I've Managed Over 100 Meta Ads Accounts. Here's Why Yours Is Broken. transcript

This episode of my podcast is an intervention. It is an intervention to get you to stop doing something that just about every brand owner I've ever talked to does at some point, and you must quit it. It is very bad for you. It is very bad for your business, and it is limiting your progress. I have uh personally managed probably well over 100 men meta ads accounts at this point across basically every category. I've [music] spent on ads tens of thousands of dollars per day, over $100,000 per day for some accounts at peak moments down to brand new ad accounts and everything in between.

I have looked at brands that are legacy brands across huge retail settings and footprints. I have looked at brand new startups and everything in between. The point is I've seen a lot of ad accounts over a bunch of years at this point, and when things go wrong, when there are performance problems in an ad account, over and over and over again, I see people who are working on the account bring up one potential explanation that just makes me crazy because it is nearly always wrong.

In fact, not only is it nearly always wrong, but even exploring it anywhere in the early stage of trying to figure out what's going wrong in your ad account is actually deeply counterproductive, and many times costs brands tens of thousands, sometimes hundreds of thousands of dollars all the way down to their contribution margin because they do this. And I'm going to tell you right now what that explanation is that you must avoid if you're going to be serious about your ad account.

Let's get into it. >> [music] >> Okay, are you ready? Here's what the explanation is. The explanation is Meta is broken. People say this all the time. Anybody who's been on DTC Twitter has seen a thousand tweets that are like, "What's performance like this week?" And then and people chime in, "Something is really broken. Something is really messed up." And there's whole accounts dedicated to the idea that Meta is constantly breaking and constantly wrong and constantly messed up.

And whenever I tweet, inevitably, something about how this is a bad idea to point to this as a major problem in your ad account. Almost every single time I tweet that, I get a bunch of angry replies. People are mad at me because they say, "How can you say that when things are constantly buggy and breaking and etc." And here is how I can say that. Okay? Here's how. I basically never see this problem in our ad accounts.

Now, there's one exception to this, which is overspend bugs. That definitely happens and is a real thing and is very frustrating, okay? And we build some things into our process to make sure that we don't hit too many of these bugs or that these bugs don't affect us too much. We put in the bug report requests. But that's not what I'm talking about. Those are isolated moments, usually over a couple of hours. They're very easy to see in the data because suddenly you're spending way more money than before, particularly if you're on manual bids, and you know, it's at a crazy CPM or something like that.

And so, we'll leave that that aside, okay? That's a big problem. And when that happens, when there's a massive overspend, uh you know, you you have to look at that. Now, sometimes people are wrong about this, actually. Sometimes actually it's user error, not a meta overspend, whatever. But beyond that, uh what I'm thinking about more is when people are saying my performance has declined for week after week after week and and it seemed to all happen on this one day and it's cuz meta made this change or something like that or Andromeda did something, whatever.

Look, meta makes changes and does things that actually affect your performance, to be clear. That for sure happens. But this explanation that meta is somehow broken or constantly breaking or a cousin of this idea, meta used to be easy but now it's hard. Uh all of these things I find to be deeply frustrating, wrong-headed, and like I said, actually really counterproductive. And you've got to get it out of your head that this is the problem in your ad account.

You've got to. You must get this out of your head because it gets you nowhere. I call this explanation for uh and I think I picked up this language from Bill D'Alessandro a long time ago. I call this explanation the ghost in the machine explanation. The idea is something in the black box of how meta works is suddenly stopped working for me and maybe it works in the aggregate for most advertisers or maybe it doesn't, but in my ad account, I am affected in some way where I just can't win because of something Meta did.

And I I think like thinking about this it this way is is a huge huge problem because at the baseline level, the first thing I want to tell you about it is that it's almost never actually true. Hey, before we go any further, don't forget to subscribe wherever you're watching or listening to this and leave a comment. I would love to hear from you about it and I do read and reply to every single one. If you like this content today, if you like where I'm going with this, if you like Meta ads, tactical direct advice opinions about how to grow your business, your DTC business profitably, you're going to like my channel. you like this, you're going to like it, subscribe.

I'm going to tell you in a minute what I think is probably actually happening in your ad account when you see this and sort of some things to explore because I don't want to just tell you stop blaming Meta, but stop blaming Meta. Okay? It's so rarely the problem in an ad account that Meta is actually broken. And the reason I know this also is not only my ad accounts just rarely having this problem if like I honestly I can almost think of never when it has had this problem where it's literally not something I caused or a team member of mine caused or something like that where it's actually just straight up Meta's fault and it's not just like a like I said a couple hour overspend bug.

Not only is that never happened, but when people go to this explanation and when I actually start looking, I can basically always find a different explanation instead that makes more sense. And and then this happened again to me recently. at an ad account where it was like, "Oh man, Meta messed up and and it's like we haven't been able to recover our ad account since." And even people at Meta were suggesting that Meta had messed up and that something was wrong and they couldn't really do anything about it.

And you know what I found when I did the audit? That actually it wasn't Meta's fault at all. Like there's really clear evidence really really clear evidence that the problem was a user error in the ad account. In fact, it was so clear it was overwhelmingly clear. I think it was a 95% plus likelihood that it was that problem. And inevitably this happens. I I you look at an ad account and you say, "Why is performance declining?" And you can find some explanation that actually has better explanatory scope than the idea that Meta is wrong.

And and so and here's the thing, if Meta is broken, I actually had this uh somebody I was talking to you this recent problem said, "Not only are you probably right, you've convinced me of your explanation, but but he also said, 'Look, on top of that, the idea that Meta is broken is a really unhelpful explanation because if that's true, I can't do anything about it.'" So, I actually want to have a bias against that interpretation of things.

And this is a critical point in this. And part of the reason why I implore you to stop saying this. And I know you're saying it, I know you're thinking it cuz basically everybody I've ever talked to at some point has this problem, okay? If if my growth strategist came to me and said, and and I've had this conversation, right? And said, "Look, it seems like Meta is really not working for this client anymore." What am I going to do?

Am I going to tell that growth strategist on my team, "Okay, get on the call with the client and tell them that." Tell them, "Meta broke, we can't do anything about it." Like, how's that conversation going to go? First of all, they may or may not believe you, but secondly, you can't do anything if that's true. Like, there's not a lot of utility in that explanation. It doesn't really get you anywhere. What are you going to do?

Beat down Zuckerberg's door and get him to change your Meta account? You're not going to. Anybody's gone through Meta support, knows that it's like an endless loop of problems, okay? So, so it's just not a productive answer. And so, what I would always say to a teammate of mine is, "I don't actually even care that much if that is the right answer. First of all, my strong prior is that it's not the right answer, okay?" That's my strong prior.

But secondly, even if it is, what what can I do about it? It doesn't help. So, you have to have the responsibility to say instead, "What can I go do to solve the problem?" Because that's actually a productive answer. This is what the client said to me as well, who said, "You know, even if Meta did cause the problem and it wasn't a user error, well, it doesn't really help us. Like, you I would rather believe that Meta didn't cause the problem because then we can actually fix it." And that simple move is the reason why this is such a concerning thing for me, worthy of a podcast episode, okay?

Because if you actually believe that Meta is the problem, then you can't do anything about it. And it cuts off This is why it's so counterproductive. What what happens is when somebody does that, they start going down this route of like contacting meta reps and doing all these things and they're going like, "Well, you know, how do I get you meta to fix this?" And then they just sit there with their ad account breaking.

Or they go start a new ad account or something like that. And and it's this huge giant amount of work to go launch a new ad account. And it's totally unproductive. In fact, it's counterproductive because then you just sit for longer bleeding money or bleeding opportunity cost or whatever it is. But, if you instead go to the go and say, "Okay, here's what I think the problem actually is. Let me go see if I can make some adjustments." Most times you'll find the problem, but secondly, it'll actually allow you to do something about it, which I think you should have a bias towards, okay?

Listen, the value proposition for hiring in the Philippines is so simple. I've repeated it a thousand times and I still believe it today as much or as strongly as I did on day one when I first told you about it, in part because everything I touch is being run in large part by incredible people in the Philippines who I love having on my team >> [music] >> at pretty much every level. And so, what More Staffing allows you to do is go find incredible talent in the Philippines across every element, across every category, [music] across every part of your e-commerce business, your service business, whatever business you're running, people with deep resumes in your space who are ready to work hard.

More Staffing identifies those people, really streamlines the interview process, delivers you just the best people with a great job description, great resumes, gives you a couple of people to interview at the end of their search process. Again, people with deep resumes in your space. And I'm not just talking about $5 an hour virtual assistants, right? I'm [music] talking about like manager, director level talent, people who are going to come in and make a big impact on your business.

And you're going to have access to that talent at a much, much, much lower cost than it would cost you for equivalent talent in the US, and it will be a great rate in the Philippines. That exchange right there is so awesome for your business and for people in the Philippines both. It is a true win-win. I have team members who have been paid at Filipino rates who have bought like four-bedroom nice apartments in Metro Manila, like a big city, city apartment with in a high-rise with a pool, like a nice place, right?

On the salaries that they have made from me, which would be dramatically less than I would pay that same equivalent talent number, you know, again, like at sort of director level positions in the US. It is just like a giant win because your money goes really far in the Philippines and it goes much farther than it would go for equivalent talent in the US. Also, people in the Philippines fall right up speaking English.

Like, I love people from the Philippines, great culture, and just wonderful people who are ready to work hard. Again, if you're paying an above market rate in the Philippines, it's very likely they're going to get a you're going to have highly loyal talent. And More Staffing makes it really easy to identify those people, get them into your business. They'll even give you a one-year guarantee on anybody you hire. So, [music] you hire with More Staffing, it doesn't work out for a full year, they will help you hire the replacement for that person at no additional headhunting fee, which is really awesome.

Go to morestaffing.co/af. Tell them I sent you. morestaffing.co/af. Get started today. Now, what you should actually have the ultimate bias towards is the truth. And if Meta broke, then Meta broke, and nothing you can do is going to fix that, probably. And if that's the true explanation, that's fine. But, let me just tell you, if you're an operator listening to or watching this, like, this is a telltale sign that your agency is doing a bad job for you.

Is that if the agency is saying things to you regularly, like, I don't know, Meta is just messed up, it reflects a lack of deep knowledge of the platform, and it reflects a lack of understanding of what's going on. By the way, let's just be really clear about this. It like it's it's it's not only blame shifting, but it's blame shifting to the wrong person. What you want is somebody who's going to take responsibility to go fix it themselves.

But, of all the places to shift blame, people are going to look at Meta and say, "Oh, that's this trillion-dollar mega corporation that doesn't care about me." And while that's technically in some ways true, right, that they probably don't care about your individual ad account that much, on the other hand, they have an enormous amount of incentive because of how much money they make from your ad account for their ad accounts to work great.

And so, they have all of this money to invest in a really effective machine that conducts this incredible probabilistic auction at large scale, and not even large scale, at unbelievable scale across billions of users all the time. And so, they are very invested in making sure their machine works. And while individual code changes or whatever it is may affect some certain things more than others, blah blah blah, and I get all of that.

I really, really do. I really get it. I am not actually even saying that Meta perfectly predicts the future, or does everything right all the time. What I am saying is that they have a very strong incentive to get this right. And that you ought to basically assume that Meta is right and not you. It's a little bit like over the last couple days I was working on a gardening project when I was off work, and I suck at gardening.

I have no I have no background with it. I don't understand science really at all. And I And so, like you get into, you know, simple horticulture here, and I'm I'm like totally lost. I've never done any of it. I suck at working with my hands. That's the whole reason I'm doing it, actually. It's like a way to learn something and be bad at something, and whatever. So, I'm I'm working on this project. And And over and over I keep having this experience, I'm like, "Gosh, something is broken with uh with this." And then I have to remind myself, it's actually probably not the thing that's broken or the piece that I bought that's broken.

It's probably that I bought the wrong piece or or bought the wrong part, right? Or I need to go back to the hardware store, or I'm installing it wrong, or something like that. The problem is very likely with me, not with the gardening instructions, okay? Not with the orange tree instructions that Claude gave me, although it was wrong about one thing. AI definitely hallucinates. So, like I have to go with that perspective first, because then that actually is the productive path to fixing it.

So, anyway, okay, that's enough. You get the idea. I I just want I'm belaboring the point. I thought at first this is just a tweet, but I want to say this because I really think this is a huge problem, and it the conversation comes up all the time. And so, when somebody, like I said, in your on your agency team does this, like we've actually almost outlawed it as a thing to say to our clients, because I want to make sure it's very clear that we don't talk this way to people.

Instead, we take ownership. So, here's what we actually do instead. And let me tell you Let me Let me break down the actual way to find a problem in your ad account, and I'm going to break it down into four sections, okay? And I think this will be really useful for you. Number one, the first possibility when you see a problem in Meta Ads that you need to fix, here's the first place to look, okay? The first explanation is there actually isn't a problem at all.

Like nothing is wrong. And I'm thinking of two really specific things here that people do with this. Number one, they'll say they'll be looking at bad performance over some period of time and they will not realize that what it really is is just small sample size noise. There is no ability for Meta or anyone else to perfectly predict the future. There's actually variations in outcomes all of the time and people are extremely bad at understanding the idea of statistical noise.

The idea that if I flip a coin three times in a row and each of those times I get heads is actually pretty statistically unlikely and yet it happens pretty often. Right? So there's like a very low chance all told that I'm going to to get heads three times in a row or especially four times in a row or something like that, but it does happen still fairly often in somebody's experience. Like you you do enough sets of three or four coin flips and I didn't go bother doing the math math, but you do enough sets of those, you're going to get some weird outlier reactions.

That's called noise. That's statistical noise, which is not telling you anything about the predictive value of a coin flip, okay? It's always 50/50 and yet you may get some weird times when you get really different than 50/50 results. And therefore, that same kind of thing happens in your ad account all the time. For whatever reason, people convert less or more. Now, there's a weird thing here also where if you're running auto bids, there's actually changing numbers of daily active users on Meta and with changing numbers of daily active users means changing amounts of changing performance, right?

Like you're going to have you're going to pay more for worse customers on day when there's days when there's less daily active users in market for your product and you're going to pay less for better performance on days when there are more daily active users or when people have more buying intent or whatever it is, right? I'm going to come back to this explanation in a little bit, but I'm not even really talking about that here.

I'm just talking about the basic variation and especially if you're spending less money total, it's more likely that you have a statistical noise problem because when you have less total conversions, you're going to have more randomness and weirdness. I've got a client right now that is you know, their average order value is is like a a hundred dollars and they're not spending a huge amount of money yet. And so, it's just like across different ad sets with different products where you're getting a lot of these ad sets with only a couple purchases per day, when that happens, like, okay, you know what happens?

Like, some ad sets look like they get zeroed purchases some days. And that's because the AOV is pretty high, the spend is pretty low, and there's noise. That's what probabilistic prediction of the future does. Means, right? It means that sometimes things go better, and sometimes things go worse at a daily level, but taken over the long view, over the long term, you're probably going to get roughly the performance that you're looking for, particularly when you're running manual bids.

So, people don't understand this idea of randomness, of noise, and therefore they look and say, "Oh, Meta did something wrong for this day or two." When actually it was just baseline variation, okay? So, there's that. Secondly, CPM is a similar thing here. CPM people look at as this really big problem sometimes, and they'll say there's like a ghost in the machine, or but actually like very often it's the case that a higher CPM is not actually a problem.

What it actually reflects is that you're reaching customers that are more valuable to you. And so, in an auction system, you pay more for things that are more valuable. You're more willing to go pay for those customers. And by like willing, what I mean is that as you deploy dollars, Meta's going to go find the best customers possible uh relative to your conversion goal, or relative to your optimization goal, and therefore, you end up with sometimes higher CPMs, cuz you're reaching, I don't know, older, wealthier women.

And those people cost a lot of money to reach on Meta. And so, like, it looks like a problem. I had this I I'm I'm I'm ashamed to say that I was like pointing the finger at CPM for a client of ours at one point. And this was for the specific reason of launching a brand new ad account. And they're really really high CPMs in the beginning. And you know what? Like, we've maintained high CPMs the whole time. We haven't fixed the problem, and I don't blame Meta at all.

I think it's our problem. I think we haven't done a good enough job selling the product yet. That's the reason why we're stuck there. CPM looks like this thing that Meta gives you that's a problem, but in fact, it's a reflection of Meta probably reaching the right customer next. And if you're still not performing well enough at a high CPM, the problem's not the CPM, the problem is your performance, okay? So, that's what I mean by there actually isn't a problem there, okay?

It the CPM looks like something is wrong. It's actually probably not. It's probably just reflective of the reality of things and you have to live and work within that because of what your category is, who your customer is, all of those kinds of things. So, so um often times those are not actually problems and you shouldn't look to fight them. Resist those explanations. There's a reality of sample small sample noise, there's a reality that CPMs are what they are, and that's that.

I mentioned in this episode the need to find better offers for your customers, to deliver more value to your customer by [music] delivering them a better offer that makes more sense. There is no tool I know of that is better and smoother and easier and more complete for testing and measuring the impact of your tests on offers and really all the things that move the needle with your e-commerce business than IntelliGem.

IntelliGem is really a complete testing tool for your website. I hesitate to call it a CRO tool because it's really as as I've as I've said before and as I've heard [music] from the the people at IntelliGem, they like to say CRO beyond the basics because they can just do so much and measure it so well. So, IntelliGem allows you to test all of the basic things like like simple funnel testing where you're you're running just two versions of you know, simple AB lander tests, that kind of thing.

But also you can do the same thing of course with offers. You can test the price of your products on your PDPs. You can test your free shipping threshold, the amount you're charging for shipping. You can charge you can test all kinds of things like your the first the new customer offer that you have on your website that 10% off you get when somebody subscribes to your email list. All of those things which really make a big impact and now they've even introduced the ability to test in checkout and in cart upsells as well as post-purchase upsells coming from them now.

So, you can eliminate some other software costs by just bring that to IntelliGem. All these things that matter so much for how your business generates value for customers and how they generate value back for you. And of course, you measure all of it actually on the basis of profit per visit not just revenue per visit or conversion rate because IntelliGem ties into your COGS data. So, you can see like what's the actual contribution profit on these tests Uh, is the thing that really matters the most.

So, go to intelgems.io if you're serious about testing, and you should be. intelgems.io, use the code Faris20, f a r i s 20, to get 20% off your first 3 months. intelgems.io, Faris20, you get 20% off your first 3 months. Do it today. Okay, category two of problems. These are things that are not in your control, okay? But, they're actually not a ghost in the machine. So, they have real impacts on performance. They're not at all in your control, but they are actually not a ghost in the machine problem.

What I mean is Meta isn't broken. Performance really is worse, but it's not actually something you can control per se, or your control of it is really sort of different than what we think of as like control from a media buying perspective, okay? There's two of these. One of them that that I could think of first. One of them is seasonality. This is another thing that people just really underrate. Like, your TAM changes a bunch based on the seasonality of your product and your category.

There're just going to be times when your product is performing like much better or worse based on the seasonality than it would other And seasonality is something that's really hard to predict. There's some obvious versions of this, right? Which is like Father's Day, Mother's Day, uh, holiday. But, I could think of another client at one point of mine that had big-time summer seasonality on a particular ad angle for a particular product apply to the rest of the angles and products.

And it was just really hard to see, but when you go look, and this is the thing I love to do, go look at Google Trends for the seasonality curves uh, for or for the search term volume around some angle that you are pursuing or for some product that you're pursuing. And what we found, for example, was that for for one particular client, customers were searching a term much, much more from spring into summer than they were in the fall.

Now, the weird thing about this is you would sort of expect them to be searching that term a little bit more with like heavy meals around holiday cuz it was related to health and some of that kind of stuff. But, actually, I think they were searching for it around summer because they wanted to look good in bathing suits. And so, it was still health related, but it wasn't like New Year, new you health-related. It wasn't about sort of like relationship to food at holidays and sweets and treats and all that kind of stuff.

It was just about looking good in a bathing suit, and therefore the the the search term was way higher in the summer, and the performance, the ad spend, especially with a manual bid, tailed off dramatically once you got to like August and September, and it tailed off consistently, and percentage-wise, it tailed off at the exact same curve as the Google Trend spend. So, it was very tempting to say we did something wrong with the media buying, and that's why we weren't reaching people a certain way, and that, you know, that's not a really ghost in the machine explanation.

That's like a a we did we did the problem with the problem explanation. It was also tempting to say that Meta was doing something, like Meta had burned out its audience, whatever. But, it wasn't those things at all, I don't think. I think it was just purely about the seasonality of the demand around the particular message around the particular product that we were pursuing. Now, seasonality explanations, or a related one, macroeconomic explanations, like, "Oh, the economy's crazy." These have a similar thing to the ghost in the machine explanation, which is that there isn't much you can do about them in one sense.

And in fact, I think this is a really uncomfortable thing about them, and it's why I try not to use these explanations as much either. But, sometimes they're really necessary if you really get to them because they help you understand more about your customers, and that's a critical thing and a big difference between seasonality explanations and ghost in the machine explanations. The seasonality explanation tells you something, like I said in the example before, tells you something about how customers are relating to your products, why they want it, what is selling them on it, what felt need, felt desire they have, and how you want to communicate them with them in a different kind of year.

And so, the implication of that seasonality explanation I I mentioned before was that we ought to go develop a bunch of creative around that angle starting in January, build up the creative machine so that we can go invest a whole bunch so that when March and April came around and the and the curve of demand started going up into the, you know, look good in the bathing suit in the summer explanation or message, right?

When that happened, we would have a whole bunch of creative attacking that angle. And that's a really important thing to get right, and that is something you can do about it. It also meant that in the seasonal off-season, we ought not keep hammering away at that same creative angle. That was sort of a waste of our efforts. And that was actually a really important thing because then it was like, well, we need a new angle for a different time period or we need to focus on different products or something like that cuz we are not going to hit if we focus aggressively over here, okay?

And And so in that respect, you can do something about it, but you have to do something about it at a broader messaging and brand kind of level. The other version of this that is sort of like not in your control but not a ghost in the machine, this is going to sound like a ghost in the machine but I actually think it's a feature not a bug, okay? And it's what I would call faster ad burnout in the Andromeda age. And basically what I mean here is that is that it seems clear to me at this point.

I talked about this in the in my episode a couple weeks ago about the uh Meta Performance Marketing Summit. It seems clear to me at this point that Meta actually has gotten materially better at efficiently delivering the right ad to the right person at the right time faster, which means your ads reach their best audiences more quickly and more efficiently than ever before. That's a great thing in a lot of ways, but in other ways it's a problem because what it means is that your ads also burn through that audience faster, that sort of potential audience.

So, and there are there are legitimately limits to how many people you can reach with a certain angle at a certain time and a certain piece of creative. This is why creative diversity matters so much. And so, that explanation for things is sort of not in your control either. It's actually the opposite of a ghost in the machine. It's that the machine is so efficient that you go fast. And sometimes what people will see is an ad that sort of like rips for a couple days, burns out, and then you end up going like, "Oh, what what happened?

How come that ad won't come back?" And it's like, "Well, nothing. Like, you leave the ad on, you let it keep spending a little bit, it won't spend as much as it did before, and that's a reality of sort of just how creative works in those moments." By the way, one other seasonality one cuz I just forgot to mention it and I want to come back to it is uh I I saw this in apparel really significantly where you'd see really crazy demand curves in certain categories, you know, like denim or jeans or something like that where it was I was working with mixed by Nazrin back in the day I've talked about a lot and we we talked about this publicly but we had an ad that like really ripped for like three or four weeks and we're like man this is amazing but then it faded away because fashion demand cycles are really fast a lot of times for certain products and the seasonality is so significant that it just it just moved really really quickly.

It's so in in categories like fashion you got to really really watch that seasonality one. It's actually not a ghost in the machine, it's not a media buying setup problem either. You're pointing at all those things there's there's seasonality. And again, what that led to was a different approach which was Nazrin turned around so I got to develop more products that have more seasonal moments to them, right? So or more seasonal resonance at different times and she was over indexed on a certain kind of product that was really only going to work at a certain time of the year.

So she realized I need to make more summer products, I need to make more fall products, I need to make more winter whatever, you know? It's just to fill out her whole collection, etc. And so that's where that is really helpful. So faster ad burnout and seasonality, not in your control but not a ghost in the machine. Number three, things that are in your control, the explanation for the problem that you're experiencing they're in your control and you got to stop doing it, okay?

Because this is these are the ones that just kill me when people make mistakes in the ad account that they feel like is something wrong with the ad account and in fact it's not the ad account's fault, it's actually just a straight up media buying error, okay? And and this is this is this is where like I said this kills me when people point to make these mistakes and then point to the machine, okay? Number one, did you change a setting somewhere like your exclusions or your placements or your gender targeting, right?

You decided to go male only or female only or you look and you you look at the age breakdown you're like oh we we suck at 18 to 24 so we're going to get rid of that audience, you know? Um and you do that and suddenly your performance gets worse because you see the ROAS of that audience and you're wondering about it, whatever. Or um or you do this with placements. Oh I got to get rid of audience network, I'm over delivering there and it's a big problem.

Look, again, occasionally something like this actually does happen, but this really, really occasional and the simple move here is to actually kill that ad and relaunch it in a new ad in the new ad set if you see a crazy placement thing, but most times by going and changing your exclusions or changing your placements to fix a problem like this, you actually are going to make things worse a lot of times. Now, you may actually need to update your exclusions to properly target your ads, but in that case, it's also not a ghost in the machine.

Your ads are going to deliver to more existing customers if that is who is most likely to buy. That's again a feature not a bug. And so, if you're updating your exclusions, you should expect some level of volatility in the performance. You should make sure you know what you're doing with that. You may want to actually keep your existing customers in there or put them in a separate ad set and give specific budget to them or you may want 18 to 24-year-olds in there and you didn't realize it before or you may want every placement or something like that.

But people go and change some of these settings, see performance get worse, and then say like, "Oh, Meta is just messed up and I got to get rid of these settings cuz they're just they're just killing my ad account, right?" And and and almost always that's actually counterproductive, but there's actually a bigger one than that. That's like a small part of things. I don't see that affect things as often. The bigger one that's more obvious than all of this and I hope you know what I'm going to say right now is turning off ads.

I mean, turning off quote-unquote low-performing ads is I mean, it is often a disastrous decision. It just really is. I've said this a lot of times. Turning off low-spending ads is not a big deal to me. If you have ads that are not spending very much, regardless of performance, turning those off because you need to get more ad space in the ad account limit, that's fine. But if you have ads that are spending, unless the performance on those ads is like wildly worse, like wildly, wildly worse than what you think uh it ought to be or what your target is.

If that ad is wildly worse, then yes, you may need to turn that ad off. I did this with a client recently. They were selling a product where the engagement on the ads was good. The ads were not clickbaity, but they just weren't converting at all and I think there's a few reasons for that related to the offer and the lander and some of those things. As they started to do those things, like I turned off the ad set because it was so so far outside of the range and they just didn't need to sell that product enough.

Like it was a sort of not a main focus product anyway and it wasn't wasn't important that we like go and solve that problem really for the long-term health of the brand. It was sort of a secondary product, etc. So you just kind of kill it. But in the in fact, I didn't kill an individual ad. I killed because it wasn't an individual ad. It was the ad set because it was a targeting a product that just was not selling through.

Okay? I mean they had an offer problem basically. But turning off high-spending ads, I have seen absolutely nuke account behavior and people think they're doing the right thing. They're like, "Oh, it's at a 0.7 ROAS and I wanted a one ROAS or whatever and so I got to bring that I got to get rid of that." But if it's one of your top spending ads especially, you can get really really bad really fast here and then start pointing to other things as the problem.

And that is a mistake. Do not do that. If your ads are spending a lot of money, you need to know that they are probably playing some role in the ad account. Meta is modeling conversions back to the ad account and if there is a bunch of a bunch of spend going to those ads, you should probably turn them back on. Basically, basically the first thing I do every time I take over an ad account is I go turn on all the old high-spending ads that somebody else turned off.

And you know what happens? I almost always get some level performance back. Almost always. It's just like it's just this crazy reality of what happens. People turn off ads that don't that that they think are hurting the account. You turn them back on and wouldn't you know it, there's 30% more spend in that ad account at the same target as always because somebody was looking at this wrong. There's a million reasons people turn off ads that are wrong reasons, but the baseline position should be that in 99 plus percent of cases, you don't turn the ad off.

It's really really rare that you should do it. Okay? Number three, the other thing that's in your control that could be a problem here is if you have a bad pixel installation. And this is where like go check your data update. What we've started to do, just like shout out to my friends at Elevar, we've just started to use Elevar for everything and the reason why is not because I have a strong opinion about Elevar versus Blotout versus Billie who was an old sponsor of mine or anything like that.

Like I've come to think that like most of the versus even a manual install or versus a Shopify install, there's a really simple reason is that Elevar is a known and trusted commodity in the space. It seems to do a really good job with the data and most importantly, it's like an insurance policy on the ad account. The reason I use it is because if something goes wrong with my pixel, I have somebody who's an expert who I can talk to who can look at it.

That's really it. And I just want to like have that installed be as good as possible. They have a strong incentive to make sure that they're passing back the most data possible, CAPI is set up well, all that stuff. And then we've got somebody to do because I never want problem messing with my ad account and I can't do that. And I have seen a lot of times where performance looks really different than other than it should and it's because of that.

I have inherited two ad accounts in the last year where we've taken them over and immediately found a double reporting pixel. And when the ad account in that case that the ad account looked extra good because it was over reporting revenue, okay? But it's still is weird thing where the ad account is reporting something really, really different than the actual performance. And when that happens, don't blame Meta, don't assume it's a modeling problem, don't think there's a ghost in the machine, go ask the question, did somebody install the pixel twice?

And you should go check go check that, okay? And yep, that was what happened. We fixed that and wouldn't you know it the the performance went back to to where Meta's reporting was pretty much reliable. It can also happen the other way where like a poorly installed pixel can under report revenue. It doesn't tend to happen as often I don't think, but just have that dialed in, make sure it's going because very often when something is wrong, it's because you did something you did something wrong with the pixel uh at the reporting level, okay?

Number four, let's say you've gone through all that, your ad account performance is still not as good as you want and you're struggling, all right? What do you do if you say if you've sort of crossed off the there isn't actually a problem bucket, if you've crossed off the not in your control but not a ghost in the machine bucket, you've crossed off the in your control and you didn't do any of those things, you didn't turn off any ads, whatever, but you still have ad account problems, okay?

It's not seasonality, it's not in macroeconomics, it's not anything else. Your ad account just isn't working anymore. Here's explanation number four and this is the hardest pill to swallow for people and it's this. Your brand hasn't actually earned any more scale than you're getting. You just haven't earned it. Like, listen. You're selling stuff to people. You're asking people for their money in exchange for your products.

Wake up to the reality that you must provide value to them in excess of the cost. You must do that. They don't think about it in those rational terms, but that is what they have to do. They have to believe in their minds and in their hearts that the value you provide to them is more than the money that they cost them. And if you are not doing that, people are not going to buy your product. And that means your creative needs to get better.

Ezra Firestone said on my podcast recently something that really stuck with me. I've heard him say this before, but it just landed with me a lot this time. He said, uh there's two things in your brand. There's product and promise. And product doesn't win. The best product doesn't win in the marketplace. The best promise wins in the marketplace. That's what good creative is supposed to do. It's supposed to be the best promise.

How do you give the best promise possible communicated in clear, compelling, effective ways, okay? So, your creative needs to get better. Your offer needs to get better. I've talked about this a lot recently, but it's this it's just this thing where you should know what the customer needs to engage with your product. And maybe that's a big bundle discount that makes sense for both you and for the customer. Maybe it's a sample offer.

Maybe it's a subscription offer. Well, it could be all kinds of things, but you need to think more clearly about the offer. Maybe there's a guarantee you need to put on it cuz people are just nervous, right? Whatever, okay? Um those those are the kind of like tactical fixes, but number three, right? And this is the thing that I think is the really really the thing at the core of this a lot of times and is the hardest thing to fix.

And it's very possibly the case that you just have limited resonance. You just haven't done the work to communicate enough how great your product is to your brand. And I believe your product is probably pretty great. Most people I talk to do a really good job caring about and efforting the product again all the time, okay? And And again, you go back to my Ezra Firestone episode and you'll hear him talk about how in his businesses he is constantly trying to make the product better cuz at the end of the day he sells product to customers and he wants them to have a great experience.

That's how you generate retention, all those things. I I see that a lot. I think most entrepreneurs I talk to care that their product is good, and they should be efforting that all the time. I don't think this is a problem. I think what's really a problem that they have not actually built a brand that resonates deeply with people, and that they care about. They haven't done the things of finding the right influencers and building the right social content and doing the sort of non-DTC, non-direct response things that they need to do to make it clear that they are the right product for a certain kind of person all the time.

Your brand just hasn't achieved the residency yet. And if you haven't done that, you're not going to reach the kind of scale that you want. You have to actually be great. You have to actually make an emotional connection to people. You must do that to reach the kind of scale that you want in some way. And or you have to expand your product line enough to go do this. Like, you know, it's not always an emotional connection.

Sometimes it's just uh cutting across categories really, really well. Unless you are the ultimate tactical monster, and I think that's even it's questionable whether that's a real thing, okay? I still think there's an element of the promise here that is really important. But, unless you're there, okay? You have to do these kinds of these kinds of things, and I think most brands love to actually do them when they happen.

They they love creating great moments for their clients or for their customers, and you ought to really go do that. It's very possible that there's no problem at all in the ad account, okay? And that actually you just haven't earned the scale that you're trying to get to yet. And that's what you have to do as an operator. You got to go do that. Listen, resist the idea that there's a ghost in the machine. Next time you see somebody tweet, "Meta is broken this week." Be the one that replies to them and says, "Probably not.

Probably not, okay?" Instead, be the person who says, "You are broken this week." You don't have to say that. You know what I mean. "You are broken this week. You did something wrong." Or maybe you're not broken, but you're misunderstanding misinterpreting something. Own it. Take real ownership of what's going on, and be better in the process. If you do that, you'll solve the bigger problems, and you won't get stuck blaming somebody else for something that's actually your fault. >> [music] >> Thanks so much for watching or listening.

If you want to work with an ad agency that does not point the finger at Meta, you should hire us. You should go to ajfgrowth.com, tell me a little bit about your business, fill out the intake form there. We'll get in contact, get on a call, and see if we're the right fit for you as an agency. ajfgrowth.com, like I said, is the place to do all of that. You can also just email me if you really want to. podcast@ajfgrowth.com is the best place to do that.

Again, like I said, thank you so much for watching and for listening. You should subscribe wherever you're doing that if you like this content. You will like a lot of my content. You will like most of my content. So, uh subscribe, leave a comment. I would love your comments on this particular episode. Am I crazy? Am I just like way too uh Pollyannaish Pollyanna-ish? How do you say that? You get the idea. About Meta? Like and and maybe maybe I should be a little more skeptical and a little more angry at Meta.

I don't know. Tell me in the in the comments. I'll read every one of them. I'll reply to every one of them or close to every one of them. Big thanks to More Staffing and to IntelliGemz for sponsoring this episode. They are both great. Check the links for those in the show notes. Don't forget the offers that you've got there. A year with More Staffing where there's a guarantee on the employees they give you. 100% off first 3 months with IntelliGemz.

So, go check those out as well. Thanks so much. I will I've got a bunch of great interviews Oh, I've got a bunch of great interviews coming up. You're Just you don't want to miss them. So, subscribe. That's the point. Do that. I'll talk to you soon.