Sometimes people say things like: "Sam only shares simple, common sense advice, he never teaches anything advanced".
Today, mostly for fun, I want to share something advanced. (really advanced...).
Most of the time I share simple strategies anybody can use to make dramatic improvements to their life and business. I do this because 90% of the time, it's the simple things people are missing, and no advanced strategy is going to save them from a fundamental (but simple) mistake. Common sense is NOT common practice!
But today, fuck it. Let's go down the rabbit hole and twist some peoples minds.
Today's video is "Ninja Shit -- Playing To The Biases Of Facebook's Algorithm".
In this module we take things to the absolute extreme... True ninja shit!
If you want to truly dominate Facebook ads you need an edge over the competition. Competitive edge is everything! You're constantly fighting other advertisers to get your ads in-front of your audience and if you don't know how to fight, you'll never get in the ring, yet alone knock the other guy out...
To get an edge, you must understand the nature of the Facebook machine. You need to know how it "thinks".
Just like a human, Facebook's algorithm has biases, heuristics and tendencies that rule its behavior and decisions. Once you know its biases, you can play to them and form the ultimate symbiotic human-machine. The result? Total domination of your entire market on a global scale. (this will blow your mind).
Check it out and let me know what you think in the comments below?
Here's what we cover:
1. Inside the mind of the Facebook machine 2. Playing with the “extremes” 3. Engagement bias 4. Lexical sentiment bias 5. Graphical sentiment bias 6. Financial bias 7. Consistency momentum bias 8. Next level tracking 9. Inside man advantage
This video is a "module" from my Uplevel Consulting training program. Uplevel Consulting is an 8-week program that shows you how to take your business to the next level using leveraged business models, machine based operations, marketing automation and Nasa grade Facebook ads. If you're tired of manual processes and relying on unreliable humans to run your business -- Uplevel Consulting will show you how to transition from "Atoms" to "Bits".
Hey everyone, Sam Ovens here, and welcome to this training module called Ninja Shit. Now, you're probably wondering what the hell sort of title is that for a training? What am I gonna learn in this thing? And what could we possibly cover or what could we possibly learn that we haven't already learned inside Uplevel's Facebook training so far?
And the truth is is that the modules which you've seen so far in the Uplevel training for Facebook ads, that is more than enough. If you just master the stuff you've already seen, that's enough to take you to multiple millions of dollars in revenue per year and just run laps around your competition. Just non stop, right? That stuff is more than enough to get you there.
However, and I just wanted to do this for fun, and to also just show you how far you can really take mastery of something. Like it's one thing to run ads, it's another thing to run them profitably, it's another to spend a million a month or more on ads, but then it's another thing to show you these things which you're about to learn.
My goal with showing you this is to really open your eyes so that you can see how far you can optimize a system and master it. Because I'm gonna show you this for Facebook ads, and I'm gonna show you exactly what to do and all of the hacks. However, you can apply this sort of thinking to any system, not just Facebook ads. So my goal with showing you this is to teach you a way of thinking and analyzing as well as teach you how to do this with Facebook ads. But my goal is that you'll be able to take this and apply it to other systems and areas in your life so that you can develop total mastery.
So here's what we're gonna cover. We're gonna start off by going inside the mind of the Facebook machine. And what I mean by that is really understanding the algorithm that runs Facebook's ads and it's content delivery system. So why does it show this thing to that person? How does it make that decision? And what are the formulas and equations that this Facebook algorithm uses to decide all of these different things? You know it's a very complex sophisticated machine, and if we're gonna master it, we have to understand how it thinks.
And then once we understand how it thinks, I'm gonna show you how to play with the extremes. And this is a fascinating thing which you're gonna learn. Really machines don't understand extremes the way human beings do. And I'm gonna show you how we can play with extremes to really get the algorithm to start humming.
Then I'm gonna show you a bunch of different biases that the Facebook algorithm has. So because the Facebook algorithm thinks and it makes decisions, it has biases. It has preferences, which it weights greater than others. And we need to understand what those are so that we can not trigger the bad things, and that we can absolutely trigger and unleash hell on the things it likes.
And these things are, first of all an engagement bias. So obviously Facebook's algorithm loves it when there's engagement on stuff. Then a lexical sentiment bias. This is basically where their algorithm can read the words inside your posts and your ads and see if it can understand them or how happy or sad they are, and it gives preference to different things. This will fascinate you and I'm gonna show you how it works.
Then graphical sentiment bias. And this will pretty much blow your mind, but you know Facebook's algorithm can also crawl images. Every image you're putting into your Facebook ads, every image you're posting on Facebook; the machine is looking at your images, and it can see what's in them. And I'm gonna show you an actual example of this so you first of all believe me and you can see what it does. And then once you understand how it works, I'm gonna show you how to play to this. I'm gonna show you the sorts of things that the algorithm likes to see in images. What things it has a bias towards. And this is fascinating stuff.
Then I'm gonna show you financial bias. You know Facebook is a company. Companies like making money. And Facebook's algorithm does have a bias for money. And I'm gonna show you how you can signal money to Facebook so that you can really make it dance.
And then consistency momentum bias. Facebook's algorithm, trust is a very important thing. The last thing Facebook wants to do is burn its users by doing something sketchy or delivering some advertiser to its user base and that advertiser was sketchy. And so Facebook is very big on trust. How trust is earned is with consistency. With consistency comes trust, and then the reward of which is momentum and extra delivery. I'm gonna show you how this works and how you can signal consistency and trust to the Facebook algorithm.
Then I'm gonna show you how to take your tracking to the next level. So it's one thing to just use Google tag manager and just let the Facebook pixel into a container and put it on your landing page. It's another thing to let your Facebook track people even if they're incognito. Even if they're using Safari. Even if they're using ad blocker. Basically being able to track everyone, everywhere, and really give your Facebook x-ray vision. And I'm gonna show you how that works and how you can implement it today.
And then finally, the inside man advantage. So a lot of the things which we're covering here are technical digital things which you can do to really get the algorithm to give you preference in advertising. However, the final one is actually human advantage. And here I show you why some employees inside Facebook are very important towards helping you win with your Facebook ads. So once you're spending a lot of money you can get a Facebook [rip 00:06:52], and rips can do things to your accounts to really help you gain an edge. And I'm gonna show you how all of that works.
So we've got a lot to cover today, let's jump right into it.
So inside the mind of the Facebook machine. So Facebook is a system. It's a complex machine built of interconnected parts. And like all systems, Facebook has inputs, processes, outputs, feedback and environments. We need to understand the nature of the machine if we want to master it. Now Facebook's brain is a machine. And algorithm that makes billions of rapid decisions all the time. And while the machine isn't human, it was created and updated by humans. Those human intents, the intent of the humans that designed it, built it, and still to this day maintain it, are inherited by the machine. And Facebook's intent is to survive. It does this by keeping users on its platform. To do this it needs to keep users engaged. And to do this it needs to provide a good user experience. And by doing this, it can sell ads, make money and grow. So that's what Facebook's intent is; to survive.
First of all it wants to make sure it maintains what it's got. And then it's ambition is to grow even bigger than what it is. So, when you can understand the intent of Facebook, you can understand the intent of the algorithm that's kind of running everything. And Facebook's eyes are the input of data, or the inputs of data, that feed the brain so that it can make decisions. Which, serve it's intent.
And to master Facebook, you start by understanding its intent, then how it makes decisions to serve its intent, then the heuristics and parameters it uses to make those decisions, then the data inputs it uses to run those heuristics. Right, you want to understand these things. Once you understand these things, you know what to do so that you can do this and expect that. And it just works.
So here's what this looks like. You know here's what Facebook ads and the Facebook platform looks like when we set it up like a system diagram. And so over here we have the inputs which we put into the system. And these are users, posts and ads, campaigns and budgets. So these are different inputs, and of course there are way more than these. These are just some of them which I've selected to help illustrate my point here. So these are the inputs that come in, and then there are processes that run in the machine. This is what we refer to as Facebook's brain, or the algorithm. And it runs processes.
To really figure out how to run these processes, it uses things called heuristics. And a heuristic is basically a rule of thumb or a bias that it uses to make things simpler for itself. And so if it's trying to decide what should I show out of Advertiser A and Advertiser B, you know it will use a heuristic to determine that and quickly make a rapid decision. And the algorithm is trying to figure out show what to who? When and how often? And then should we accelerate it or slow the delivery? And is this advertiser safe or risky? Should we trust them or not trust them? Right? It's trying to do a whole bunch of things, billions of them, all the time.
Then we have outputs. And this is the result of the inputs coming in with the processes applied to them. The outputs our results. You know the things that we get. This is things like reach, so how much of our audience do we reach? Impressions. How many people do we serve impressions with? Then clicks. You know how many clicks do we get to our landing page or our funnel? Which is also traffic. And conversions. Like how many people actually buy or register or whatever? And then we have feedback. Feedback is basically the feedback mechanism that comes from the output and feeds back into the inputs so that ... it'll come back and attach to this input so that the input is different when it comes back in again.
So even if you create an ad and you put it in, then it's gonna process a supply to it and output's gonna come out; like impressions, clicks and whatever. But then, that output's gonna feed back around and change your input again. And so even though it's the same ad, the output is coming back and feeding back into that input.
Now let me give you a perfect example. You create an ad, you put it in and it starts working and lots of people like it, it starts getting conversions, and you would get a relevant score of like nine or ten. Well, but that output of people liking it and clicking the like button, commenting, sharing and all of these things, and clicking the link; that creates a quality score of ... sorry, what's it called? A relevant score of like nine or ten, and then that comes back around and it attaches that to the ad. And then it comes back in and it's got an advantage. So that's how feedback works. And it can rapidly accelerate and improve your ad, or it can rapidly throttle and harm your ad based on the feedback that comes back around with it.
Feedback can be things like attention. So if someone's scrolling through their news feed and they see your ad and they pause on it and hover on it, that's attention. And then likes and comments. So if people are clicking like on it and commenting, sharing, clicking the link and converting; these all basically feed into different performance metrics.
Then we have the environment. And ever system operates within an environment. You know there's no such thing as a system within a vacuum that's totally independent and air locked from the outside world and influences. And you know Facebook's the same. It has political, legal and financial things that are going on. So you know investors might be expecting money, then Facebook's gotta make money. Legal things might change like the GDPR. The political winds might shift and you know now privacy's a big concern, and we've seen this a lot with Facebook. And it does affect what happens in this thing.
And then we have competition and technology. So as Facebook has new competitors like what YouTube might do or AdWords or Snapchat, what all these different things might do, it might affect Facebook. And also competition that we face. You know if there's a lot of people using Facebook and advertising then that changes their environment too. Same with the technological advancements, things like that, and popularity and demand.
So all of these different things are going on all the time, every thousandth of a second, billions of things going on across this network. And this algorithm here at the heart of it, it has to decide basically everything. And understanding how this thing works is really the key to mastering Facebook. Because this is the thing that makes all the decisions.
You know the times have changed. Back in the old days, if you wanted to run an ad on the TV or in a newspaper, you called up that newspaper and you would say, "Hey, I want to run my ad," and they would choose whether you were allowed to run your ad or not. And then they would basically choose where you got put in the newspaper, how often it ran, and all of that. N so a human made all of these decisions.
But, today, with most online advertising it's not humans making these decisions. It's an algorithm. And it's making billions of decisions all the time. So, just like you might build a relationship with a human to get advertising advantages back in the day if you were buying print ads or whatever, you want to really build a relationship with the machine so that it can give you some advantages. Basically things are the same, it's just a machine instead of a human.n like a human it thinks. And like a human it has biases. And like a human it likes people that confirm its biases. So, let's understand this thing.
So it starts with its intent. Like a human, it has intent. And intent is what it really wants overall. And its intent is to balance all variables in order to survive and grow. So it's looking at all of the stuff that's coming in. All of the outputs or the feedback, the environmental changes, and it's running all of these processes and it's just trying to balance things so that it can survive. That's its most important intent; survival. Second one; growth. Overall that's what it's trying to do. And if its intent is to survive and grow, and it's decisions are things like show what to who, when and how often, and should we accelerate this delivery or slow it, and is this advertiser safe or risky? You know it's trying to make all of these decisions.
How it's really going to make these decisions is things that serve its intent. So just like a human, if a human has a goal, then a human's gonna make decisions. And those decisions are going to be heavily influenced by its goal. And it's basically gonna make decisions that are aligned with its goal. But it also has to balance these things because it's very complicated. Because if it just wanted to make money, I mean it could go crazy and just start giving advertisers all of the stuff.
But remember, it has to at the same time balance its users. Because if its users get pissed off and leave, then there's no one to show ads to. And if there's no one to show ads to then there'll be no advertisers. So it really has to balance these things. It's like we have to keep our users on, keep them happy, deliver a good user experience. But also make money and also grow and also survive and also make sure that we're helping these advertisers get results so they keep advertising. So it's very complicated and it's doing this for billions of people all the time. Just, it's an insane job. No human being could do it. And that's why it is the job of an algorithm.
But obviously it can't make perfect decisions. Because it doesn't have all the information necessary. It can't call people. You know it can't go and interview people and do all of that. So, to do that it has to use heuristics. Heuristics are basically rules of thumb. They are ways that you can make a decision fast just using a mental model or a rule of thumb or something like that. And it's not perfect. It means that you're not gonna make a perfect decision, but you're going to make a near perfect decision very quickly. And that's what Facebook's algorithm uses. It uses a ton of heuristics.
Heuristics would look like if this then that. So, let me give you an example. So if lots of people like this ad and comment on it and share it, then quality score equals nine. Or quality score equals ten, or whatever. And then this signals that. So if quality score or relevance score equals ten, then that signals that this ad should be given preference in the auction pools, and it should also be given additional reach. Right?
Now whether that ad is actually like by people, and whether that ad should really be given more reach and more preference in auctions; we don't know. But, Facebook will use that heuristic to basically take that action. And most of the time it probably is right. But, some of the time it probably isn't. These things can either serve you if you know how they work. Or, they can harm you if you don't know how they work. And it's very important to understand them. It might also weight the importance of something 80% and this thing 5%, and for example, it might weight a comment at ... let's say it weights a comment or a share at 80%.
I know shares are weighted the highest. And then comes comments. And then comes likes. So a like might be 5% percent. And then tolerate up to this threshold might be another heuristic. So, you know tolerate this many trigger words before disapproving ad, and tolerate this many ad disapprovals before banning account. And tolerate this many banned accounts before banning fan page and business manager. Right?
So it has all of these tolerances built into it. And it also has biases. So you know a bias is just a preference of one side of a scale versus another. So like a human being's bias might be to prefer night time over day time. Right? That's not necessarily true that people should prefer that over that, but it's just a bias. And machines have biases too. They look like this in machine logic. They value this more than that. So it might be value credit card payments more than PayPal payments; which is true. It actually does have a bias towards credit cards versus PayPal in terms of payment methods, and then give preference to this. So, you know if somebody has more history of ad spend and they're coming up against you in an auction, then Facebook will use a bias and give preference to the person with the most historical spend.
It also might have a bias saying throttle this. So if a post of an ad or a post receives so many negative complaints or don't show this ad feedback requests, then it might throttle the delivery and stop giving it additional reach. And another one might be accelerate this. So if lots of people are clicking like and the sentiment of the comments they're posting and if different people who have more influence start commenting and sharing, then Facebook might be like all right, this looks like one of those things that people really like and can really help us, so it just starts rushing the delivery to it. And then it might have another one which is like trust this, don't trust that. So it might trust certain keywords, and it might not trust other keywords. Or it might trust certain domains and not other domains.
So really to understand the Facebook brain and the algorithm is to first start with its intent. Remember it's trying to balance everything to survive and grow. And remember it has to cover two sides there. It has to look after the businesses. It also has to look after its users. And if it can look after its users and the businesses, then it looks after itself. So it's always trying to balance those things. And when you're using it, you have to keep that in mind.
Sometimes I see a lot of advertisers and they don't. They just behave, like Facebook doesn't care about the users that are seeing the ads. Which is crazy. And so they end up getting themselves in trouble. They just think oh, I'm just here to use this thing and just be aggressive and sell stuff. And it doesn't work because if the users get pissed, then Facebook won't let you even advertise. So you need to make sure the users are happy, but also, you're not just there to entertain you're also to get customers. So it's a balance. What you're doing. It's a balance of looking after the users, adding value to them. But also, being direct and aggressive enough so that you'll get a response. So it's very nuanced in there. And you need to understand how Facebook makes its algorithm decisions using heuristics and the biases that are present within all of this behavior.
Don't worry if this sounds complicated. I'm gonna make it dead simple and show you what all of these things are in this training module. So influencing the machine. Now, we can't change the machines heuristics and biases. But we can change its inputs. And you can feed it the inputs it prefers so it makes the decisions you prefer it to make so that you can get the outputs you desire.
So you know we can't go into Facebook and plug our laptop into it and just start changing the weightings of the different algorithms and its decision making criteria. We can't do that. And so in order to influence Facebook, if we can't change those things, then we ask ourselves how can we change the inputs that we give it to change the outputs it gives us? So that things are more preferential and better for us so that we can do better. And I'm gonna show you how to do this in this module.
And in this module, I'm gonna show you how the machine thinks and what to feed it, so that you can make things go boom. And by boom I mean really make this thing start climbing up additional gears and really getting into the performance zones that it can get to. And to get into those zones and to start getting to that next level, you have to play to the algorithms biases.
Now, playing with the extremes. So humans and extremes. Humans are skeptical when they're presented with extremes. And if somebody is overly happy we think oh, that person's strange. And if something is really good we think oh, it's too good to be true. Right? We human beings, you know our heuristics, they have like too much warning signals. So if something's too much, like too much of a good thing, then we start to be skeptical and we're like oh, something's going on here, something's wrong.
But, machines are different. And machines and extremes. So machines don't understand too much of a good thing. If something is programmed into the machine is good and they receive an extreme signal of good, they consider that to be preferable. You know, more good. They don't really understand when there's too much of something. And they can be programmed to understand when there's too much of something, but most of the time most algorithms aren't really programmed to do this. And a classic example would be like the Google search algorithm, right?
So back in the day, not even ... oh, I guess it was quite a while ago, but still not that long ago really. Like five years or something like that. People figured out that if you put the key word you were trying to optimize for multiple times throughout the text in your page, that Google's algorithm would favor it and show you if someone searched for that keyword. For example, if someone searched for the key word Sam Evans, and I had a page that had Sam Evans like a hundred times on it; then it would rank better than other pages that have the key words Sam Evans [inaudible 00:28:40]. So that's how a lot of people started hacking traffic from Google by just doing key word stuffing.
However, once Google started to catch on, it changed that and it now is much more sophisticated. But, there's still ways to really get Google's search algorithm to give you traffic by playing with extremes. And with Facebook there's still tons of ways to get it to favor you and give you traffic by playing with extremes. There are still tons of them. Because it's not as blatant as the Google search algorithm one. You know it's very simple and blatant to think oh, if someone's searching for this key word then I should include this key word multiple times on this page. Right? You can see how people came to that conclusion pretty quickly. And how Google shut it down pretty quickly.
But with Facebook it's nowhere near that blatant. Most of the things that Facebook's algorithms like people aren't aware of. And if they are aware of them, they're not sure how to manufacture them and provide them to extreme levels. So that's very different. And that's good if you know what to do.
And all or nothing. So, machines have an all or nothing tendency. Most content gets no reach on the internet. Now most people's posts, most people's photos; they get nothing. Barely anyone looks. And some content goes viral for the entire world to see. I'm sure you've seen this too. Like how come some things just get seen by everyone in the entire world, and the vast majority of things just really don't get anything? Well that is because machines have an all or nothing tendency. And those all or nothing tendencies are triggered by extremes. With machines neutral gets you nowhere. And extreme goes viral. So remember that. Neutral nowhere, extreme viral. And understand that machines have this all or nothing kind of tendency.
So biases plus extremes. The key is to understand the machine's biases and then feed it extreme inputs of things it considers good, and avoid feeding it inputs it considers bad. And this is how machines fall in love. You know if a machine has a preference for this thing, and it doesn't like this other thing, and if you just keep giving it the things it loves, and you give it none of the things it doesn't love; then it's gonna fall in love. And humans are very similar too. So that's how it works. And when it falls in love it's gonna treat you well. It's gonna give you lots of traffic, good prices, preference over the competition, lots of reach, and the best traffic at the best times. It's just gonna give you everything. So we need to understand what these extremes are and how we can use them.
So knowing where to push. Facebook's algorithm has biases towards the following things; and our strategy is to satisfy the machine's biases by feeding and extreme amounts of the-
... machines biases by feeding it extreme amounts of the inputs it considers good and, I'll add to that, we try to limit the amount of things that it perceives as bad. When we get a huge waiting on the good side and not that much on the bad, it's just going to go for it. The different things are, these are the different biases that we're going to play to. Engagement bias, lexical sentiment bias, graphical sentiment bias, financial bias, consistency momentum bias, next level tracking and inside man advantage. That's what these things are. These are going to be the tools that you use. Now let's jump into them and cover each one.
Engagement bias. Facebook loves engagement. It wants users to be on the platform as much as possible and to do this, users must be engaged. To do this, the algorithm serves impressions to engage in content and ads. What is engagement? It's a word but how is it really defined when it comes to the algorithm? How does the algorithm see engagement? Engagement is a measure of human attention and interaction with a post or an ad. The algorithm starts measuring engagement by people slowing down when scrolling past your ad and then hovering over it and then pausing and reading it and then interacting by liking, sharing or commenting on it and then clicking it. Most people don't understand that Facebook is able to understand how much people like your ad just as people, just as soon as people start slowing down in the newsfeed when they see your ad. It doesn't need to wait until someone actually goes all the way and clicks. It starts pretty much immediately, as soon as people start slowing down, where their mouse hovers over the screen. Then if they start reading it, it can start sensing all of these things. Then a mouse hovering over it, then actually reading the whole thing, then interacting and then clicking. It reads all of these things.
The combination of all of these things, it's able to determine how much people like it and how engaged people are with it. The more engaged people are with it, the more love Facebook is going to give it. It's one thing to understand it but how do we cause engagement? Or if we've already got something engaging, how do we make it more engaging? Humans run passive recognition processes when scanning their newsfeed. They look for things that serve or threaten their own personal interests or anything out of the ordinary way that they see life. Humans are basically just going into Facebook and scrolling through there looking for things that are going to help them serve their goal or also threaten their goal and also things that are just out of the ordinary. Basically a human being is looking for, all right make sure there are no threats, if there's any things that can help me, let's consume them and also let's make sure there's nothing kind of out of place here. That's basically what's going on in your brain, even if you don't think so consciously. Unconsciously you're doing that.
It hasn't changed much since the old days when we were doing this in our environment by scanning the landscape looking for a partner or that might have been what we were looking for. We're looking for food or things to eat or things to drink and then threats, dangerous animals and then also things out of the ordinary like, hey what's that doing over there? Then creating engaging ads. Back to causes of engagement for a sec. When something really serves our interests it gets engagement. When something really threatens our interest it gets engagement and when something is out of the ordinary it gets engagement. Those are the triggers.
How do we create engaging ads? To create engaging ads you must understand the minds of the people in your market. You want to understand these things, you can create engagement by inverting popular belief, countering culture, polarizing opinion, exposing corruption, extreme truth and transparency and disrupting existing an established order, et cetera. Let me explain these things a bit. Inverting popular belief would be like finding something that is believed to be true and just turning it upside down, just inverting it. Now one extreme example of this might be like the flat earth theory. People who say the earth is flat. Because that totally inverts our understanding of the earth, it really causes a ruckus especially on the internet. That's why that thing has so much massive momentum and it gets so many views and it goes viral on the internet because it inverts popular belief. That's an example of it but what we're trying to do is we're trying to invert popular belief and be right at the same time because we don't want to invert popular belief and be wrong because then we're just going to cause a big ruckus and then everyone is going to think that we're an idiot or they're going to hate us and then we're going to get kicked off Facebook.
When I say these things, we have to try to invert popular beliefs but we have to be right in what we say. Now let me give you an example because I'm talking conceptually here and you might be like, "Sam this is a theory, give me an actual thing." One thing I talk about a lot is deleting social media, getting rid of it. How it's actually it doesn't cause you to get results it actually makes you fail. That you shouldn't use it. That's basically inverting popular belief in my market. Most people in my market just say, "Follow me on this, follow me on this. Post a million times." Gary V is out there telling doctors to basically, "Don't serve your patients, don't hone your craft, don't get better at being a doctor, don't save any lives. Just go sit on your phone and look at hashtags and just try to post and get people following your hashtags. That's basically what the popular belief is. I just invert that because by inverting that, one it's controversial and it gets engagement but two, it's actually true. I get to win with that one. You're looking for things that you can invert, cause a ruckus but be right about.
Now countering culture that's kind of going against the way that culture is. If everyone is putting Lamborghinis in their damn ads, I'm not going to put a Lamborghini in my ad. I might instead put a scooter or a Hyundai or something in my ad just to go totally against it because that sort of stuff gets attention. It is probably more true that you don't actually need a Lamborghini to actually solve problems for people and help people and make a lot of money.
Then polarizing opinion. Whatever the public opinion is you can just polarize it. Exposing corruption you're saying this is what everyone wants you to believe and this is why it's wrong. Then extreme truths you can really tell a story about an experience you've had or something like that that is extremely transparent and true to the point that it shocks people. Then disrupting existing order. That's basically when you don't really expect, you don't really respect the authority or the established order of your market or your industry and you start disrupting it by pulling into question why people are above other people and why people should start questioning things and basically just causing a big ruckus. The key is to cause a ruckus for the good of the people otherwise you're just going to get yourself into a lot of trouble. These are dangerous tools. These are tools that start revolutions. You don't just go around using them just for the sake of making a ruckus because you certainly will cause one and you might even get yourself killed. You've got to be careful with these things and you've got to use them for the right reasons. You've got to be right with it.
The real magic, the magic combination of this is to be extreme controversial and true. This is the best combination for engaging ads because it causes controversy and most importantly it gets away with it because it's helpful for people to understand truth regardless of how hard it is to swallow. If you cause a bit of controversy in your ad but your overall message is good for people and it's actually helping people and by countering the thing or inverting the thing that you did you were more right than the status quo, then people will, you won't just have a bunch of hate you'll have a little bit of hate but you'll also have people who really get it and they're like, "Yeah this is true" and thinks it's hard to swallow but it's true. That's good. That's important. You need to have both sides of this. Also, you will have Facebook on your side too because you're not just out there trying to make a ruckus, you're not using hate speech or you're not causing mayhem for no good reason. Then the people like your ad, you're doing good for the people too and Facebook's algorithm is going to like your ad and people inside of Facebook if they manually review it, they'll like it too. That's the key.Now these are hard to get all in one thing but practice makes perfect.
How can you actually create engagement in your ads? What are some tricks to doing this? Well the first one is don't look like an ad. I don't know if you guys have noticed but people don't actually like it, they actually hate them. If people don't like ads, don't make your ads look like ads. That's a dumb move. That's like thinking, "Somebody likes meat" sorry, "Somebody likes" let's say somebody is a vegetarian and they don't like meat. Let's just serve them meat. That's not a good idea. They're not going to like that. It's as simple as that. If people don't like ads, your ads shouldn't look like ads. They're still ads but they shouldn't go out of their way to look like them. What you want to do is you want to look at the ads serving your market and do everything you can to not look like them because people run a pattern recognition process for seeing if something is an ad or not. Let's say in my market everyone runs ads next to Lamborghinis and stuff. People, as soon as they see someone doing that they're like, "That's an ad. That's an ad. Ignore. Ignore. Ignore. Ignore." But then if something doesn't fall for that same pattern where there's a Lamborghini or something in it, then somebody doesn't catch it. It gets through their kind of like their barrier. It gets their attention now because they're not just auto blocking it.
It's important to understand what ads look like to the market that's seeing them from all of your competitors then to purposefully and deliberately not look like them so that you can infiltrate. Very powerful trick. Of course, it's still going to say sponsored on your Facebook ad. You can't get rid of that but it doesn't matter about that. Most people are actually detecting what's an ad and what's not way before they really look at the sponsored thing or not. They're doing it using much more advanced processes. You can get past that. Then you want to look like a friends post. Facebook is for people to learn about other people and what you want to do is you want to make your ads look like friends posts. You want to write as if you're writing to the individual person whose reading the thing and avoid corporate speak and you want to get personal. You want to evoke emotion. Humans engage with emotion. You want to create shock and awe, you want to agitate, you want to trigger anger through truth, you want to tell stories, show vulnerability and humility and let people know who you are and build a relationship.
You're not writing this as a business. You're writing this as a human and the person on the other end is a human. People totally screw this up when they think they're B to B. They're like, "I'm a business and I'm selling to a business. Let's talk like I'm a business and this other business is going to be reading this." They just make all, they just write the most horrid piece of text I've ever seen in my life that really doesn't appeal to anybody because I don't know if anybody realized this but businesses aren't on Facebook searching. A business can't search. A business isn't alive. People use Facebook. People read things. People click on things. We're talking to people and we ourself, we've a person. It's a person to a person. It's not B to B, it's P to P. You want to get into this frame of mind and picture the person and how they might read through it and their reactions and their responses to things. That will help you build and write a better ad.
Be human, look human. Humans engage with humans. Wow. That's very insightful isn't it? Humans engage with humans. This is a no brainer. This is so obvious. If you put a human by themselves in a room and you just give them a cog and a can, they're not going to be very happy. If you put them with another human now they're really engaged and interested in stuff. Humans like humans. Advertise using a person fan page. What I mean by this is, if your advertising, advertise as yourself. You should have a personal fan page. Now business fan pages can work too but from everything I've tested and seen with other people too, personal fan pages work really well. They work better because people are on Facebook and people like seeing posts from people. If you've got a business fan page that you've always been using, I would suggest creating a personal one too and just testing it next to it. I bet you it works better. Now with your personal fan page you should use your headshot as the profile picture because people will see that and people expect to see a photo of a face there and they expect that face to be you. They expect you and that face to be named with those words that are the name of the fan page. These are just simple things that make you look like a human.
Include yourself and others in the image of your ad. We call our fan page Sam Ovens and what that fan page and the profile picture there's a profile of me and then people can put a face to a name. They're like, "Okay this must be Sam Ovens. This is what he looks like." Then as they start reading it they're like, "Oh he must have written this. This is that guy whose face I can see whose name is Sam Ovens writing this thing." Now they can start forming an image in their mind as they read, which is the magic of words. It's not the words themselves, it's the image people visualize in their mind as they read. That's the piece that does the work, not the words. You need to be aware of what things create this image people see in their minds as they read. When people read those boring B to B corporate vomit ads, they don't see any image in their brain because there is none. That's why those ads work so bad. All good, captivating things allow people to visualize as they read. By including images of you in your ads, people further are able to build a relationship with you and visualize you.
You should also include other things and people that important to you and your life. Things like it might be your girlfriend, your wife, your son. It might be you playing a sport, it might be you just working at your desk. Just things that are actually you and make you you, include them. These things help. Then this one is a powerful one. Tell the truth. Humans are incredibly perceptive and can tell what's true regardless of how well you mask it. Don't pose for your photos, have somebody take them when you're being you without you knowing or posing for the photo because photographers know this well. When you're photographing someone, you want them to just be themselves and just pretend that the camera isn't there. Then the photographer can really capture the essence and the moment and everything. Those, they look so much better. Then when you try to pose it just looks so awkward and fake. That's a good trick. Also for ad images you want to look at images that were taken of you just by your wife or girlfriend or partner or boyfriend or husband just when you were doing normal things and you weren't expecting to take this photo as an ad but use it as an ad. Those things work really well.
It's amazing how good humans are at perceiving the truth, like in anything. People have very good perception of this and it runs subconsciously. It's not consciously. In the conscious, when someone is analyzing it, they can only really see what they're seeing and make their judgements they're making but there's some things going on subconsciously that really, something doesn't add up or a personal is really skeptical or they've got a bad feeling. That's the thing that is detecting stuff that isn't true. It's very good. It's exceptional at it. You want, we don't try to make it look like we're telling the truth, you just tell the truth and that is so much more effective. One, it's easy. It's easy as hell because you just don't have to make anything up. You just say exactly as it is. But then it works so damn well. While other people are trying their hardest to fake it and with each ounce of fakery that they're putting in, which is a huge effort comes worse results. You can try not very hard at all and get amazing results just by telling the truth. When you do this, you'll look less impressive but people will engage with it 10 times more because they can tell it's true.
Example. If I had an ad next to say 20 Lamborghinis that weren't mine, then people would probably see right through that. It wouldn't work. The ad wouldn't work and people would just be racked up in the comments. The ad would probably get banned and then my whole account probably gone. All right? But instead, if I just had a photo of me on a scooter like on one of those bird scooters I ride around here in Venice, then because that's true it's probably way more likely to work. Now I look certainly less impressive on a bird scooter which you can rent for a dollar compared to standing in front of $25 million worth of Lamborghinis but because one is true it just works way better and it's way cheaper too. That's what you should do. Whatever the truth is, just capture that and then tell it in your ads too.
Another one is don't delete negative comments. This is another you're telling the truth thing. Negative comments, it's normal to have people that take a negative point of view from your ad or post. That is normal. It's actually weird when there isn't any negative comments. You want to leave them there and you should be getting about a balanced 50/50 between people who like your stuff and people who don't like your stuff. If you're getting 60/40 or 70/30 or 80/20 negative, you need to, you're doing something wrong there. It should generally come about 50/50 and that's when you generally get it right. When you've got it balanced like that you want to just leave them. Leave both sides of the argument there. Leave both sides fighting each other because all of that energy that they're putting in just fuels the fire of the ad. Just let it go and use it. I honestly love it when people start going at each other in the comments. I don't reply to anything, I don't pay attention to anything, I don't even read them. I just use all of that human potential energy to just fuel my ad and they're doing it. Good on them.
Now there's a type of ad and a style of ad that really plays to these things. It really gets good engagement on Facebook. It's something that I called, something that I named Trojan Horse ads. Trojan Horse ads are ads carefully designed so that they get attention, look personal, not like an ad, come as a gift, get accepted, draw the reader in at the right moment, make a pitch and then a call to action. Pretty much the same strategy as the Trojan Horse strategy that was executed way back in time. You know the Trojan Horse story where they basically assembled a massive, giant horse and left it outside the gates of the enemy's castle and the enemy looked out and saw this massive horse and thought, "What the hell is this thing? It looks like a gift or something." They sent some soldiers out to assess it and they're like, "Just looks like a big, giant horse." Then they thought, "All right, let's just wheel this big, giant horse inside our gates and just bring it into the heart of our castle because we wouldn't want to leave this gift out in the rain out there." Then they wheeled this thing in and all these soldiers came out of the horse and then attacked everyone and then they won. Genius strategy. Whoever thought of that back then, that's some serious next level strategy.
But we can use a similar line of thinking for creating our Facebook ads because most people's Facebook ads look like armed uniformed soldiers beating on the doors of the castle. They look like ads so blatantly they're literally screaming, "I'm an ad, I'm an ad, let me in." Obviously no one is going to let that in. You don't look like that. You've got to come as a gift so people accept you and bring you in. They let their guard down but there's all these different things that we have to achieve in the sequential order to really create the ultimate Trojan Horse ad. Number one is to get attention. Nothing gets attention like something out of the normal or out of the norm. You want to be relevant but out of place. For example, with the Trojan Horse strategy back in the day, a massive giant wooden horse, no one had ever seen one of those things before. When one of those is sitting out, it takes their attention. They're looking out from their castle walls and probably just are used to seeing mountains and fields and maybe a little river and maybe a bird. Then one day there's a giant horse there. People aren't going to be able to just let that one go. It's going to steal their attention.
You want to think like that too. In the modern day on the Facebook ads landscape. How are you going to get people's attention? You've got to remember what the landscape looks to their audience. If they're just seeing people in front of Lamborghinis and then you just look like a person in front of a Lamborghini, you're just going to look like a soldier and people aren't going to let you in. You look normal. It has to be out of the norm. Be relevant but out of place. I'll show you some examples of this soon.
Come as a gift. Genuinely add value and help people. This will get your ad read and accepted as a gift. In order for somebody to, with the Trojan Horse example back in the day, the enemy thought it was a gift so they were like, "Oh let's take this horse inside." Our ads should be like that too. It should actually add value. We're not really trying to pretend like we are, we should actually genuinely add value. How do we do that? By solving problems, giving them insightful advice and giving them stuff they can actually use to improve their life and their business without having to click the link in the ad. It's actually getting it accepted, getting it read and people are like, now they're consuming it. Now it's inside the gates. Then get accepted. Ads shouldn't be pattern recognized and judged too soon. How people pattern recognize and judge really soon is through stereotypes. You want to disobey the stereotypes. Make sure people can't class you in the category or stereotype as they do everyone else. That's why it's important to understand what everyone else looks like so we don't look like that. That way we're not getting judged. People are going to take the manual long process to understand us and then judge us instead of just snapping to a fast decision.
We don't want to pitch or claim too early. There shouldn't be a pitch or a claim in your ad too early because that allows people to quickly pattern recognize it and reject it. You want to try and bury that down a bit further, deeper into the text. Then you want to make your ad long. This is the text of it. You don't want to make it long for the sake of being long. That's stupid. But longer than normal. This gets buy in because people have to read through it and it also gets the read more link. The read more link is basically what can I show you as a quick example? If I just show you this ad example, you can see here it starts saying the text and then it says, "See more." Actually it's not a read more link, it's a see more link. If somebody wants to read more than what's up here, they have to click see more. Once they click this, now they've kind of done a micro commitment to say I'm interested I want to know more and also they're now deeper into a zone of focus.
By making longer text ads, we get the see more button which is a benefit. It helps us. We're also able to bury the link, sorry bury the pitch and any claims and stuff deeper into the text of the ad so that only people who are actually interested get the pitch. People who are interested are more likely to positively accept the pitch and click it. But people who haven't read really anything and get the pitch or a claim, they're not likely to accept it and complain. That's step five is to bury the pitch. Don't make any claims, pitch or call to action until someone is deep into reading the ad. Just explained that. Then six.
Earn respect. Ads that pitch without value or ads that pitch without giving value, they get rejected and it's very important to make sure you add value and help people and earn your right to pitch. How we do that is by writing a really good ad, actually adding value. Helping people so that if they pause and spend five minutes of their time reading this ad, they're actually going to be like, "Wow. That was a good piece of information I just read" or "Wow, that's insightful. That's going to change how I do this and that's going to improve my life." You can actually help people without them having to click in all of that and that's important because remember Facebook-
... Click and all of that, and that's important, because remember, Facebook's algorithm, it's trying to balance things. It's trying to make money from advertisers and get them results. But at the same time, it's trying to look after its users and give them a good experience. If you want to play to Facebook's algorithm and how it's trying to balance, you want to try and balance too. You want to try and get customers, and you want to try and get people to take action, and to do that, you need to pitch people. But, you also need to earn your right to pitch so that you don't piss people off in the process. And you're going to piss people off, but not piss off too many people. You're going to have way more people like it than not like it. That's important. And it's important to earn that respect.
When you create a really good Trojan horse ad, Facebook will love it to bits and it will just let it go and run with awesome results for years. Honestly, years. One of our Trojan horse ads has run for three and a half years straight. I kid you not, three and a half years, it's made more than $3 million, and it still works to this day, because it earns respect and it earns the right to pitch, and it still does.
And then the final one, number seven, is to twist minds. You want to invert public opinion, counterculture, disrupt order, and leave their heads spinning and be remembered. Why we want to do this is because a lot of the time, our ads, they have a long tail attached to them. What I mean by this is that someone might see an ad today. They might not click it, they might not read it or they might read a little bit. But, just because they don't do that, it doesn't mean that we shouldn't care about their perception of us.
We want to try and get people to click and buy from us today and do the actions we want from them today, but we also want people who don't click to be left with an impression in their mind that they can't get rid of, because then they're going to remember us, and then they're going to be thinking about us all the time. Then, they're not really going to know why they're thinking about us all the time. The reason why is that we just broke the conventional patterns in their mind, and so they went to run their pattern recognition to check that it all was well and normal in the world, and they found something to be massively out of place to the point that it twisted their mind and they can't stop thinking about it. But they don't know why they can't stop thinking about it, but they are.
Then, when humans think about something a lot, they assume that it's important. When someone thinks about something again and again and again, they have this bias that, oh, this must be really important, because I'm thinking about it a lot, and the things I think about a lot are important, regardless of the context of why they were thinking about a thing. So now they think it's important, and they're thinking about it, and now they see another one and bam, you're good. You're good to go. You're doing work for the future as well, and you're laying ... You're sowing seeds for future harvests by doing this properly. This is what creates a really, really good ad. A true Trojan horse ad.
That's how it works in theory, but what does one actually look like in practice? I'm going to show you some Trojan horse ad examples, and specifically, I'm going to show you five of my all-time best performing Facebook ads. All of them have made more than $1 million each and worked for six to 12 months minimum. One of them has worked for three and a half years and has made $3 million. Let's take a look.
This is called Trojan Horse Ad Examples. It's a resource available for download beneath this video in the resources section. Facebook Ad #1 is one of our best ones of all time. We call this one Reflections. The headline says, Reflections On Reality: 3 Strange Concepts That Made Me $20,000,000 By Age 27. Then it starts off, "Never aim for your goal directly. Always move in angles. This defies basic human logic because most people believe that if they want to make some people, they should try and make some money. And most companies that aim to just make some money rarely make any money. This is because money derives from value and value derives from helping people."
This is a Trojan horse ad because first of all, the headline is Reflections on Reality. It's not normal. And then "3 Strange Concepts"? That's not normal. It's pattern ... It's not able to be pattern recognized, and then it also, though, includes some money, which can help. If you can get away with including a claim of a little bit of money ... This is very hard to do, by the way. The times have changed, and the preferences on the internet have changed with advertising. It's pretty much impossible to get a financial claim in a headline anymore. You probably can't do that.
But if you can, I mean, it can work great. If you're not in a niche that helps people make money, this would be weight loss or getting a relationship or doing yoga, whatever the thing is that you're helping people achieve, this is where it would say that. Then at the top in this ad, there's no talk about a claim. There's no talk about making money or an offer or a call to action. Instead, what I tell people to do is move in angles, and that's like, what? How the hell do you move in an angle? You mean you want me to walk horizontally everywhere I go, really? So it's kind of just gets your attention a bit.
And then people who believe they should make some money should try to make some money, and I'm saying that people who aim to do that don't make any. Now I'm inverting this public opinion. I'm just twisting it right upside down, because a lot of people who want to make money try to make some money. I'm out here saying that that's not how you do it. I'm just turning it upside down, basically. That's what I was saying before. Inverting public opinion, and it's a great way to get attention and to get engagement. You can see, this ad's had like 300,000, almost, likes, 3,000 comments. It's had 9,000 shares. This ad's made ... I think this ad's made like $2 million or something. Maybe. Maybe it's made more, but last time I checked it was $2 million.
I've included this, the headline and the body copy, and you'll see this body copy is long. It is all of this. It's long. I do a big intro piece to it, and then I actually give advice. I actually start teaching some stuff in here. The first one I teach people is the self illusion. The second one is the mirror, and the third one is patterns of existence. Each one of these is an actual full thing, you know? I teach quite a lot in here.
Then after that, I basically say, I hope this helps you on your journey. I'm actually providing information that helps them. Then I turn it around and I say, did you find this post valuable? If so, I have good news. I put together a free training here. So I basically just tacked on at the end of ... This is a helpful post, which could be a blog post. Then I just tacked on at the end, did you find this valuable? And if someone's like, yes, then it's like, if so, I have good news. So you can see how I'm earning the right to pitch. If they didn't find it valuable, they'd be like, did you find this post valuable, they'd be like, no. Then I say, if so, I have good news. If not, then I wouldn't have good news, right? And they'd probably just be able to ignore this.
So only if someone finds value in this am I really saying that they should pay any attention to what I've got. Then I give them the free training, and then you can see there's a claim down here, but it's buried underneath a lot of text. And all of this text is interesting. It's abnormal and unusual, and it actually adds value and it actually helps people. That's a true Trojan horse ad.
This is like another one. This one here, I say, 9 Millionaire Consultants Were Created With This Advice (Ignore At Your Peril). So, it's kind of interesting, especially with that image there, with like a boat and stuff in the background. Then you can see, again, it says see more, and I talk about these three things. Then, How A 26 yo Punk Started A Consulting Business In His Parents' Garage, Moved To Manhattan & Made $10M In 4 years. This one's old, because I'm 29 now, and I ran this back when I was 26. It ran for about two years, and it made more than $2 million. This one was a really good ad. It's basically a story, but it looks like news. It looks like it's a news article or something, and with the image, it works really well, too. There's the body copy and the headline and everything.
And then, How Did These 9 Entrepreneurs Make 7 Figures And Get On This Private Jet?, that's another good one. But one of my best ... I think this is the all-time greatest ad of all. This is the greatest ad of all time that I've run. It's just called Advice For Consultants Struggling To Get Clients From A Consultant Who Has Too Many. This one's made more than $3 million, and it's run for more than three and a half years. This one crushes it. It's because it just provides really good advice. It's like a Trojan horse ad where I give people advice by saying, you know, generalism, you shouldn't be a generalist. You need a predictable way to get clients. Generate appointments, sorry. And you need a predictable way to convert appointments into clients. It just adds value.
That is the Trojan Horse Facebook Ad Examples swipe file. It's available for download in the resources beneath this video. The best use of the Trojan horse strategy is the Reflections one by far. This one is a killer ad. It still runs really well. It gets 10 out of 10 on the quality score, even after like years, and it just still crushes it. It's a beast. I also have some other examples of some Trojan ads. I've got them here. Put these guys up.
So, this one here, this is real Trojan. What we did is we actually didn't even put a headline. We didn't even run a normal ad. As you can see here on this ad, there's no headline. If I was to look at a normal ad, it would look like that, right? It would have a headline and then the image would be like rectangle, landscape. That's what an ad looks like. If we're using a Trojan strategy, a true Trojan strategy, then we shouldn't try to look like that. Instead, we didn't. No headline, nothing. Then if you click the image on an ad, you usually go to the landing page. If you click this image, it just opens up the damn image, which you might think is kind of counterintuitive. Like, why would Sam do that? Does he just want people to click the image and not have a headline? Well, it works. Not all the time, but it's something you should test, because it's a Trojan horse ad. It looks like a post from a friend. This is coming from Sam Ovens, image of me, and then a photo.
Towards the end, there's a link to the actual ... That's our call to action. This is how you can make it an ad. It's basically a post, and then we include the link deep in the text, it's buried, and then we use no headline or anything. This gets under the radar. It gets accepted. It's a true Trojan strategy. You can do this by just creating, first of all, an organic post. You can go to your fan page and you create an organic post, and towards the end of it, you put the link in, and you make sure that you delete the link preview if it pops up.
Now, this might be hard for you guys to understand, so let me just give you an actual example. I have to go to my actual ... I'll just do this on my personal page. If I go to paste in this link here, it just pops up here. See how it's pulling in this link preview? This is called a link preview. You just get rid of that. And then you get rid of this here, too, get rid of that. Then you just add a photo or a video. What you're trying to do is have an image post, but then you can have the link within the image. Sorry, you can have the link deep within the ad text.
That way, you do it organically first on your Facebook fan page, and then it looks like a normal post but it's got a link in it to the landing page, and then what you can do is you can use that ad, you can use that post ID as an ad. Or, you can boost that post and then you can find it in your ads manager, the boosted post, and you can copy that, and you can actually put it into another one of your campaigns, so that it's not just a boosted post. You can then just use that for targeting whatever you want. That's a really good strategy.
Here's another Trojan horse ad. This one is about as Trojan as it gets. There's a video of a black panther and then me, and it's like the most random thing in the world that you've ever seen. This doesn't look like an ad, and it doesn't even have a damn headline, again. This one's kind of pushing it, to be honest. It really pushes it a bit, but it's kind of relevant, because I actually have a way to talk about black panthers in this ad and make sense of it. Remember, you want it to be out of place, but relevant. This black panther's certainly out of place, but it doesn't have relevance, and so it works. It works really well. It's a nice Trojan strategy.
Then here's another one. This one was just a photo of me and my parents. Look at me there. I just look like just a normal person eating a burger and some chips. I do not look like a successful businessman or somebody that's going to say that they're holding a webinar. Really, this photo wasn't taken for the purposes of being in an ad. I was actually just hanging out with my parents eating a burger, and then someone took a photo. Then later on, someone else got that photo and used it in an ad and it worked. This is what I'm talking about.
You can find these sorts of ads and they work really well, because this is the sort of stuff people see on Facebook and they like clicking on, because it's people eating and it's family and smiles and there's no headline. It's Trojan. And if you click on it, it just pops up the image, and then it starts off by talking about how to retire your parents to Saint Lucia. It doesn't start talking about how to buy 50 Lamborghinis and 10 helicopters and 19 Rolexes. It does, really, the exact opposite of that.
What's good about it is that it still gets to make the claim and everything, and it still gets to give a call to action link, but it just does not look like and it's not perceived and it's not received anything like an ad. This is the trick and the art to Trojan horse ads, to not look like an ad. The more it doesn't look like an ad, the better. But remember, it still has to have some soldiers in it. If the Trojans just created the Trojan horse and made a big, nice horse for the enemy to accept, then they would have accepted it, put it inside their gates, and nothing would have happened.
So remember that they had to load soldiers inside it. The same way they had to put soldiers in their horse, you have to put soldiers in your ad. And your soldiers is your call to action and your pitch. It has to be there, and it has to be strong. You don't just want to have a horse with nothing in it. Remember, it's about achieving that balance. Keeping people happy on the platform, but also being aggressive enough so that you get people to take action and you get customers. Trying to strike that balance.
Now, when you do this, here's how you influence the system. When you have engaging ads, so let's say you input your ad and Facebook's algorithm runs its processes, it chooses who to give it delivery to, and then people see it, like it, click on it, and what not. That's the output. Then that's going to come back round through the feedback mechanism and change the ad and its future. When the feedback comes through the Facebook algorithm, and you're using a really engaging ad, let's say you're using a Trojan horse style strategy with a really good ad. Your quality score is going to look like nine or greater.
Now, in order to have an ad that works, you really need to be a six or above. If you're below six, your ad ain't relevant and you need to make it relevant. At six, it can start to work, but if you want something that's really going to work, it needs to be nine or above. All of those ads I showed you were 10. Reflections has been 10 for forever, three years. We've never seen it at nine. That just gives you an idea of the sort of relevance quality score you need to be at to get things good. You know you've got a good ad when it's a 10, and you know you've got a good ad when your CTR (Link), and I mean CTR (Link), not the normal CTR, is greater than 1%.
And you know your ad is good when it's got balanced 50/50 comments. Like, 50% of the people are for it, 50% are against it, and now they're starting to fight each other and attack each other and argue with each other, and you don't have to step in and do anything. You just let it happen. Best to stay out of it, best to let them do it, and by them doing it, they're helping your ad, which is all we really care about, and that's you it works.
When the Facebook algorithm sees a quality score, a relevance score, of nine or 10, and when the algorithm sees a CTR of greater than one, and when it sees the sentiment of the comments, that they're both for and against and balanced 50/50, it's like, oh, this is good. Let's give this thing an advantage. So, if you can signal these things to the Facebook algorithm, it's going to then have a bias towards showing your ad to more people, and it's going to show your ad to more people. It's going to give you more reach. It's going to give you a higher quality of traffic. It's going to give it to you at a better price, and it's also going to give you advantage and preference over your competition, which is how you win. This is how we play to the algorithm's biases. By understanding how it thinks and then playing to that. And engaging ads is at the heart of this.
Now let's talk about something called lexical sentiment bias. So, precognition. Facebook's algorithm knows the content people like and don't like without people seeing it and providing feedback. It has records of all liked and hated posts and the text contained in them. Most people don't understand this. Facebook's got saved, perfectly, in its databases and warehouses and things, every single post that's ever existed on Facebook ever, Even if it was deleted, even if it was moved, they've got record of it.
Within those posts are words, and it can pattern match things. It can look at, hey, this post was really hated and caused a big ruckus, and it contained these words. And, hey look, this post was really loved and it went great, and it contained these words. So, if you create a post that has word patterns that the negative one had, it's not good news, and if you create a post that has word patterns that the positive ones used, it is good news.
This is how complex this thing is, but it's fascinating to understand. Facebook's algorithm doesn't have to wait until humans click and give feedback and comment or whatever. It can forecast how it's going to go before a human's seen it just based on the words within it. There's things that we can do to make sure that the words within our ad are things that Facebook likes, and we don't include words that Facebook doesn't like so that we can get a further edge everywhere.
How Facebook's algorithm does this is with something called a lexical sentiment analysis. Lexical sentiment is the view or attitude a human will likely take based off a combination of words. So if I'm a human and I read a bunch of words, then I'm going to have a feeling or I'm going to have an image in my mind that creates a feeling and a reaction and a response, and lexical sentiment is understanding what combinations of words create what sorts of responses from a human being. That's how it works, and I'll show you examples of this soon.
Then there's trigger words. Trigger words are words perceived as bad, such as "make money," or anything anti-religion or religion oriented at all, because that's kind of a trigger point for Facebook. You can cause a big ruckus if you talk about religion. And then hate speech and racism. All of these things. If you go and put a post like that on the internet, that will cause a ruckus, and Facebook has ... They're watching out for those things in a big, big way. Obviously, none of us are going to create ads that do that, but sometimes we can accidentally trigger the wrong thing just by saying particular words. Remember, the algorithm isn't perfect. It can't understand if something is actually bad or not, like in the ultimate reality. It just uses heuristics to make roundabout assumptions and guesses and judgments that are true most of the time, but obviously they get things wrong sometimes.
These trigger words, you want to avoid using them at all cost. If you do use them, Facebook will likely disapprove the ad, ban you, and your time on Facebook will be short lived. But then, Facebook's algorithm, what's fascinating about it is it has a happiness bias. Facebook's algorithm can tell if a post or an ad is negative or positive based on its words, based on the words within it. It can read those and be like, oh, this is happy, or oh, this is negative.
The algorithm has a bias towards positive content, because it wants its users happy, not sad. This is a no brainer. Facebook, what do they want? They want users on their platform and they want them engaged. Now, if you want users to be on your platform and happy and engaged, obviously you want them to be happy and not sad. So obviously, Facebook's algorithm has a bias towards serving positive content over negative. It's going to serve more happy content over sad content, and if something's super happy, it's going to really serve that to a lot of people. If something's negative, it's not going to serve that. Or if it does, it's going to serve it in low doses, and then it's going to stop serving it completely.
How do we make our ads more happy? Hacking happiness. You want to take your winning ad angles and copy paste the text into a lexical sentiment analysis tool to see its score. Then what you want to do is change words to increase happiness without changing the core message. Now, this will kind of blow your mind. It's pretty cool. I will show you how it works. I've included a lexical sentiment tool for you, and the link is on the screen. Consulting.com/lexical-tool. You want to use this free tool to analyze the sentiment of your ad angles by copy pasting their text into here. Let's take a look at this.
This is what this tool looks like. Pretty simple. You can enter words here, select your language, and then click analyze. If I put in the word, say, "love," and then analyze it, it's going to say ... Or if I put in ... It's fascinating to see what combinations. "Love" actually comes up neutral, but if I put in "laugh," positive. It's quite fascinating to see how this tool judges whether something is happy or sad based on the words or the combination of words. Then if we put in something like hate, obviously, it's going to be negative.
We don't really use this for individual words. I put in individual words so I could how you how it works. What we do is instead we put in combinations of words. So, let's grab one of these angles here. This is probably one of the best ads we've ever run of all time, and it's called Reflections On Reality. Let's grab the body copy here. Let's grab all of it and see how it works. Now, hold on. This is probably going to grab that copy, too, which is going to be annoying. That's going to be very annoying. So, instead, let me just find it for here. This'll allow me to copy it without grabbing all of that. I'm just grabbing all the body copy here, all the way to the end, and I go copy, and then I paste it.
A ton of words, and it analyzes as positive. So a 0.6 pos, neg, 0.4. This probably plays in, in a huge way, into this ad angle and its success. You're really wanting to get an ad angle that's positive, and you don't want to have an ad angle that's negative. If your ad angle is neutral, that's all right, but you want to try and make it positive. They will do better when they're positive. One, because humans like being happy, and also, the algorithm likes it when humans are happy, and if you make humans happy and you make the algorithm happy, then you really get the whole machine to start dancing.
Now, it's important not to just write a really happy, feel good ad with no intention or no call to action. You still have to have a call to action, and you still have to be aggressive-
... a call to action and you still have to be aggressive enough in order to make your point and get people motivated enough to click and take action, but you can still leave people feeling havoc. Now, if you get an ad, let's say we put an ad in here and it's coming up negative, then we want to start going through in here and looking at words that aren't really that positive. We want to start changing them. We want to start removing negative words and we want to start rewording things to be more positive, rewording negative sentences and we want to try and put more positivity into things. We want to adapt it. Now, the best place to start with this is to find your already winning proven ads 'cause we know those work. We know they are your best ads. We want to grab the copy from that, paste it in here and then make it better. Then take that back to Facebook, create a new ad and compete them head to head. Compete your already proven winning ad versus the sentiment modified winning ad and watch the difference. That's how it works.
It's a very cool tool and Facebook's algorithm absolutely is doing this and absolutely has a bias towards these things. It's actually very simple technology that really fuels this. This has been around for a long time. It's not [inaudible 01:37:47] stuff, but it goes deeper than that. I've also created a list of lexical, happy and sad words. The hundred most happy and the hundred most sad words in the English language in order. You want to avoid the negative words and then you want to use the positive one. Let's take a look. This is available for download beneath this video in the resources section. It's called lexical, happy and sad words. These are the hundred most negative sad words in order with terrorist being the absolute #1. This is the most negative word you can possibly use followed by all these other words. Don't really encourage you to read through this list here because it's pretty bad. I am showing you this so you do not include these things.
Some of these words you can see how you can use without really being aware of what you're doing like robbed or hated or painful or hurt or depressed or cried, severing. Some of these words you could use them in an innocent way without really thinking about it, but you could really hurt your ad by doing so. You want to try to avoid all of these 'cause they are the most strong signals of negativity and then here's the one hundred most positive heavy words in order one with #1 being laughter more than happiness which is fascinating. More than love. You want to try using these words. Now you're not trying to key word stuff your ad and making it stupid. Remember I said you wanted to start with an already proven winning ad and just slightly modify it. Just see if you can fit in a few more positive words.
See if you can remove some negative words and improve its sentiment score in the total I gave you which I've linked to up here as well so that you can copy, paste the right angle in. Make sure it doesn't have any of these. Make sure you can put in some of these. Bump up that sentiment and then I've also included things that trigger Facebook's red flag and don't use those words related to these things. When you include words that relate to certain things, Facebook flags you and red flags your account and it will either just ban your ad, ban your entire ad account or it will just flag your account and throttle your delivery of impressions and reach and things. Now I don't have the exact list of all the actual words, but what I do have is a list of things that it really doesn't like. Don't promote MLM or get rich quick or make money from home, et cetera and don't promote weapons or stuff related to weapons. Don't break the law. Don't be offensive. No spam. Don't promote illegal things. Don't promote adult products.
You can use your common sense about words that relate to any of these things. 'Cause words that are going to relate to any of these things are going to be trigger words. Trigger words, forget about your sentiment analysis just being positive or negative, trigger words will just get your ad flat out, it won't get approved. Even if it does get approved, it'll just get your ad eventually disapproved and it might cause your whole ad account to come down. Avoid the trigger words and increase your sentiment and positivity with more positive, less negative. You got the tool and the resources to do it. When you improve the sentiment of your ads, you're going to get... When you input your ad into Facebook, it's going to run a quick process on it and it's going to determine the sentiment. Very quickly it's going to be like, "Is this positive or negative? It's going to be positive." It's good. If it's negative, it's not good. If it's neutral, it's okay, but obviously positive has a big advantage, but not just if it's positive or negative. There's a scale.
The more positive it is the better. If something is more positive than something else that's still positive, Facebook's algorithm will give preference to the thing that's more positive and it's also scanning for trigger words. Does it have trigger words? If it does, it's going to decline. It's going to disapprove your ad before it's even had a chance. If it doesn't it's going to pass it and it's going to let it go. All right? You want to make sure the algorithm is receiving these signals so that you can have a preference and an advantage when advertising. Now let's talk about something which is spooky and if the lexical didn't blow your mind, this thing's going to blow your mind, graphical sentiment bias. Most people don't know you can do this thing or most people have absolutely any idea this thing is even doing this thing and is capable of this thing. Let's see what it does.
Precognition. Facebook's algorithm notices the photos people like and don't like without people seeing them and providing feedback. It has records of all liked and hated posts and the photos contained in them from all time. Even ones that have been deleted. Everything that's ever been put onto Facebook it has warehoused and tagged. Unlike words, words are easier because Facebook's algorithm can just look at the words and it can find, "These words were here. These words are there. It's probably like this," but with images it's harder because images don't contain words. What do they contain? They contain pixels and objects and things. Facebook's algorithm can actually run image recognition and it can find objects within images and if it runs a nearest neighbor sort of algorithm through its database of all posts and images of all time, it's looking for images that look like this image or images that containing objects that this image contains and with those images in the posts containing those images were positive and well received by the audience or negative and hated by the audience.
A few images contain objects like the positive ones, then you're good and you've got an advantage and if your images that were like the negative ones, you have a disadvantage. That's how it works. Most people have no idea that the algorithm can look at the images. It absolutely can. It's been doing it forever. How do you think when you upload a photo it can detect somebody's face? That's the start. It's pretty advanced. It can detect a face. If you upload an image, it'll find the people in it and it'll tag them and also if you upload an image with text on it. Facebook knows when there's an image with text on it. It has the 20% text rule. It's obvious that Facebook's algorithm is scanning images, but what it can see is the question. That's what I'm going to show you in here. Not only what it can see, but what it likes. What things that it can see in images does it like? How can we give it those things? Image recognition... Facebook's algorithms scan images and detects the objects, scenes and emotions in them.
Graphical sentiment is the view or attitude that a human being will likely take based on a graphical combination. Based on a bunch of pixels and a human looking at those pixels, will the human respond positively or negatively? If positive, then the sentiment is positive. If negative, then the sentiment is negative. This is how it works. That is the graphical sentiment. Really it can detect so much more than that. Not just as a negative or a positive, it can detect a huge amount of things. Here's one of the ads that I'm running. When Facebook's algorithm really assesses this image, it can detect faces like my face and my wife Ashley's face and it can also detect that there's humans. First of all it might detect that there's two people in here. Then it might detect the sex. Well, it doesn't might, it actually does detect sex. One male, one female, two humans and then it can detect the rough age range with pretty good accuracy, 25-34. It can tell if there's kissing and it can tell if it's outside or inside. It can tell what a beach is, so here's a beach.
It can tell what Nature is. Nature is a category containing things within nature, so if one of these is true, nature is true. Then it can form a sentiment from the overall combination of images and the sentiment could be happy, neutral or sad which is positive, negative or neutral. There are things that Facebook's algorithms likes. It likes happy sentiments, but it also likes certain objects, which is fascinating. Facebook's algorithms has biases towards objects and images. We want to play to those biases, but I want to back up what I'm saying 'cause I'm sure some of you are like, "Does it really do this? Does it really know how to call these images? Sam, how do I know that it actually does this? Can you show me?" Well, if you don't believe me, watch this. I have on my screen here, my mobile phone. I've hooked it up to my computer and if I show you on my phone here, I will be able to show you how this thing works. I'm on my Iphone right now and I'm using Google Chrome which is an app for, it's a browser.
Google Chrome on an Iphone. I'm not using the Facebook app. I'm just going to the Facebook website and logged in. Then you can look at any person on Facebook. I'm just going to use my Mom here, Ann Evans and then you can go to an image, like say this image here. Once you've found an image you can just hold down your thumb on that image and then you want to just hold down and then it should say Copy or Look Up. Then you want to select Copy, all right? Now if I just go to a message, any sort of text. I could go to a notes or a new message and now I just want to paste what I copied. I don't think Facebook's aware of this, but I found it. This is what it's reading in there. This is what the algorithm is doing. It says "Image may contain three people." It can detect the number of people and it's detected one of the person's names, my sister Isabella Evans.
It can detect that people are smiling and it can detect that people are standing instead of sitting and it can detect that somebody is wearing a suit. It can detect that it is outdoors. How about that? That was the one that I just showed you. If I go back to Chrome, that was that. It could detect my Dad's suit, three people, one of their names, outdoors, smiling. Let's try another one. Let's try this. Copy. This one's found two people, smiling, standing and hat. It's funny what it can detect, right? It's detected a hat and smiling and standing, but it's detected two people and this is the thing. There was obviously only one person in there, but machines are funny. Let's go back and look at it again. You see there's somebody's face in the background there. Even though out here in mine when we looked at it that fast, I don't know if you did, but I didn't really see that face the first time I looked at it.
I just saw one and that's why when it first pasted it and said that there was two I was like, "What? Did it make a mistake?" but there's actually two in there. You can see how accurate this thing is. It even detected my sister's cap, that hat. What's fascinating about this is that if you've ever noticed, graduation photos seem to get really, really a lot of likes and comments. Now in order for them to get a lot of likes and comments it has to get a lot of reach. A lot of people have to see it. How does it know? It's algorithm has to detect certain things. It can detect things like hats and if someone's wearing one of these hats, it might help it to get a lot more reach. This is how advanced this thing is. What I showed you is just an actual way to prove it. This isn't everything that the algorithm is seeing in here. This is just the things that I can prove to you that it's seeing in here by this example and that hack that I found.
You can do that on your own phone whenever you want. Now that we know that Facebook's actually doing this we want to play to its biases. I already explained these things. Just like with lexical sentiment, these trigger objects, not trigger words, but trigger objects and then there's happy objects and then there's positive objects, negative objects and neutral objects. Trigger objects are objects perceived as bad such as violence, gore, text overlay, sex, nudity, et cetera. You want to avoid using these objects at all cost. Facebook will most likely disapprove the ad, ban you and your time on it will be short lived. Don't include things like blood, violence, gore, sex, nudity... It all detects these things. Facebook's algorithm is very good at detecting nudity and pornography. Do you really think humans at Facebook have the time to check every single image? No way. They've got algorithms that do a lot of this and so do most platforms. It's not just Facebook. It can detect all sorts of stuff. Don't trigger any of these things.
Now, happiness bias. Facebook's algorithm can tell if a post or an ad is negative or positive based on its image and the algorithm has bias towards positive objects, scenes and emotions because it wants its users to be happy. Now hacking happiness. You want to take your winning ad images and load them into an image recognition or analysis tool to see its objects and emotional sentiment. Then you want to test your images with more positive objects, scenes and emotions. How do we do this? I've got two tools to show you. The first one is an image sentiment analysis tool. I'm sorry, both of these are image sentiment analysis tools. You can use these free tools to analyze the objects, scenes and emotional sentiment in images. Object and scene analysis is the first one and it's consulting.com/objecttool. Let's check this one out first. This is called Amazon Recognition and honestly you don't even really need to try this out. You can try it out if you want. You can just get a free AWS account. You can just create an account. It's free. You don't need to pay them any money and then you can just try a demo.
Then when you come here you can see that you look at the sample image and from the sample image it's detected that there's a skateboard, sport, human, people, all of these different things. What we want to do is we want to upload one of our photos here. Let's say I'll grab this one. Here's it's detected human and people. Person, face, portrait, female, so it's getting sex, leisure activities. This is an interesting one. Smile, apparel, clothing, head, [inaudible 01:56:31], underwear, woman, dress. You can see that it can detect all of these different things from here. These are the different objects that it might find in an image, but there's another one that we can use which is called emotional sentiment analysis. This one's consulting.com/emotiontool. This one you don't even need to log in at all. You can just go to this and then just try this demo. Here you can use the sample image or you can upload one. If I upload an image this time I will use, I've just got to enter in this and I'll use this image.
You can see as soon as it's loaded in, it can detect the two faces. Mine and Ashley's, my wife. Now if I click on mine, my response here is happiness. There's happiness here, but there's also fear which is actually interesting which can be caused sometimes by the glasses, but if you look at me there it doesn't really look fear, but this is fascinating to see what it's seeing here. Then this one here is happiness, but it's actually reading disgust and it's probably because of the sunglasses here. That's why it's fascinating to see what a machine sees sometimes. That's why sunglasses can really play with it, but let's try a different one like this. This one's detected Ashley for happiness, but it's quite blurry. Let me try another one, this one here. This one's got me neutral and it's got Ashley neutral, too, but you know what? These images are actually really blurry because I can see it up here because I've tested this one before and it gives me neutral and it gives Ashley happy. Let me try one more.
This one here, let's see what this finds. It's not going to find a face in there. It's also fascinating to check if its detecting a face. Now this one here isn't Facebook's algorithm and what it can see. This is just another tool that can detect things, but it's important to load in some of your proven images and see what the algorithms can see in here. You can see from this image here it's detected surprise and in this one its detected happiness. This one's happiness. This one here it's detecting neutral and really what you want to do you want to make sure that yours, it's detecting a face first and foremost because Facebook... What do people like to see on Facebook? People. What do people have? Faces. It should detect faces and in those faces it helps if it can detect a positive emotion. Happiness, surprise, something like that. Also, you can see the different images, the objects that it can detect in here. It should detect human people, person, face. It should detect these things, smile as a positive. That's how it works.
Now with these tools alone, we know we want people to be in them. We know we want them to read happy or neutral at the least, not negative, but how can we really take them to the next level? How can we really play to the extremes? Because remember algorithms like extremes. That's why I created a list of graphical happy and sad objects. These are objects, scenes and emotional sentiments considered to be happy and sad. You want to avoid the negative ones and use the positive ones. Let's take a look at this. This is when we stand to get pretty [inaudible 02:01:09]. This is a resource, a PDF ready for download beneath this video in the resources section, graphical heavy and sad objects and introduction to Facebook's algorithm performs graphical analysis on the images in your ads and it has a bias towards positivity. This means that positive ads gets priority, more reach, impressions, lower prices, et cetera. I've given you the link to those tools again and here's the positive objects, scenes and emotions that the algorithm can detect and likes. You want to try using these.
The first one is a big one and it's humans. It loves humans. More is better than one. An optimal is one male and one female. Why this is is because it's quite, it's instinct. It's in the human species instinct. If we weren't programmed to like one male, one female, them we wouldn't be alive because this is the combination it requires for life to exist. It like that. Also, when there's a male and a female, females like images of females and males like images of males and that's why most female's heroes are other females and most male's heroes are the males. When you have a male and female together, you get both sides. This is quite fascinating. That's why a lot of the ad images you'll see now are me and my wife. Now faces, it loves faces. If you've got humans in it, make sure that there's faces in it of those humans. It likes smiles and other positive facial expressions and gestures. It just doesn't detect smiles and things, it can also detect hand signals. If someone's waving, thumbs up, whatever and it also likes eyes open and gazing at viewer.
Humans attracted by other humans' eyes and the gaze of the eyes. When you look, your eyes are really powerful at pulling other people's eyes into them. Your eyes should be open and you should be staring at the other person on the other side 'cause when they catch that gaze they're more likely to look back because we're programmed to do that. Then another one is babies. Baby images goes nuts on Facebook and so these things work really well, but some of you may be thinking, "Sam, why don't you rent a baby and just use it in your ads?" Remember what I told you about telling the truth. Don't let any one of these things... All of these things are interconnected. Remember engaging ads tell the truth. If we're just trying to hack engagement by including a baby, but we don't actually have a baby, then we're going positive on one but we're going negative on another and it's just not worth it.
One of the most important things is to tell the truth. I'll try that when I have one or I guess I could get a photo of one when I'm hanging out with my friends or something. That could be a way to get one, but that's why I haven't tested one so far. It also likes animals and pets. People love pet photos on Facebook. It loves nature, so it likes trees. Outside it likes lakes, alpines, mountains and it likes the sky and all these things. Sunsets. Sky's a huge one. It really loves the sky, especially when the sky is prominent. It also likes sunsets. Daytime is preferable over nighttime and outside is preferable over inside. If you had an image inside and then one outside, the outside one would work better. I've already said sunset, clear skies, sunny and good weather helps. It can detect the ocean. It really likes the ocean and mountains and trees and grass and it likes human standing, interestingly enough.
Standing is better than sitting. It likes humans standing up. And colors? It likes colors, bright and colorful works and it also likes talls and objects that are fun. Things like bikes, scooters, boats, et cetera. Leisure activities. These are things that it likes. Now, you might think why would it like these things? Well, humans like these things and it likes what humans likes 'cause it's been programmed that way. That's why it likes these things. When you think about it it's kind of common sense, but most people don't think about putting these things in their ads. You don't want to put these things in so that it looks awkward, but you want them to naturally, you want to just naturally get photos in these sorts of spots. Nature and outside with humans and faces and things like this. Here's negative objects, scenes and emotions and you want to try avoid using these.
No humans. No humans is pretty depressing for a human and so why would you put no humans in your advertisement? It also doesn't like one human sitting alone. One human standing, but smiling might be good, but one human sitting alone it doesn't really like. Just think about it like that. And then no faces. If there's a human, but that one human is looking away, it doesn't like that so it doesn't like humans being near, but the faces are looking away. Also negative facial expressions and gestures. No natural light, no outdoors or nature and no color or lack of color or brightness. It also doesn't like violence, blood, gore, war, weapons, gambling, drugs, smoking and it also doesn't like text overlay or text included in the frame of the image. Even if you've got text showing in the image in the background or something, it just doesn't like text anywhere in there.
These are things to include to make your ads work better. Now I've got one more resource which is Top Performing Ad Images and these are some of my all time best performing Facebook ad images that play to the biases of the algorithm. Let's-
... Play to the biases of the algorithm. So, let's check these guys out right now. I actually gave it to you in a cheat sheet. It's called Top Performing Facebook Ad Images, and this is what it looks like, and it's available for download in the resources section beneath this video. This one works quite well. You can see, it's unique. First of all, it's outside. It can detect alpine, it can detect aircraft, and it can also detect two people smiling, male, female, and dress. This one works well.
This one also works well, and it's got sky in it. This one works well, because it's got sky. It likes sky. It also can detect two people, male, female, and it can detect pet, animal, and smiles, happy. Sky, human, male, female, cat, happy. Then this one, sky and ocean and two male, females, and smiling, happy, outside. Again, male, female, outside, sky, alpine, snow, nature, smiles, happy.
Then, here, two people, male, female, trees, sky, outside, light, happy, daytime, dress, smiles. Then there's this one here. It detects sky, it detects animals. One person and that one. That one does quite well. It's because it's got animals, and it also just kind of looks weird. This one does well. It's outside, one person, smiling face, and then it can detect trees and yeah, you can see that. This one also does well. It's got one person, animal, sky, outside.
These are all good ad images. You can start to see that none of these were posed for a Facebook ad image. This was actually my wedding. This was just a photo my wife wanted to take on a Sunday one day just randomly. This was when I basically first met my wife ages ago and we just went on this trip. Then this was just when I was hanging out in Saint Lucia. This was actually just us hanging out in Santa Monica, and this was just me hanging out with my cat having an ice block. This was when I was actually just looking at some stuff on my phone and my wife took it. This was when I was just working and then my wife just took a photo.
You can see, none of these are really stages, and this is what actually happens in my life day to day. This is what it normally looks like. They've got the truth element and they've got outdoors, nature, animals, friendly, male, female, face, human. You can start to see why these things probably work. What's fascinating is that none of those things really make me look rich. None of those are me shaking hands with Tony Robbins. None of those things are me with 50 Ferraris in the background or with 10 Rolexes on my arms. They just make me look like a normal person.
This is what I mean by Trojan horse. You're looking like a friend's post, and at the same time, you're using humans' faces, male, female combinations, you're using outside, nature, sky. You're using things to get the algorithm to favor you and at the same time, have a good image so that human beings find it interesting and like it, too. You're playing to both things. When you do that, you get an edge.
This is what happens when you use good images. A graphic sentiment might read, sentiment equals positive in the image. Scale of positivity in the image might 0.9 out of one, and then genders might read male and female. And then age range might read 25 to 35. Trigger objects equals pass. So, trigger objects might be looking for text overlay violence, gore, blood, guns, negative hand signals, things like this. This one, you want to make sure you pass it. Male and female, I'm telling you, is the best balance to have if you're trying to reach both males and females. If you're just trying to reach females, then having just female subjects is great. If you're just trying to reach males, then just male subjects is great. Positive sentiment helps a lot. The age range, you want to have the age range of the audience you're trying to target, too.
Obviously, I sell to people from like 18 all the way to like 65 plus, and there's only really ever images of me in it and my wife, and we're both in our 20s, so it still works people outside of that age range, but you'll find that the algorithm seems to most latch in on this range. Now that's for many things, not just the images, but the images are certainly a large part of it.
Now, let's talk about the next bias, which is a financial bias. Facebook loves money. It needs money to stay alive and grow. It also needs money to satisfy its investors. Facebook's algorithm has a bias for making money, but it will never take money at the cost of user experience. People don't understand this. People think, oh, Facebook just wants money, so why won't it take my money? It won't. Its user experience is more valuable to it than the money. One comes before the other, and money signals.
So, the algorithm sees money as individual ad set budgets, cumulative ad set budgets within a campaign, cumulative campaign budgets within an ad account, bid prices, average daily spend, average monthly spend, average annual spend, lifetime spend, reliability of payment, which is how many declines has it had. These things it sees as money signals. Then it has a money heuristic. If two advertisers are competing for the same traffic and all things are equal, Facebook's algorithm will go with the advertiser that has the best money signals every time. They get the traffic, they win.
To get an edge with your ads, you want to pay to the algorithm's financial bias by signaling the right things. You want to look like money. Here's how we signal money to the machine. Ad accounts. You should not have spend limits on your ad accounts. You should do your budgets and control your spend at the ad set level within campaigns. Don't set a spend limit on your ad account. It's not a good signal and it's not necessary anyway.
Payment method should be a company, not a person. Make sure you use the Facebook business manager. Make sure that your payment method on file, which will be a credit card or a bank account, make sure that credit card is attached to a company. Do not put a personal credit card on that thing. It can tell the difference. It knows that people don't spend a lot of money. It knows that people that spend a lot of money have businesses. Make sure you do that. And then use a credit card or a bank link, direct bank link, not PayPal. It doesn't like PayPal. Yes, you can use it. It does not like it very much. If someone's using PayPal, someone's using a card, it'll choose the one with the card.
If you're using a card, don't use gift cards or temporary cards. Use a proper card, a business card. And make sure you never decline and always pay without any issue. If every time the Facebook machine tries to bill you and it always declines, just like anybody, that's going to not trust you. It's not going to like you, so make sure it doesn't happen. We purposefully use our bank links, so no cards at all, because cards can decline. Even if there is money, cards can decline because of fraud protection, so we use a bank link so it never fails.
Campaigns should optimize for conversions. They should not have lifetime budgets. Like I said, set budget at ad set level. They should have a good history of consistent spend, and you should be spending more than $100 a day. $100 a day is going to start to get you good momentum. Less than that, it works, but at $100 a day, it starts to work quite a lot better, and it starts to work really well at $1,000, too. $1,000 a day seems to be where it's highest, like where it's most efficient, sorry. Above that, it still works great, but at $1,000 a day, it's really efficient. That doesn't mean we don't go past $1,000, obviously. We want to spend as much as we can, but it really likes its efficiency around there.
Past $1,000 a day, campaigns have entropy and should be decentralized and load balanced across multiple campaigns separated by ad angle, and I've explained this in previous modules in this week's training. You want to have good architecture. Entropy exists at a lower scale, so entropy exists at the campaign level beneath $100 a day. You should try to have your campaigns more than $100 a day. But they also have entropy at over $1,000 a day, so if you're spending more than $1,000 a day in a campaign, split it out. Load balance it across two campaigns or three campaigns. It's better to have 10 campaigns spending $1,000 a day each than one campaign spending $10,000 a day. It likes to be at that efficiency zone, and that's what good architecture is all about.
Ad sets. Ad sets should have a daily budget of at least two times KPI to function. Now, KPI is the thing it's optimizing for, so if it's optimizing for leads, if the conversion event is set as leads, and you're getting leads for $10, then two times KPI would be $20. Two times $10 equals 20. That's where it's going to really start to function its best. It will actually start working at one times KPI, but not very well. Two times KPI is when it's actually going to start working properly, and then it's going to be most efficient at 10 to 15 times KPI. So, 10 to 15 times the cost per lead, if you're optimizing for that, which would be $100 to $150. That's when it really starts to sing.
You're going to experience entropy at the ad set level at less than two times KPI and greater than 20 times KPI. So, beware of these zones. Stay out of these. Stay within these if you want maximum efficiency from your ad sets. And stay above $100 and beneath $1,000 if you want maximum efficiency here. And never change the budget or settings of an ad set. If you want to make changes or budget changes, duplicate it, and make the changes in the new clones. Don't disrupt the initial conditions of the existing ad set. It's not a good idea unless you absolutely have to. Like if you have to turn something off, you can turn it off, but it's best not to make any changes in existing stuff.
Now, bids. For the most part, bids should be auto for cold traffic, so you're setting those to auto warm retargeting and anything to do with conversion objective campaigns, and for hot retargeting, you can use daily unique reach and bid manually at higher prices to win auctions. If you really want to signal money to the machine and get preference in auctions, then we want to really only do that to an extreme level when we're retargeting to hot audiences.
In our hot retargeting campaigns when we're advertising to people who are deep within our funnel and recent, as in within the last 30 days, then we know they're hot. We want to market to them. So we have to signal, we have to really do something to get our ads in front of them. To do that, we want to use the daily unique reach objective and we want to use manual bid. And we want to bid high. You can try like $150 a day per 1,000 impressions all the way up to $999, but that's pretty steep.
Most of the hot traffic or hot retargeting campaigns we run with the daily unique reach objective, we're going to be in between $100 and $300 per 1,000 impressions. That's our bid. We won't necessarily pay that, but we'll bid that just to climb the pecking order in the auction pool and make sure we're getting delivery.
And spend. Your spend should be as consistent as possible. Ideally, it should be every day without ever stopping. The algorithm favors consistent spend over time with no spikes or blips, and you want to be consistent, build momentum, and build a solid history. Just like building a history with a bank or somebody lending you money, you want to do this with the machine, with Facebook's algorithm. Make sure you don't decline, make sure you always pay on time, make sure you advertise consistently. Make sure that there's no dips and spikes and no question marks. Make sure that you look trustworthy and reliable. And the best way to look like you're trustworthy and reliable is to just be trustworthy and reliable, because that looks like that, because it is it.
When you do this, you signal these things to the machine. When you input a campaign and ad sets and ads and you push them live, Facebook's algorithm's going to look at you and scan it and look for a financial forecast read. It might read all the campaigns and all the ad sets within your account for today, its total threshold spend is about $1,000, right? So okay, this is how much this person's willing to spend. Now, their historical, how far away is their historical from what they're trying to do today?
If it's not too far away, it's like okay, this person probably not only is willing to spend that, but they're highly likely to spend that. Now, if you try to scale from like $100 to like $10,000 a day, it just won't even work. There will be a spend cap on your account at $5,000, and when you get to that level you just want to contact Facebook support and they'll take it off for you. You don't need a rep or anything like that to get it removed. But, you can't ramp spend too quickly, because it's going to look at your threshold spend versus your historical spend, and when things are too far apart, it looks sketchy and it doesn't like that.
And you want to make sure that you ramp things up gradually. You can move fast and pretty rapid, just not overnight. And it likes consistency over everything, so don't ramp things up and then take things down and do nothing, panic, wait a couple of weeks or a month, then come back and do it again. Keep it consistent. Advertise daily.
Then it will look at your bids. If you're bidding auto, then you're pretty good most of the time, because it's like, okay, they'll willing to just pay whatever and you're good. When you're doing manual bids, you need to be way better, because if you're not bidding enough, you won't get any traffic, and so yeah, that's why auto-bid's for most people unless you're doing hot retargeting in which you use manual and you go high.
And then risk. It's going to do a risk assessment on you based on your payment method. It should be credit card, should be business credit card, should be within the business manager. You shouldn't have a history of declining. If you have, make sure you don't decline again. You should also not have a PayPal account, and yeah, all of those things ... Or you should not have your PayPal linked to your Facebook. Use the credit card or a bank link.
If these things look good, then the algorithm is going to give you an edge in the auctions. It's going to give you more reach, it's going to give you better traffic at better prices, it's going to give you preference in everything. Another edge.
Now, let's at least about consistency momentum bias. Facebook loves consistency. It views it as reliability and trustworthiness, two things it cares about greatly. The algorithm has a bias for consistency, and it rewards consistent advertisers with reach and momentum. Now, this is fascinating, because it's kind of like the same way human beings build trust. Why do we trust certain people? Think back. It's because we've just been exposed to them over a long period of time with a lot of impressions. That's really what serves trust. It's familiarity and consistency and reliability, these things form trust. Same with the algorithm. It views things like that. When you're advertising consistently over time, you're really going to get a higher trust and it's going to favor you.
Now, consistency signals. What inputs tell the algorithm to perceive trust or distrust? The algorithm sees this as consistent use of ad accounts, fan pages, domains, landing page URLs, landing page content, ads, audiences, campaigns, conversion events, ad spend, etc., and the consistency heuristic. So, if two advertisers are competing for the same traffic and all things are equal, the algorithm will go with the advertiser with the best consistency signals every time. And they get traffic and they win. To get an edge with your ads, play to the algorithm's consistency bias by signaling the right things.
How do we signal consistency to the machine? Say your fan page. You want to use one fan page consistently. Don't use all these different things all the time, don't look like a spastic. Just use the same one consistently. Then you want to pair its relationship to an ad account, so you want to use one ad account and one fan page together, because that's what normal looks like to the algorithm, and it likes it, so do it.
And make sure that your fan page has more than 5,000 followers. If it doesn't, run a likes campaign to get more than 5,000. It has some sort of trust thing at 5,000 followers, so if you don't have that, get that. In the Facebook ads training, I show you how to get that. Just look back at one of the previous modules. You also want to post organic content on your fan page daily. Just share posts and things like that, and this shows that it's being used. And also build it up as an asset with photos, videos, and posts, etc., over time. That's what makes a fan page consistent and trustworthy.
And an ad account. You want to use on ad account and use it consistently. If you constantly use different ad accounts and they have relationships to the same fan page, it will decrease trust in the fan page and any new ad account that touches that fan page. It can be contaminated, so keep it clean.
Then verified domains. Sorry, verified domain, singular. Make sure you verify your domain name with Facebook, and then use that domain as the CTA link in all of your ads. This signals trust to Facebook, and it will give you preference in auctions. So, a lot of people don't know how you do this. Let me show you on the screen right now. If I go to my business manager, and I'm going to go to ... What was I doing here, my domain, got to go to my business manager settings. Then I want to go to brand safety down here on the left hand side, domains.
You can see here, I have two domains, samovens.com and consulting.com, and these are verified, verified. You want to click add domain, put in your domain name here, and then just follow the instructions and verify it. Then when you've verified it, you've increased your trust for this. Now Facebook, it knows that this domain is good. Then, you want to make sure that the links that you use in your ads are on this domain. Now, I have both in here, but we only run ads from this. We only run our ads from consulting.com. Stick to one domain, stay on one domain. It makes life a lot easier. The moment you use a different domain, it's a nightmare. Like, cookies don't travel across domains, and so you lose a bunch of tracking and stuff there. Also, Facebook doesn't like it when you use different domains, so use one main domain.
Then the link that you put in your ads for the call to action that goes to your landing page, that should be on this domain. Our domain that's verified is consulting.com. The landing page URL we use is consulting.com/webinar. Now, another thing you want to do is use, always use, https. Always use the secure https instead of http. This makes a difference. So, if you don't have a SSL certificate on your landing pages and your website and stuff, then you should get one. I recommend just using Cloudflare, it's the easiest way. It's free and it's simple. You can do it pretty much immediately.
Then make sure that your landing pages and your website can only work on https. Prime example. If I go to consulting.com without putting in the https or anything, it will go to https. It's impossible to go to a non-secure version of this. That's important. Why? Because Facebook's algorithm uses two pixels. One is a pixel for non-https sites. The other is a pixel for https sites. If you have some landing pages that are not https and some that are, it's going to cause mayhem. And, if you don't use https and you're just using http for everything, it's not going to work as well. There is more tracking options on https than http, and also Facebook has a bias towards https, so just use https, and verify the domain, use the same landing page URL.
Now, the landing page. Use one landing page and keep the URL the same, and avoid changing anything on your landing page, including split tests, and never use a link splitter or redirect traffic somewhere else, as this throttles your reach. If you've got lots of different funnels and lots of different landing pages, then maybe you should think about just having one funnel and one landing page into that funnel.
We're making more than $30 million a year, we spend more than $1 million on ads, and we have one landing page. I kid you not. We've had one landing page for more than four years now, and it still works. And we've driven I don't know how many millions of clicks to that thing and it still works. You want to just keep one consistent one, and then you want to keep that URL the same because every time you change URL, Facebook has to earn trust for that URL again. This is a fascinating thing to watch. If you just have your ads running and then if you change the content of your landing page just through ClickFunnels or any landing page tool, your ads stop working as well and your traffic gets throttled.
Why? You didn't change the link, you didn't update the link, you didn't do anything. How does Facebook know? Because it has crawlers that crawl your page. This is what classic scammers do. What they'll do is they'll create an innocent page and then they will create an ad that points to the innocent page through a link. Then they'll put the ad through and get it approved because it's all innocent, the landing page. Then once it's innocent and it's gotten approved, then the scammers will change the landing page to be the thing they actually want it to be changed to, which is their scam thing.
Now, Facebook has realized this, and whenever there's a change to the page, it freaks them out. Any change. A headline change, an image change, a video change. Any change, it can detect it via the HTML. So, any change in HTML, which is basically anything, then Facebook realizes there's a change, and then it throttles your traffic because it doesn't trust you anymore. Now, it will regain that trust and then serve traffic again, but it takes time for it to do that, and you don't want that to happen because it really hurts your ads.
If you absolutely must change your landing page, do it. If you must change the landing page URL, do it. But, avoid it at all costs. Avoid running split tests. Just try to find a page that works and stick to it and leave everything consistent. When you run consistent things over time, Facebook really starts to trust it and it just starts to build momentum and it just starts unleashing its powers at you. Because it knows, hey, this guy's willing to spend a lot of money, and this guy also is consistent, and we know we can trust these things, so let's send this guy lots of traffic, because we know he wants to pay lots of money, and we know he's good to show to our audience. But when you change stuff around, Facebook ain't going to gamble on you, so there's a huge bias in this.
Another one is never use a link splitter. Don't redirect traffic or route people through eight different links. Just use a simple link. Don't redirect it through all of the stuff, it doesn't like that. Conversion events. You want to use one pixel, crucial. You don't need more than one pixel. Some people tell me, "Oh, but Sam, what if I've got two products, shouldn't I create two pixels?" I'm like, "No." "Oh, Sam, what if we have two different websites, shouldn't we use two pixels?" "No." "Sam, we advertise in two countries, one's Australia and one's America. Should we use two pixels?" "No."
We advertise in more than 72 different countries. We spend more than a million a month. We have been advertising for seven years. We've made more than $10 million with advertising. We use one pixel. We continue to use one pixel. You should only ever use one pixel. Anything more is unnecessary and confusing and you're going to screw something up, so don't. And you want to keep your conversion events consistent, and you want to build the number of conversions on them and continue to use the same campaigns with ad sets optimized for the same events. You're creating your variation at the ad set level.
Yes, you'll be creating new ad sets with different images and different ad angles. Yes, you will be having variation and testing and scaling, but the campaigns will stay consistent. Campaigns can stay, ad accounts can stay, fan pages can stay, and same with the conversion objectives. They can stay too. When you start to build momentum on these things and consistency on these things, the algorithm favors you. Here's what consistency, AKA trust, reads with the algorithm.
Now let's talk about next level tracking. Garbage in, garbage out. Facebook is a system with inputs, processes, outputs, and feedback. To improve Facebook's ability to self-improve, what you want to do is improve the quality of its feedback. This is done with improved tracking. If you're a human, which you are, and-
So, you know, if you're a human, which you are, and you want to learn more, which you do, then improving the feedback mechanism is dramatic. You know feedback, with the absence of feedback, we would never learn. Feedback is what allows us to learn. And so the same with us, the algorithm, feedback allows it to learn more rapidly. It's like X-ray vision.
And that's my next point, X-ray vision. So you want to turn on automatic advanced matching for your pixel. And this allows you to track and target people who use incognito, adblockers, ghost modes, safari, and other attempts to hide their usage. So how do you do it?
Well, you go to your business manager. Then you go to your business manager settings. Then you go to data sources, which is where your little pixel hangs out. And there's our pixel, consulting.com pixel, which you should have one of. Then you go to the pixel by just clicking up here. And then when you come here, the pixel page should load. Might take a little bit of time, so just click on this. You come here. Then you click on settings.
And then what you want to do, is over here on the right, you can see it says automatic advanced matching. Now yours will be off. Everyone's is off by default. Now, do yourself a favor and turn this guy on. And then you want to see all configured parameters, and you want to make sure all of these are on. And here it says, "Use customer provided data from your website to find more conversions, improve optimization, and remarket to more people." That is basically a euphemism to mean see everything. Right?
Now how does this work? Well, what it does, so you wanna turn this on by the way. But what it basically does is it allows Facebook to track people based on their inputs on the page instead of just the URL cookies and pixels. So let's say someone's got like adblocker, or ghost, ghostery, or something, or cookies turned off, right? Now, most people don't have this stuff off, like most people don't use this, but some do. So if someone's using a ghostery app, an adblocker, or an incognito window, or you know any browser that basically blocks tracking, right?
Now if someone comes and visits that page, and we're not able to fire a pixel on them, then we're not able to track them. If they opt in, we're not able to track them because we can't fire a pixel. If they purchase, we're not able to track them because we can't fire a pixel. But there's a way.
And it only really works, well it's quite fascinating how it actually works. You know when someone clicks on a Facebook ad and then lands on your landing page, there's like a handshake from Facebook over to your landing page where Facebook hands people over to your landing page. So there's a handshake there where there's a connection.
And so if someone lands there, Facebook knows who they are, because they just came from Facebook, so they've got them identified there even if they're not pixeled. Now, if somebody enters in like their email address in a form and registers, or purchases, or does something where they input information, then on the thank you page, we can't fire a pixel either because we're not able to do that to them because they've got cookies or whatever it is disabled.
But instead, what Facebook does is it reads their email or their input, and then it sends that back to Facebook, because we have the Facebook code on our page and it can call out forms, and it can receive the inputs from our forms. And then when it sends that data back, it can search inside the database to see if there's a user account that correlates with that email that was input, and then if there is, then it can assume that that person did that event and it can track it.
So pretty ninja shit, as this module is about, and probably why Facebook hides that feature to be honest. It's pretty buried in there. Kind of like the Trojan horse pitch in that ad, Facebook's buried that one. But you can turn it on and I'm guessing they've just hidden it because of all the heat their getting in the media. They probably were excited to roll it out, but then not so excited given the heat. But turn it on, it's good for your business.
And there's other ways to get to different levels of tracking. So tracking level one is what most people do, and most people copy and paste the Facebook pixel and Google tag manager along with a dozen other scripts, and then they copy/paste that Google tag manager code on their landing pages, and this is a bad idea. It's just too heavy. There's too many scripts. And so, you wanna make sure that you go to at least level two.
And tracking level two is when you delete all of the tags out of your Google tag manager to reduce the weight. And think about it this way, every one tag in there is like the equivalent of one second of load time. So if you've got a huge amount of tags in there, then you're making people wait seconds. And Facebook hates seconds. It likes speed. And so if you're making people wait, Facebook doesn't like that, and then they're gonna punish you. So get rid of your scripts.
People think the more scripts they add into Google tag manager, the better they're going to do. It's actually the opposite. The smaller number of scripts you have the better, because the faster your page speed, and page load speed is more valuable than most scripts.
Now, you're gonna need really only one script in there, which is like your Facebook smart pixel. Most people think they have to put Google analytics in there. Not really. Like I barely use Google analytics to do anything. I don't think I've ever made a useful important decision using things Google analytics gave me. I can't remember the last time I logged in.
So you know, you probably don't need the stuff you think you need, so get rid of it. Reduce the wait, make it faster. And when it's faster, it'll track better. And if you do need more than one script in there, like if you absolutely do need it, that's all right. But then, you wanna give Facebook load priority number one, and you can set the firing sequence in Google tag manager fire Facebook number one.
And then tracking level number three, which is pretty again, and you want to ditch Google tag manager all together. Get rid of that thing. And get a developer to install event based tracking for your site and your funnel. And this allows you to use events instead of URL visits, and events are the future of tracking.
The days of URL based tracking is basically gone. It's turning server side, and it's turning to events. And when you use events, it improves your audiences, the ability for you to create lookalikes. It improves like your retargetting audiences, it improves your conversions, your accuracy of your reports, your attribution accuracy.
And to show you how it works, I'm gonna show you an example of creating custom conversions and audiences based off of events. So we use events, and they're quite a lot better, but if you're just starting out, don't worry about level three, just get to level two. And if you're making lots of money, like then you can go to level three. You'll need a developer to help you with it, because you have to actually code stuff, right?
So it looks like this. If I come here, and if I go to our custom conversions. Like if I go to data sources, custom conversions, and if I find like one of ours here, like let's see ... I found webinar registration. And I go to this specific custom conversion.
Well, you can see here that ... Where is the settings for it? Yeah, so I can't actually ... Oh maybe I can. No, I don't think I can actually edit this thing. Oh no, I can. Sorry.
What I'm looking for down here is, you see where it says event: landed on thank you page? So it's not using a URL. So if I go to create a custom conversion, normally how this would work is you would select your pixel, and then you would fire it on, you know, all URL traffic, and the URL would contain, you know, this URL. Like consolting.com/whatever. And then you would name it.
And it's like server-side tracking instead of client-side tracking. It's more accurate, it's the future, it works way better, and it's definitely where everything is headed. It's harder to install, you need like a developer to install it, but it is worth it if you're making lots of money and you wanna go to that next level. Also, when you create audiences and do everything, I mean you're basically using events instead of pixels. And you wanna ditch those pixels if you wanna go to that next level.
And when you improve your tracking, you improve your Facebook. And the Facebook might look like this. So Facebook sees if the pixel is initializing. And initializing means is it loading. And if the Facebook pixel isn't getting to initialize quickly, then Facebook is going to hate it, and it's going to punish you badly in your auctions, and your delivery, and your reach.
You know, it'll grade your execution and better execution gets better results, because you look like a more serious operator, and Facebook likes that. So these are ways that you can take it to the next level.
The final one is an inside man advantage, and the human side of things. So Facebook's algorithm isn't 100% digital. It has major inputs from Facebook employees, so humans. And when you have a strong relationship with somebody inside of Facebook, you can get a massive edge of everybody else. And getting a contact inside.
So, to get a dedicated account rep at Facebook, you have to spend over $5000 US dollars a day for at least 30 days, and at that point you should reach out to Facebook support manually and ask them if you have an assigned rep, because what we noticed happens is that, you know, we had a rep, but we didn't know we had a rep. And then one day we were just talking to Facebook support, and they were like, "Oh. Yeah I can see you've got a rep assigned." And we were like, "What? We have a rep?" And they were like, "Yeah, you've got one." And we were like, "Who are they?" And then we actually reached out to them and it turns out we had one.
A lot of people who spend more than five grand a day and have done so for more than 30 days probably have a rep, they just don't know about it. And you can ask support. If you don't spend more than five grand a day for at least 30 days, then don't worry about it. And you shouldn't spend this much money just to get a rep. That would be mindless activity. Just wait until you're there. You don't need one before you're there. Even when you're there, you don't even really need one. We didn't realize we had one for years. So they're not essential, but having one can certainly help.
And even if you do spend five grand a day for 30 days, that is not like a guarantee that you're going to have one. This is just roughly the kind of zone you need to be at to get one. But, you know, it's changing all the time. You know, Facebook has internal changes, or more demand, or whatever, then that might change. You might need more than five grand, and you might need to go more than 30 days. And you can just reach out to Facebook support every 30 days once you're in that zone and see what's going on.
And flagged versus white listed. So your account can be flagged by a Facebook employees without you knowing. And what happens here is if your ad account, let's say you've got lots of ads. Some of them get disapproved. And now all of a sudden some of them get flagged for human review. I'm sure you see this sometimes, that some of your ads get held. Like a bunch of them go live, but one of them doesn't, and it's just pending for a long time.
Now that little thing that's pending for a long time, the ones that got approved immediately, the algorithm crawled those and just approved them. It was machine approved instantly. But the ones that just get held back, those aren't machine approved. They're not machine rejected. If they were machine rejected, they would instantly be disapproved. That means that a machine can detect trigger words and things in there and it doesn't like it.
But if it isn't approved and isn't rejected immediately, then it's queued for human review, which means a human's gonna have a look at that thing sometime in the next couple of hours. And if a human looks at it and they don't like it, then it will get disapproved, and they also might flag your account if you have lots of disapprovals. And when your account is flagged, they won't tell you. You won't know.
But when it's flagged, you know you get throttled reach and results, and it's harder to advertise, and you can also have trouble with disapprovals. And reps can get you white listed. And this gives you like improved reach and results and less disapproves. And white listed is basically the inverse of flagged. And so nothing is fine. Flagged is bad. And white listed is the best.
So you'll start your life as nothing. If you get flagged, that's not great. And white listed is where you should want to end up. Now, once your white listed, it means that you know, you've got a lot more trust. And so basically it means an internal compliance at Facebook have reviewed your thing, they've approved it, they know you're not really sketchy, or political, or selling scams or whatever, and you know you're good to spend quite a lot. And they're more lenient on you with all sorts of things.
And then ad approvals. So sometimes, ads are auto disapproved by a machine, even though they're fine. And you should appeal these and try to get them through. And sometimes you have to provide evidence, et cetera, through the approval process when you're chatting with a rep, or when you're chatting with a Facebook support person.
You don't need a rep to appeal your ads. And if you think your ads are fine, truly, and you're not breaking any rules, you should appeal them. And quite a lot of the time, your appeals will get approved, and then you can run those ads that were originally rejected. And reps can get you approved more often. They can get your appeals approved more often, and they can also white list ads which stops them from ever getting caught in the disapproved section. Once they've been white listed once, you can use it again, and again, and again, and it won't get disapproved.
And private betas and info. So when you build rapport with a Facebook rep, they will give you inside info about disapproval reasons, so why your ads got disapproved. And they'll also give you info on glitches, bugs, and outages, so you'll know what's going on with your account. Or faulty reporting, whether it's Facebook, whether it's you, whether it's whatever. And they can also give you info on internal preferences. So you know, what's Facebook's going to be favoring more or less. What Facebook's gonna be concerned about more or less.
And so, you can kind of get a gauge of the internal political wins and compliance wins inside the Facebook machine. And you can also get access to private betas for features that other people don't have. And you know, we've had these with Facebook with our rep, which have been awesome. And also Google reps, what with AdWords and YouTube and stuff. And so, our media buying team builds a strong relationship with our reps at Facebook and Google. And it's worth it.
Because, you know, we do get actual advantages. And this, in turn, these advantages, signal inputs to the algorithm, and then the algorithm gives us better advantages. So there's some ways to get an edge over the competition through having an inside man that can push buttons for the machine. And this is something you wanna to when you're spending in this range.
And this is the form of input. So it could look like this. Client privileges, managed account equals yes or no. So, you know, is this advertiser managed? If yes, then they're probably higher value, and you know, they've probably you know have better clients to use, more valuable clients. So this can help.
Flagged equals, it might have none, which is normal, which is fine. Or it might be red, you know red flagged, which is negative. Or white listed, which is the best, which means, you know, this is the best signal you can have.
And then the threshold. So threshold might be at what, you know, up to what limit will we tolerate disapprovals or negative feedback pieces on things. And if you've been white listed and you're managed, then you've got a higher threshold for things like disapproval ads and things like that, and it's much less likely that your ads and your ad account are just gonna get shut down. So it gives you a higher threshold and a higher tolerance to operate within, as well as advantages.
And then it might look at your access and see that you have access to a private beta or different things, and it might give you those features and unlock advantages for you as well. So that's how you can improve things with an inside man. Or woman.
Now let's talk about creating the perfect storm. So, we've talked about a bunch of different things. And the perfect storm is when we combine all of these things and we really take them to the extreme. Because any one of these things on their own will give an edge. All of these things together, on top of a really good ad, on top of a really good funnel, and on top of a great product and offer, comes just a monumental advantage that will just absolutely annihilate the competition.
And so we've covered all of these things. It's been a journey. But we've covered the engagement bias, so you know, creating an ad that gets attention and it gets engagement, and we want to use the Trojan horse method for this.
Then lexical sentiment bias. So we want to make sure that we're using combinations of words that the algorithm likes. So we're trying to get a positive sentiment, and we're also trying to ignore, or we're trying to not include any trigger words and have a positive sentiment.
Then graphical sentiment bias. So we're trying to include objects Facebook's algorithm likes. Humans, faces, standing, male/female, and smiling, eyes, gaze, and nature, outdoors, sky, hit, animals, things it likes. And we also want to make our stuff look different. Also, Facebook's algorithm should be able to detect positive emotions. Smiles, happy faces, not things that are negative.
Then financial bias. We want to look like money to Facebook. And we do this through having consistent spend and having campaigns, and ad sets, and budgets set at the right tolerances. Not too low, but not too high. Right on the knife's edge of efficiency before entropy sits in.
Then consistency bias. We want to build momentum, and we do that through trust and consistency. So we want to use the same fan page, we want to have content on our fan page, we want to have a verified domain, we want to use the same domain again, and again, and again. We also want to use the same landing page link, the same landing page. We don't want to change the content on the landing page. And then we want to use the same campaign shells, and the same ad account attached to that fan page, and we want to spend consistently over time. And this gives us momentum and a bias.
And then next level tracking. We want to improve Facebook's self-improvement mechanism through improving its feedback, and we do that by allowing Facebook to have X-ray vision by turning on that additional tracking setting under the pixel's settings, which is the hashing thing. Turn that on, and then you want to make sure that if you're using Google tag manager, that you've stripped it of all of its tags. And if you really want to go to that next level, then you want to install event based tracking, get a developer to do that, and go server-side.
And then finally, the inside man advantage, so having somebody at Facebook that you build a relationship with so that you can get white listed and get access to features before others don't, get your account managed, and get an edge in the auction pool.
Now all of these things, like I said, on their own are powerful, but when you combine all of them together with a great funnel, with a great offer, with a great product, then you just dominate. And by playing to the algorithms biases, we get an edge over the competition. We beat them in auctions, we take the best traffic, we get the most reach, and we get it at the best prices. The result? Absolute worldwide domination.
Now you know what to do, so go put it to work. Go implement these strategies. Find your best, proven, winning ads. Make them stronger using these methods. And then go give your competition hell. Because these strategies are seriously next level, and I feel sorry for anybody that has to compete with you when you unleash the power of these strategies on them.
So go put it to work.