the ENTIRE google ads algorithm explained in 24 mins

All right, so in this video, I'm going to show you how the entire Google Ads algorithm actually works. And so a lot of people, like especially these gurus on this platform, they think that Google is a super complicated thing or like every single thing is random. And they always say you need like a million different campaign types or a million dollars in the bank and a real brand name to actually make Google work. But that could be further from the truth at all. Like I started Google with $200 to my name out of a college dorm room. And since then, I've made millions of dollars through Google, but I've also lost hundreds of thousands of dollars doing all these mistakes that gurus say is the right way to run ads. So in this video, I'm going to show you the four stages that the algorithm actually runs through that no one else talks about. And we can use these to scale quicker and more profitably than any single other person. And then also my system of how I run Google Ads in a way where I basically take all the demand on this platform for myself. I'm able to scale past every single other person on this platform. So unlike every single other video on this platform, it's not going to be about how much money you can spend. It's not about making like 10 different campaign types and adjusting all these tiny little settings. Like the only thing that matters with Google is just your ranking, right? How high up on the page you actually are. So for example, search ads. If I type in Asana into Google, the first ranked result is going to be for monday.com, their competitor. They say best project management alternative, monday.com is so much better. So to be completely honest, like a lot of people don't know what sponsored results mean and they just click on the first link because it is the first link and they end up signing for their direct competitor. It's also the exact same thing with shopping ads, right? If I typed in outdoor lighting into Google, chances are most customers are going to click on the first option, not because they like it more than all these other options, but because it's the first immediate choice. is crazy, but this is what the chart actually looks like. All right, so the fourth position gets 7.2% of clicks, the third position gets 10% of clicks, second is 20%, and the first position gets 40% of all the clicks to an organic search. So no one else has actually focused on this, but this literally means that the first position on that search gets 40% of all the money. All right, so obviously the goal is to get to the first slot as often as possible. The way you rank first is just going to be based on this formula. So it's not quality score like everyone always talks about. This is a Google's internal formula that they rank all the ads on their website, determine who shows up first, and who actually makes the most amount of money. So, before I explain this equation and show you guys how you can rank first every single time, first I have to explain how the algorithm actually works. And so, before we get super technical, the first thing you have to know is that whenever someone types in anything into Google, it was going to run through a four-step algorithmic process to determine who to actually show and where they're ranked. And so, within this four-step process, Google's going to take their hundreds of millions, if not billions of ads they have, and then first condense them down only to a couple thousand. And then after condensing down to a few, maybe like 10, and then lastly, it's actually going to rank them and then set the price for them. And so, these are the four steps, right? First, you're going to have retrieval, then elimination, then ranking, and then lastly is the auction. So, I'm going to break down all four of these. And so, the first step is going to be retrieval, and retrieval is the most intensive part of the entire algorithm because Google's going to find out out of all these advertisers, who's actually relevant to the search, like who's actually in the ballpark. As an example, if a customer were to type in orthopedic shoes into Google, Google's going to find all the ads that are somewhat relevant to the search. And so, like work boots, construction boots, indoor shoes, socks, like all this stuff is somewhat relevant. But if you were selling like yoga mats, you're going to get thrown out, right? And so, to actually get shown here, Google's going to scan through basically every single thing you have, which is including but not limiting to your landing page, your titles, your descriptions, your images, your actual price point if it's relevant to that user, your ad assets, like your keywords, and then also your demographics, right? If your images have college students in them, Google's going to scan that and show them to more college students. And this is why you always hear the phrase that your feed is your target. And because all these things are technically your feed, this is all the data you're bringing to Google, and now Google has all this data to find out who's actually going to buy from your store. And then the next step is elimination, right? So, it's taking those millions of ads it got in the retrieval process, and it's going to cut it down by 99.99%, so there's only a couple thousand ads left. And so, the way it actually goes through the elimination process is just based on the formula I gave you guys before. And so this formula isn't like public documentation or anything. It's actually given out by the DOJ because they legally had to do it. Literally just by understanding this equation, you're going to know Google Ads better than any single other person on the platform. So first off, for the ad score, it's just scoring every single ad that was taken in the retrieval process from -100 to 100 based on the actual search of the user. All right, so this is going to be equal to your bid, right? How much you actually paying per clicks, like 50 cents per click or like $3 per click. And normally this is automated by Google unless you're running a standard shopping campaign. And then next is the most important part, which is your predicted click-through rate, right? So how likely is that customer who placed that search to click on your ad? And then lastly, you're going to have beta, which is going to be subtracted from your bid times PCTR. And this is basically like a penalty on if you give users a experience. So basically, if your landing page is slow or if you're just flat out a scammer or you don't ship your products, then your beta is going to get subtracted from your ad score. So Google uses this equation to give you a number between -100 and 100. And the only thing it does in this elimination phase is that if you're negative, you just get thrown out of the ranking completely. So after elimination, you're only going to have a couple thousand ads left. And then Google's actually going to rank those ads and figure out who gets the number one spot. So for the ranking, Google's going to determine this based on the ad score one more time. So before determined who should eliminate based on anyone that's negative, but now it's actually going to rank them based on who's closer to that 100 mark. So then if your bid is just automated by Google anyway, and if you're not a scammer and you have a good-looking store, you don't have any beta, the only thing left to focus on is how can you increase your predicted click-through rate as high as possible? So this value is not only determined by how good your ad looks, but it's also based on all the data that Google has about the person who typed in that search. All right, because at the end of the day, like Google is a data-based platform. If anything, they're a data collection platform because they have so much data on every single person that uses it. So it's pretty crazy to think about, but if you go to myadcenter.google.com and you click on manage privacy, you can see all all information that Google has on you that you never really gave it. Right? So, it has your gender, your age, language, relationships, income, education, industry, employer size, home ownership. And it gets all this data from all the times you've been on Google, your YouTube history, areas where you use Google, and then also all the payments you made through Google Pay. Cuz Google knows all these things without you ever telling it. Like, imagine how much data they have on every single customer that ever searched on Google. So, of course it knows who you are, right? Your identity, what your hobbies, what do you like, but it also knows what you're about to do based on your history of the entire G Suite. So, for example, like if you just bought running shoes a couple months ago, Google knows that, and then based on a million different data points to know when you're about to buy another pair, and then show you ads based on that time frame. So, of course this is perfect for us because we want those types of customers clicking on our ads and buying our products and services. So, Google determines ECTR based on that customer's intent, and then also based on how relevant you are to their buying history, and also their current search. Right? So, Google's going to use all this data to determine your ad score, and the higher your ad score is, the higher you rank on the page. And of course, if you really look at it, like you're not up against a million people or a thousand people, you're just up against a couple other people that you're listed next to. And then after the ranking, it's just the auction. This is the easiest part. It's just going to determine what advertiser is going to pay what amount per click. Right? So, most other ad platforms are PPI, meaning pay per impression. So, if I'm on Facebook and I watch an ad for 5 seconds and I don't end up buying, that advertiser is still going to get charged. With Google, it is PPC, meaning I only get charged if someone sees my ad, sees my price point, sees what I'm selling, and clicks on my ad in order to buy. Right? So, since we are paying per click, Google's going to give us a CPC, which just means cost per click. Right? So, if a customer clicks my ad, how much is Google actually going to charge? And so, actually in old school advertising, CPC literally just meant your relevance. So, you'd have a bunch of keywords, and every single advertiser would bid on those keywords, and that would be your cost per click. But in recent years, CPC, Google defines that as your intent of the customer times your relevance. Right? So, for example, right here, this is a graph of the intent of the customer times the relevance of your ad. Right? So, these are three different people. I'll just say this person doesn't buy anything at all from Google Ads. This person sometimes buys from ads. And then this person she just buys every single thing she sees, right? This is all graded based on the intent and the likelihood that that customer is going to buy from an ad. All right, so for example, maybe the most likely buyer is a 3x, the next person is a 2x, and the last one is a 1x. So, let's say that these three people just typed in orthopedic shoes into Google. And then on the relevant side, let's say there's three ads. And this ad is selling like socks. This ad is selling like running shoes. And then this ad is actually selling orthopedic shoes. And this is the most important part because this actually acts inversely, right? So, the person who's the most relevant only has like a 1x multiplier. The second relevant has like a 2x. And let's say the last person has like a 5x. Now, if you actually connect them like the most relevant ad to the most likely person who's going to buy is only going to pay a $3 CPC for that click. And the same works the other way around, right? If you were the least relevant seller to the most qualified buyer, you're paying a $15 cost per click on something that should have charged you $3. But just to tell you the truth, like the person who's most likely to buy something is going to be buying from the person who's the most relevant to what they're looking for. This also extends to your conversion rate and your average order value, right? If this person buys everything he sees, he's typically going to have a higher likelihood from buying from your store and then also spend hundreds if not thousands of dollars per order. So, clearly this is the type of customer you want, but we only get that if we're relevant enough for them to click on us compared to anyone else. And if we really look at this algorithm from the retrieval to the elimination to the ranking to the auction to your conversion rates and your average order value like all of these things is just based on how relevant your ad is. All right, so your relevance is first going to get you into the algorithm, then it's going to keep you from being eliminated, then it's going to get you ranked higher, and then also get you the most intent customer with a higher conversion rate and a higher average order value. All right, so once you know all this nerdy algorithm stuff, let me just show you how you can scale quicker and more profitably than any single other person and basically own a monopoly over the entire Google Ads search tab. So, every single person on Google is doing one of two things, right? Either they're going super deep on one landing page, one product, and one customer avatar, right? Or they're going super wide, right? So, as many products as possible, as many landing pages as possible, and you can't really target an ideal customer because you have thousands of them. And so, for the longest time, you basically have to choose between them, right? If you go super deep, then you can't cover as much, but if you go super wide, then you can't really target a specific customer. But, in the past couple months, like this trade-off is completely gone because with all the AI tools you have, like you can go deep and wide at the same time. And have hundreds or thousands of different products and landing pages and have each one of them be perfectly relevant to one customer, all at scale. And this is what we call surface area scale, right? How much surface area can you cover on Google where every single time a customer searches something in, you always pop up. And so, for example, with shopping, like if you only have one ad in the search, you really only have one chance for a customer to click on your ad compared to anyone else's. And so, what we do instead is that whenever we have one product that's working, instead of just scaling up the budget, we basically duplicate this product five or six more times but by changing the customer avatar. And so, now when a customer searches up our product, we're not only in slot one, but we're also in slot two, three, four, five, six. And this is crazy to think about, but like there are specific products on Google where whenever you type them in, you have a six out of seven chance of clicking on one of my stores, right? And so, you basically leave the customer with no other choice but to click on you. And then even with search, right? Not only are you getting the 40% from position one, but you're also getting the 20%, the 10%, and the 7% from all the positions below it. And so, at this point, like you're not even outcompeting everyone else. Like you literally have no other competition since you're the only person a customer can click on. And the way you do this at scale is that whenever you duplicate a product, you also change the price point, the images, the angles, and the demographics. That way, not only do you control the entire search bar, but you also hit for every single type of relevant keyword. And so, since the entire algorithm is just based on relevance anyway, like changing all these things is going to make you relevant to every single type of customer. And so, starting off with pricing, which is definitely the most important part. Like, let's just say you're selling one t-shirt for $30. And so, whenever you duplicate this product, you also want to change its pricing tier. Right? So, you want to have a standard version, a budget version, a discount version, a premium, a luxury version. And just by doing this, you basically capture the entire market where maybe like the high-end people would buy this one, the standard people would buy this one, and the people on the more budget side would buy this one. And the reason why this works so well is because people love typing in the price into Google. So, for example, people would type in like best shirts under $30, for example. And if you didn't have these two options, then you're completely thrown out of the auction. The same works on the other side, right? If someone types in best shirts under $100, you're more likely to appear for these two listings than you are to appear just for your standard one, right? So, it's all about being as relevant as possible with every single pricing tier. And then for images, it's basically the same thing where people just buy from people that look like that, right? So, how can we be as relevant in the images to every single customer avatar? And so, if you go to Google right now and you type in like black jacket, these are the types of ads you're going to see, right? Just a piece of clothing on a white background. And so, what we do instead is we basically take this product and take this image and duplicate it against a bunch of different customer avatars. Right? So, this person maybe is more like leisurewear, this person is more like activewear, and then this person is more corporate. And literally just by doing this, like you're tackling three different angles and three different use cases, and each one of these customers can actually see themselves using the product based on what they actually like to do. I mean, like even just looking at these images, like this is what every single other person does on Google, and these are how other marketing our products. Right? So, clearly we're going to become more relevant to a lot more use cases than any single other person. By far the most important is always going to be your titles and your descriptions. Right? Because your titles and descriptions are directly relevant to a customer search, and they act as like SEO towards your brand. And the main thing I always see people get wrong is that they write these titles and descriptions for the customer and not for the algorithm, right? But if you really think about it, like an algorithm's going to scan your titles and descriptions within 200 milliseconds. You have to make sure you have all the keywords stuffed in there that with the algorithm actually picks it up. So for example, if I was selling a heated jacket, I'm obviously not going to only have my title as heated jacket because I only compete for basically two keywords. And so way better title for this would be like heated jacket for men, USB rechargeable, temperature control for hiking, skiing, and outdoor work, water resistant, black. Because really look at all the keywords you're popping up for. Right, we're popping up for a jacket, heated, for men, USB rechargeable. Like basically every single word in this title, we're entering in a completely different keyword pool. So now instead of only for one search or rather two keywords, we're showing up for like a hundred different keywords. And this is how I format all my titles. Right, so I would have a main adjective, then the product, then who it's for, then all the USBs, so all the unique selling points and what actually product does, then the use cases, so like hiking, skiing, and outdoor work, and then lastly just leading it off with some specifics. And the way you actually find these keywords is just in Google Ads Keyword Planner. It's just a free tool you can use to just see what people are actually searching up for a particular product. So for example, if I just typed in heated jacket and set this to my target country and click on get results, I see all the keywords that people actually search up when they type in heated jacket. And so for example right here, we have electric jacket, self-heating jacket, heated coat, heated hunting clothes, and I can actually download all these keywords, throw them into a cloud, give it my format, and then I can have titles built out in like less than a minute. And then literally the same thing with descriptions. Right, so your description just give you more room to target every single customer. So you basically want to have like two paragraphs and just stuff every single keyword as possible. And then in these paragraphs, like you have so much more room to talk about like winter, hunting, construction working, bad circulation. But now all these different keywords are now searches that you pop up for. Right, so just by optimizing these three things, right, your pricing, images, titles, and descriptions, like that's going to make you more relevant per user, which is going to rank you first. But then by duplicating all these products, it basically puts you on the first, second, third, fourth, and fifth slot. Right, so now let's actually run Google Ads in a way where you can scale up past every single person on this platform and not waste a bunch of money doing all these mistakes that everyone else always talks about. And so, the most important thing with actually running Google Ads profitably, and I see a lot of people always get this wrong, is that consolidation will always beat segmentation. And so, while every single other guru is telling you to run like 10 different campaigns and separate all the data and segment everything by niche or by product lines, like this is the complete wrong way to go about it. And so, your goal is literally the opposite, right? How can you become as consolidated as possible where basically you're only really running one campaign? And so, of course, sometimes you're going to have to make adjustments like brand and search campaigns, PMAX campaigns, shopping campaigns. You might have to layer them in between each other, but the main goal is how close can you get to basically only running one of them? And so, in e-commerce specifically, like the main campaign you're always going to run will always be a PMAX campaign. All right? It's just short for performance max, and it's like Google's AI campaign type. And so, by default, a performance max campaign is going to hit on Google search and Google shopping and then also YouTube, Gmail, display, Google Maps, and Discover. And then if you you look up how to start a PMAX campaign for e-commerce, you can find a million different tutorials on it, but every single person just does it in the standard way that actually wastes you a bunch of money. And so, for e-commerce, you don't really want to start out with Gmail ads or display ads or any single one of these other ad groups except for just search and shopping, right? Through search and shopping, you're going to make 99% of all your revenue, and most of it's going to come from shopping. And so, the way you actually do this is you just click on create and then campaign. And then right here, click on create a campaign without guidance, select on performance max, and set this up for purchases, your account goal, and then select your GMC. And then right here, I basically leave all this stuff blank. I'll get into what a target return on ad spend actually is cuz this is super important. And we just click on next. And when you get to locations, just always click on presence. I think it's auto set to presence or interest, but that's going to waste you a bunch of money. Just click on presence. And then also another big thing is when you click on devices, just make sure you turn off TV screens. I don't think I've ever got a purchase from a TV screen. But then on this page, when they actually ask you for all the assets, like whatever you fill in here, it's going to be shown as headlines, YouTube ads, Gmail ads. So, you want to leave every single thing you see here blank, so where you only have product selected at the top right. And then just going down a little bit more, you can turn off text, image, and video enhancement. That way you're actually not shown as YouTube or Gmail or all these other different ad placements that actually don't make you money. And then you can scroll down and click on next, and I mean you're basically done. You really don't have to do anything else but just set your budget. And so, this worked for me, but sometimes it will give you an error. So, if that ever happens, just fill everything out and actually make that asset group. And then if you just click on duplicate, you can leave this one entirely blank, publish that one, and then you can go ahead and pause the one we made before. And one thing you should never look at will always be ad strength. So, my ad strength has been poor in every single campaign that I run because I'm not really using the campaign on how it's meant to be ran, but I'm using it in a way where I can make the most amount of money possible. And literally no one knows about this, but now you're using Google's AI system, but you're not being placed for Gmail or Discover and Google Maps and all these shitty placements that burn you money, and you're really only being targeted where customers actually buy, which is once again, search and shopping. And so, this is the main campaign that I run for e-com. Of course, you can also just do a standard shopping campaign, which is not going to be AI, but it will give you a little bit more control on your cost per click. But basically, all the brands I start right now, I just run a PMax campaign until I'm spending like maybe 5 to 10K a day, and if I'm ever at the point where it could be more optimized, then I'll either add in asset groups, tweak the target ROAS, or maybe layer underneath it a standard shopping campaign. And this campaign is going to result in like 99% of the revenue I get, and the other 1% sometimes come from a branded search campaign, which is also pretty easy to set up. And I actually rarely see these two cannibalizing against each other, but a branded search campaign is whenever someone types in my brand name, I actually pop first instead of a competitor or a view site. And whenever you turn on a campaign, you're always going to hit the learning phase, right? This is common with every single ad platform and every single campaign type. And so, Google defines this learning phase as having 30 conversions in 30 days. So, whenever you hit one conversion a day for at least 30 days, then you're outside of this learning phase. And all this means that whenever you launch a campaign or start a new ad account, like Google has no idea who you are. They don't have any data on your customers or what price you're selling or who's the ideal customer for your service. Or so the thing is if you don't focus on relevance and all the things I talked about before like aren't really in an alignment, you're basically going to be in this learning phase forever because you're making it super hard for Google to gather the data on your customers. But on the other side like if every single thing is in alignment from your titles, your descriptions, your images, your price point, the actual products you're selling, like you can exit this learning phase within like the first like if you're launching ads and it only goes up from there. Because especially in these early days like you're just stacking on a bunch of conversions, you're doubling down on all the customer data, and then Google uses that data to target better customers for your store. Or so this definition of 30 conversions in 30 days really only applies if nothing else is in alignment and it takes you a long time to actually gather that data. But for the average trendline at least across all the stores I see is that maybe the first couple days is a little bit shaky, but then on day three maybe you get your first order. Or so now Google has some data to work with or maybe the next day you start hitting like two or three orders. And so Google basically goes through this week-long cycle of just gathering data and then improving your next conversions by targeting better customers. Or so even on the first store that I ran like I got my first store order two days in and literally the day after that I got five. And so now with your TROAS a lot of people don't know this even exists with Google since it doesn't exist on Facebook or Instagram or TikTok. But TROAS basically means you can set a target for how much revenue you want to make based on your ad spend. Or so on Google if you set a 300% ROAS, right? Which basically means a 3x return. Out of every dollar you spend with Google, you want Google to make you $3 back in revenue. Or so this is pretty crazy to think about because you're literally telling Google the target of what you want to hit. It's almost like you're selling a dollar bill for three dollar bills, right? Or so TROAS just means target return on ad spend. Right? So your target is $3 for every dollar you spend with ads. And so fun fact, actually ROAS just means return on my ad spend. What I see a lot of people do which just kills their entire campaign is you never want to set a target to early. So typically I never set a target for the first 3 months and I just let my PMAX campaign just do its thing. But whenever I see some sort of plateau or drop in efficiency, then I would just add in a target over here and just let Google actually reach for that target. But typically a lot of the campaigns that I run, even one that I scaled up to like 500k in ad spend, I actually never even added a TROAS because the campaign was so profitable that I didn't want to limit it and segment it. But if I keep on going with this PMAX campaign here, eventually when you start scaling up the budget and you start spending like 20, 30 thousand dollars a day with ads, you're eventually going to hit a plateau where every time you increase the budget, you're going to have a marginal return. And so whenever we start spending more than like 20 to 25k on ads, we basically just layer this PMAX campaign with just a standard shopping campaign. Because you notice the early signs of this campaign will always be better than just increasing the campaign that's already being spent. But even if I just go into the earlier stages of the campaign, let's say a total of like a hundred dollars in spend, you can literally see what products are getting the most amount of clicks, what keywords are getting the most amount of clicks, and then what products aren't selling well and what keywords are just wasting money. And so from this point on, like basically every single Sunday, the only thing I change in the Google Ads account is just either tweaking the TROAS, upping the budget, or removing the keywords and removing products. Right, but even going into long-term scaling, like at the end of the day Google is a super long-term platform. Like no matter what you build on Google, it's going to have longevity for two, three, even up to five years long with just consistent returns. So at the end of the day, like every single other platform, like Facebook, TikTok, and Instagram, they're going to have basically a sine wave, like up, down, up, down, up, down. Then with Google, if every single thing is in alignment, of course you're going to have some peaks and some dips throughout the week. But if you look week after week and you set up everything properly and you just kind of let Google do its thing, you're going to have a trend line that basically looks like this all the way up. But just by going back into this formula, like you can increase your predicted click-through rate today just by being more relevant to what customers are searching for. Right, but with beta and customer quality, this is going to be something that you earn every single day of you gathering conversions and giving customers a good experience. And just these two things alone are going to decrease your beta, right? Decrease your penalty. That way you can actually rank higher consistently every single time. Right? And that's honestly why Google works so well for the long term. And then also the reason why I keep on building more stores, right? Because the longer you run a store, the more Google's going to trust you, the more they're going to reduce your penalty and reduce your beta, and the cheaper your clicks are going to be, and the higher you're going to rank on the search bar. Right? And then you layer this long-term approach where you're just being super relevant to every single user, service area scaling, you're building out more stores, is exactly what allowed me to outcompete all these big-name brands like IKEA, AliExpress, Amazon, Home Depot, because I'm simply more relevant per user and Google gives me more trust because of that. And even on the first store I built, like I think I started around mid-April, I did around 3.5k in revenue with around 40% margins of 1.5k in profit. And then literally 3 months later, we had a 50k month with $20,000 in profit. And then 3 months later in quarter 4, I did $270,000 with $56,000 in profit. All while remaining above a 3x ROI the entire store's life cycle. And the only reason this is possible is because I was super relevant and super deep on one customer avatar, one major pain point, one major use case, and then I basically multiplied that like 10,000 times across the entire store. And of course there's like a million different nuances and a bunch of like tiny little adjustments where if I covered all of them, this video would be like 100 hours long. And so if you want me to literally hold your hand through this entire process of starting Google Ads from A to Z, there should be a link below. But other than that, that's the video. Thank you guys so much for sticking to the end. If you enjoyed, definitely like, comment, subscribe, all that fun stuff, and I'll catch you in the next one. Peace.