viewing it, but did not click on the ad.

Okay, if you think about that versus an engaged view conversion, meta is telling you there that some large number of view conversions are not engaged, right? If there is a distinction between those two, then it means that some amount of the view conversions didn't engage with the ad at all. So, why would you give it any why would you give the ad any credit for the purchase? Why would you do that if people are not engaged with it?

Why would you do that at all? So the engaged view conversion is is sort of a leading is is an indicator there that lots of view conversions are noisy and so just get rid of them. Especially if you're I'm fine with you running engaged view. Engaged view seems to make sense. It might give some extra signal to meta. I don't run them. Uh but that's not a big deal to me. It's not one of the things that's costing you boatloads of money like or whatever the title of this video says uh or this podcast says, but uh but still still a big deal.

So um excuse me. But but the but the view conversion thing can be a really big deal as view conversions not engaged view conversions. The reason that that's a big deal is because uh uh is because you're going to get over reported rorowass in your windows uh or in your meta ads dashboard with view conversions in there. Um and particularly if you use manual bids, which I'm going to talk about more in a second, this is a problem.

Plus, if you combine this with the lack of exclusions, now you have few conversions on existing customers who are much more likely to buy without clicking an ad anyway, right? So, if you combine those things up, you're going to suddenly look like your rowass is way better than it is. If you go to um in your uh there's two tools you can look at in the ad account here. The first is um in your column setup, you can click columns compare attribution settings and see the difference between your click and view attribution there.

Um or in your breakdown settings, you can look at audience segments. So, search audience segments. Those two things together can give you exactly uh both the last thing I was talking about with exclusions and attribution and get a sense of how different ads are being attributed across your funnel. I recently looked at an account where it was like showing a 3:1 in the sort of like default conversion rorowass optimization window for a brand that had view conversions mixed in with click conversions, but they also had their existing customers with their new customers.

You put that all together and what was happening was um actually like at least half of the revenue that Meta was showing them generating was view conversions on existing customers. And so it was really messy and it was a problem. The simplified way of saying all of this is new customers click basis. You do that and show the full 28 day click window, right? You're going to have a pretty true sense of what's going on. Most likely that's actually still under reportporting the true value of your ads even on 28 day click.

Uh most likely um but it'll get you a lot closer than anything else. So there's that. All right, measure everything on click. Um, uh, especially, oh, by the way, especially if you have organic reach, like if you have any organic presence and you're getting people engaging with your stuff outside of the ad account, then it's even more important to to ex get rid of view conversions because that's going to throw things off all the more.

Um, at the end of the day, measure all of it on your AME, so new customer revenue divided by total ad spend. Uh, track that religiously and measure it on revenue. If you do those things, um, and you look at the profit of your business, forecast according to the full P&L, um, you know, you do those things, you're going to be in really good shape across the board here because you're going to see what's really impacting the outcomes as you make changes.

The bank account never lies. That's the key. Okay. Number six, traditional creative testing campaigns. I'm not going to belabor this. I have run uh I have launched uh, excuse me, I've put out plenty of other content on creative testing campaigns, but I see this all the time. People run new creative and creative testing campaigns where they try to analyze they they force spend to new creative. They try to analyze the results and then move that creative to scale campaigns.

That is maybe the biggest cost center in your business. And I mean that it's it's it's probably in the top three if you are doing that. Just look at any creative testing campaign and try to justify that expense any way you can. Like it it's just like many of them are performing at half the rorowass of your traditional campaigns. You're allocating huge amounts of money to those. It's actually limiting how much creative you can test because it's too cost prohibitive to do more.

Um, I'm going to put in the links of this episode, a couple of episodes I've done on creative testing campaigns, but you got to get rid of them. It's a it's a really bad way to run things in an ad account. Get rid of them. If you want to launch new creatives and new campaigns, I don't really mind that as long as you launch them with a manual bid. Um, if you do that with a manual bid, again, bid cap, cost cap, TRS, that's fine with me.

I don't really care so much that you launch them in campaigns against all of your other ads as long as there's a manual bid. uh there, you know, whatever. That's not really a big deal. The point is, you just don't want Meta to be able to spend on ads that are going to underperform and then justify it because you wrote the word test in the campaign name. Stop it. Stop wasting money. Let Meta decide whether or not your performance works according to uh the measure, the economic measures that make sense for your business.

Okay? Give it those measures in manual bids and then do it. We'll talk about manual bids in a second, but again, just one more time, check the show notes for this episode and you'll see a couple of uh longer episodes I've done on creative testing and why I think it's such a problem and what I think you can do about it. Okay, number seven. And by the way, subscribe to the show. If you if you've gotten this far into this episode, you like this content.

You're going to like a lot of my content. I promise you. Uh just go subscribe right now, wherever you're watching, listening. It'll help. Okay, number seven, manual bids. This is the other big giant cost center in these in these brands. But like when I see businesses that are trying to scale based on their own sense of how to allocate budget each day, I'm just amazed. I'm not going to belabor this either because I put out a lot of content on manual bids in the past.

Um again, I'll I'll link some things that I think are some helpful resources here. But running bid caps, running cost caps, running to Ros allows Meta to um to allocate budget in the ways that are actually best for your brand according to the machine learning that it has. Use that tool. Use that tool. If you need a $100 CAC for your business to be profitable, then tell Meta with a cost cap or with a bid cap. My preference is bid caps, but tell Meta with a cost cap or a bid cap. $100 CAC, go spend as much money as you can.

It will spend more at some times, less at other times. It will be frustrating at times because it will be volatile and daily spend, but on the whole, Meta will much much more efficiently allocate your dollars. This moves you away from you making budget decisions, which is the thing I was talking about earlier, and puts Meta in charge of distributing money according to your um according to the goals that you have for your business.

It will also make it so that Meta suppresses your worst ads and um and uh uh what what was it? Uh distributes your best ads, okay, more. Okay, so it suppresses it suppresses your worst ads, which is really important. And it more fully distributes amplifies maybe is a better word. Amplifies your best ads more quickly. Uh that is what you want. You want meta's machine learning. You want all the power of that machine to um to be in charge of making those budget decisions.

The decision you want to make that meta doesn't have as much clarity to is the actual economics of your business. Where do you need to how much volume do you need um to move the the inventory you have which is definitely a question and then how much what at what cost at what CAC are you willing to move that volume? Those two questions are yours to answer. Meta's question to answer is how do I most efficiently distribute my dollars within that framework?

If you can do that, then you can leverage the whole tool for you. At the end of the day, if I wanted to summarize everything I've said in this episode, it all comes back to one principle, which is this. Who is better at making decisions about how best to allocate my dollars? Me, human media buyer or meta with its um with its machine learning basian probabilistic forecasting machine. Which one of those do you think is right?

In my opinion, and I think this is really hard to argue against, Meta is drastically better at making these decisions than any human are than any human is. And so that's the way I'm trying to build my entire ad account. All of these are the clear and leading indicators. If you're doing the things that I've talked about in this episode, uh then these are the indicators that you're not thinking about it in the same way that I am.

That that somebody is not thinking about how do I leverage the machine as much as possible and let it rip for me. Okay. Um you know, you may need help with this. If you want recommendations for agencies who could help you with this or um or or uh tools places to learn about how to think this way about meta ads more holistically within these sort of financially driven machine learning driven approach to the whole thing reach out to me podcastfgrowth.com.

I'd love to help you with that. Um but this is the way to think about things. This is this is what will unleash for you the best use of an incredible tool for driving efficient profitable customer acquisition in your business. All right, thanks so much for watching or listening. Like I said, subscribe wherever you are doing that. I think if you like this episode, you really will like a lot of the rest of the content that I am putting out.

So, subscribe wherever you are doing that. And please do email me, like I said, podcastfgrowth.com with questions, thoughts, follow-ups, anything like that. I'd love to hear from you. You can find everything I'm doing at ajfgrowth.com, including if you want to work with me specifically. I don't really have any space right now for that, but I might have a recommendation for you, but check it out there. Afgrowth.com. And uh I also put out a lot more of this kind of content on xandrewjer and my newsletter at ajfgrowth.com.

You can sign up for that as well. just drop your email address in the footer or in the popup. I'll send you my four free essential e-commerce resources, uh, documents, things like that that I use every day with my clients that I think will help you as well. Don't forget to follow up with admission, um, which really is the best learning community I know of to get more of this kind of thinking, this exact kind of thinking, um, so that you can get better there.

And also coaching for your ad account, consulting there with people who are really experienced thinking this way about ad accounts, as well as for who is really awesome. I'm using them for a bunch of my clients, for a couple of my clients now. Um, probably gonna add a third soon. Like, for's great if you're a performance marketer and you want to control funnels yourself at scale, offer test, all these things without touching the website.

Forat's an incredible tool. Go check them both out. Uh, thanks so much for watching, listening. I'll see you next time. [Music]