WEX talking about Pax, what used to be like, it was the savior of all things and light at the end of the tunnel. And we were so excited about it. And now you're, more jaded than I am with Pax I've been, I've been burned, like, how did I lose like 1% of my revenue? Like something out. Little bit messed up here. So that's, that's that what we're gonna be talking about today is how to set it up to limit any sort of over attribution and get it to a. Diverse campaign. Welcome to The Ultimate Guy for Google adss, part five, Synco. That's block. Yeah, that's a lot of parts. Sorry. Yeah. You know what's funny, dude? Is we used to do all these as one video. I know where people who've been longtime watchers as a channel and it just got too complex. It was like Well, it's like just rapid fire curriculum. Yeah. Like when you're running, your brand might use manual CPC and go over to the top place, but don't over fence. It's like, what? So Yeah. Thanks everyone. Yeah. It just became an auction so fast. Yeah, Yeah. Yeah. You're really good at that. Dude. You missed your, you could have been an auctioneer or you could have been a, what did the A four step caller. You know what's really funny about that? If you Google John Mo. There's an auction house named John Moran. That's why I couldn't get that url. But yeah, it was actually, there's an auctioneer called John Moran. I'm like, oh, it's probably like a long less cousin. I'm gonna ask my dad if he has like any, you know, long lost brothers or something of that, like maybe he had like half a medallion somewhere and he is just gonna find the person with the other half a medallion and yeah. And then we're gonna link up his long lost brothers and that's the found by my cousin. Love the turn. This video is taken like all of a sudden. It's a Nicholas Cage movie. I'm gonna steal the Declaration of Independence. That's why you're so good at Google Ads. Auctioneering is in your blood. That a d d? Yeah. Yep. and energy drinks. All right, PAX. Cool. Yeah, we're talking about Pax. What used to be like it was the savior of all things and. Light at the end of the tunnel and we were so excited about it. And now you're, you're, you're more jaded than I am with Pmax I've been, I've been burned, you know, like Yeah. I still like it. I like, you know what, I dunno why I like Pmax. It's because setting up Pmax is like setting up a Facebook campaign. Like it's so easy. You know what I mean? It's just like, oh, you just, you just follow the yellow mc road. Right. And it always shows good roaz. Yeah, right. Yeah. And no matter what, it always looks good. So. Yep. Exactly. And that's where I've been burned by it so much too, cuz it's all of a sudden, like when you start to try to make it do something, that's when you get, that's when you kinda see, oh, that's what was going on. What you mean by that? Sometimes you wanna scale. And then it's like, okay, it's scaling, but then you check like overall mer and it just goes down. Like I'm, I'm looking at campaigns that are like 20% better than 30% better than 40% better. Then I look at analytics, I'm like, how did I lose like 1% of my revenue? Like something, something got a little bit messed up here. Mm-hmm. And so what we found too, oh, that happens sometimes too often. But also what is the best, feature of the campaign is it will hunt for any conversions it possibly can. Even if it didn't originate with them, and that's where we started to see things go a little, haywire. You're just gonna cannibalize everything you're doing if there's not easy conversions to capture. Yeah. So you start to crank it up and Google will spend that money and then it will earn a remarketing click somewhere, and then it'll show that it had roaz, on that sale. Like let's say 2 3, 4 x doesn't really matter. It was a returning user from. You know, and so that's when it's like, did we really do anything? Not really. Mm-hmm. So when you try to scale, it will attribute more sales to itself, but how is it attributing more sales to itself is where either is, getting good non-brand cold traffic that can't scale and it's, rightfully, earning itself those conversions or. Is sort of cannibalizing other channels, like, direct channel or your, organic or your email. So it starts to over attribute itself to itself a little bit. And, that's where things can go wrong, especially if you have a lot of good brand recognition and you're not, Necessarily wanting to max out the brand, it'll start to capture that. or if you have people that are coming in from Facebook or Instagram or organic, it doesn't matter wherever they're originally coming from, it will remarket, and it will show an ad to it. And the bad part about PAX two is because it can have an engaged view, convers. And it will make YouTube videos by default. It doesn't even need to earn a click to re attribute or to attribute conversions to itself. And that's where it gets a little bit screwy is because you can have a person that's getting ready to convert and as long as they see a YouTube ad, within by default three days of converting Pmax will say, yep, I got someone. so that's where things get a little bit haywire, is it, it can over attribute itself too much. Flip the coin for us cuz we're teaching people how to build this thing. And so like right now everybody's like, well, I'm gonna stop watching this video and go watch how to build Facebook again. Like when it works. Yeah, it's amazing, right? these are the, the pitfalls. But when it works is, it's unbelievable. And it will, it will be everywhere. it will be on search, on shopping, on YouTube, on G S P, on Discover, on display. It'll do fantastic remarketing. It can show up, just like a d s a campaign for all your search categories. And it will give you good non-brand top placement in shopping. And it, it works extremely well. that's what we're gonna be talking about today, is how to set it up to limit any sort of over attribution and get it to a really diverse campaign. and some of the learning lessons that we've learned over this last year because we found out that, oh, you know, doing it one way is way worse than doing it the other way. Or, oh, that's actually how asset groups work. Our signals are actually a lot less, a lot weaker than we thought. Mm-hmm. um, all the little tips and tricks that we found out about how to build an. To a high degree of, scalable non-brand cold traffic. and that's what we're gonna be doing today and that's the kind of the reasons why we found out how this works. it's pretty cool. I'm excited. All right, so let's dive in cuz I have an account that we can share. I had Tom Brees audit our YouTube account by the way. Mm-hmm. and he looked at all of our videos and he found the inflection point where the views go up and it's right when you start sharing your screen. Swear to God. I bet. Yeah. I can imagine. Yeah. That, and this is crazy, but we get more views, uh, if, for my videos anyway, when my hair. Not kidding. Tom's like, yeah, so he is like, when you share your screen and when your hair's down not up, and I'm like, that's the strangest thing I've ever heard, but okay, so I just share my screen. Your hair's down. Look at that. We're, bro, I know this is like, this is the John Wick of solutions. Eight videos. Oh man. Here we go. you kinda look like John Wick too, which is pretty funny. You're like, that's the nicest thing you've ever said to me, with you every, uh, share with you just last 30 days of performance, so, so that we're. Knowing that it's working well. 19,000 in cost, 82,000 in conversion value. So it's, it's getting a really good return. It's a little bit, little bit above a four. So when we're looking at the asset groups, we built out 59 different asset groups. most will build out, most companies I've seen. the very large company I just audited recently had one. and it was, uh, it was a PE every, every PAX audit I. Every Pmax audit I do. Yeah. Is one asset group, sometimes three. Yeah. Yeah. And crazy. Overachievers. and then it, there's three. And then at least one's paused. Well, all the products are always disabled. Right. That's the other thing that I don't understand, like why is Pmax, is it more difficult to check or see disabled products instead of pmax? Uh, no. What, what's interesting, there's actually some really advanced strategies on why you would disable a, products. And one of 'em is actually if you wanna run a standard shopping campaign in addition to it, which we'll talk about, that actually works really well because this campaign will still display remarket and will still capture everywhere else, but will not cannibalize your, We're marketing only. Um, yeah, so the smart shopping hack, basically the problem is when I'm auditing these accounts, I don't think that's what they're doing. I don't think that's Yes. I can't tell why they're doing it wrong. I can tell you why I do it wrong. So that so here's what's cool is there's 59 different asset groups. is important. What we found out is, Google will take a sum total of all of the impressions and evenly distribute them, throughout the asset groups, sort of regardless of signal. And so it gets a little bit, it, it's a very interesting thought process because what happens is if you have a hundred thousand impressions in a week, and you have one asset group, that one asset group will get all a hundred thousand, a hundred thousand impressions. If you have two asset groups, each one gets about 50, 50 ish. Like one might get 40, one might get 60. But then if you have three asset groups, Each one will get 33 ish. 33,000. So it kind of evenly distributes them. It's odd because it, makes sense because they, the campaigns can go broad, like in, they can search for pretty much anything that's around your products or service or site, whatever it may, may be. And it doesn't necessarily need to be stuck to an asset group. And that's the part where I think things get a little bit haywired when people build these out. They think, okay. But what it looks like here, is this is typically what we see if I have this button down, Asset group peer, and people say, aha, the interesting demo. Perfect. So people that are interested in men's shorts are only gonna go to the, the button down. Obviously it, it's not, that's why it's set up this way, but that's what people think. They think whatever they put in this signal that's gonna go inside of this asset group, it's, it's completely false. so this signal actually does very, very, very little. That's what we found out. The signal is sort of like, I, think it was sort of put there just to give the people feeling that, okay, at least I'm sending it off in the right direction. It really doesn't work that way at all. your site's data and your product feed, if it's e-commerce, is 10 times more powerful than what that signal says there. Hmm. So what's interesting is we started to kind of just group these signals together, and then diversify them with add. So what happens is you'll see I have an interest in demo, and it says men's shorts, men's apparel. There's six more here, but this is going to a button down because yes, there's people interested in men's apparel easy, and if we see another interest in demo, we see like men's, short men's apparel. This is going to the polo. So what we found is that the just doesn't need to be your. It needs to not be your own data. It needs to be something that is categorically around the type of product that you're trying to sell, but then you are using your ads to differentiate the traffic. You're very rarely gonna get a person that's looking for shorts and then click on shoes and then go and buy shorts. It does happen, but not as much. Not as often. So your ad copy is doing the heavy lifting on directing the users to the areas that need where they need to go. It's not the signal that's bringing in the user. The signal is coming in. But if someone's looking for shorts, the asset group's listing group will say, here's my shorts and the shorts. Because that data in the feed matches up to the search term will naturally show this asset group as the asset group that needs to be shown. And then people are coming in. So it's very, very strange. It's almost like running a standard shopping at a DSA campaign together, and this audience signals almost like an observation. It really doesn't make too much of a difference in the audience signal. It's nice spot. Do you think that'll change long-term, John, or do you think that's the way that PMM Max is built? I think that's the way that Pmax is built. I think that because things can go so broad, I will get. men's polo, if I go after the broad keyword of men's apparel, Hmm. Here's why I'm asking because, uh, Google for, you know, federated learning cohorts flock. Mm-hmm. which they say they've deprecated and I don't believe, and now, what is it called? topics A api. Topics api. So I think flock is just the, topics API is winner dressing for flock regardless. It would make a lot of sense to me that topics api. The audience targeting. And if they really invest heavily into topics api, I then Pmax starts to kind of veer in that direction. Yeah I think that's kind of like what's going on here. if you think about it, your, flock, your, flock of people essentially, or your topics, api, I, of these people have these three cookies for these three weeks. One of those cookies is gonna be like, well, it's, he was looking at socks. Well, that's men's apparel. Perfect. That's an interest of men's apparel. Well, men's apparel on a men's apparel company, how many asset groups does that? You see 59, 48, 50, 50, 44. There's so over. It's a broad match for interests. Exactly. That's a great way to put it. It's like a broad match for match for interest, and so they overlap so much, but you let the person dictate their own path, and that's where you'll see, like for example, I have this one, the Grateful Dead Golf. Well, if I go into the Grateful Their golf shirt that had 44 asset groups, that's going into button down polo and all products, and by all products one They chose what they wanted based on the ad that they saw, the product that they clicked on. So this is more just suggestive targeting and How do I describe this? Like a, just a, a big wall of ads and then people are like, well, that's what I'm interested in. Okay. Was that from the signal? No. The signal didn't do much. The signal was, yes, this is a male apparel interested user. Now choose 2000 SKUs. And so that's what's crazy is the way that you can really help, win non-brand cold traffic. Inside of Performance. Max is very, very well diversification of your asset groups. when you shove everything into one. The learning algorithm basically just is a, is a, you know, it's a stupid machine. When you're talking about recognition of what a user would, or what the owner of the ads campaign would want. All they know is this word does good, fine, that's good. But when this word does good and it's your brand, why? Well, it didn't really test much else. Why? Well, it had five pictures and all of your products. So people that are Googling your brand always click on one of your products and always. So it's just gonna have you do that all the time. It's not diversifying the signals, which I know sounds stupid, but the more signals you can put in there is the better that it will go. But you really need to diversify your asset groups. Most people will say, well, you don't wanna diversify your asset groups too much. I would agree. I think that it's solely dependent upon your ad spend. if you look, we have a $500, day budget, which is not a small budget. it's a healthy budget. More than enough for 59 asset groups. When you look inside of, I don't know why that does that. If you look at the, all of my acid groups here, like there's activity in each one of these. you know, it's not, necessarily saying like, oh, there's only, This fish vest. Yeah. I didn't really get much activity for, that's really not a popular product. but yeah, I would say that, this is the last 30 day period. there's more than enough to identify kind of what is my top performing products and, why. so I do see that around it looks to be, for some reason the chatter on the, interwebs is going more towards like reducing down your acid groups. I still say don't do that. I still say that you, should have quite. The more the better You're, you say you reduced on your asset groups. What have you heard? It's a few people on LinkedIn. Really? Yeah, it follow. People are just like, Hey less asset groups the better. Having too many takes too long. There might be some use cases. Oh, that's what it is. It's a bunch of low spend. That's what I'm saying. I think spend dictates. Yeah, that's what we should be telling everyone. It's like how many asset groups? Well, how, I mean, I'm not gonna put, yeah, I'm not gonna put 59, 59 acid groups. There's only $20 a day Right? Then yeah, that makes sense. I don't have enough ad spend to fill up the activity for all those products. So what we really should be telling people is, depending upon the level of spend, that's what's gonna dictate how many asset groups you have, more asset groups, you have the higher spend you're going to need. you can break them out into separate campaigns. It's hit or miss. It is completely hit or miss, and it's out of everyone's control. And I will fight people over that, that topic. Mm-hmm. reason being is that you are gonna have people that will visit multiple times. And so when you visit multiple times, you're gonna jump between PAXs. So where it started and where it ended. If you don't care, that's okay. But for example, if I pull up into this Pax campaign, cause we're gonna, kind of jump into the brand issues. I jump into this campaign here, you'll see I have two PAXs, PAX for Women's and PAX for men's products primarily. The problem is though, is I did make $3,400 last week on my brand campaign that I can't tell you where they. They could have started in women's, they could have come back inside of the section one 19 brand clicked on a brand search ad and bought a women's apparel, and now this campaign here says I sold more men's apparel. That's not true. So just know that there is going to be overlap post first click between the PMAX campaigns and if you have a PAX campaign that is trying to sell a product and it's showing up for more generic search terms. It's not really like a standard search campaign, like they didn't search for something and click on that problem or product. that's something I think a standard search campaign would be better if you need to make a push on a specific group of items. PAX is just kind of whatever sell sells and it will do very well. Just know that you have less control over it. So multiple Pax, I'm using it so I can have the spend dedicated to getting the product awareness, but whatever sells. So that's why I'm splitting mine out. It's like, I know I have X amount of dollars pushing men's products. I have X amount of dollars pushing women's products. That's all I care about. I just need, both of 'em have exposure if they're gonna jump back and forth and sell. I don't look to see how is men's performing? How is women's performing? I look, I look at analytics for that if I wanted to, or look in the backend to Shopify, but I'm not gonna use Google as to dictate how, or to dictate my understanding of is this product doing better or not? So just know that they're going to jump back and. With the brand, showing up in Pax because it is gonna cannibalize the return traffic between the two of 'em that be, makes it a little bit more difficult. what you can do to mitigate that is to run your own brand campaign. And when you run your own brand campaign on search, in addition to pax, you do help pull that out. What I mean by that is you'll see this brand campaign here. Last 30 days, I spent 85 bucks. So $85. I made almost $14,000 in sales. So this campaign here, because I have 9 cent cost per click, and my cost per conversion is pretty much like. Like 18 or like, it's like 80 cents, something ridiculously cheap. I'm taking this 14,000 out of this Pmax campaign here. That would've captured it, it would've shown up for it. because I have another campaign that's doing better than my brand campaign or than my, branded search inside of Pax, my pmax is then more free to use that total daily ad spend on, you'll see one brand here, one brand here. Everything else when you sort descending by conversion value is all non-brand except for like here. which that could have been a non-brand. They were just return traffic. But this is just last seven days, by the way. Like it's, it's really nice and clean. Yeah, I think you had a question about the, categories, right? The ask you just mentioned earlier, the importance of categorization, and then when you started talking about cross-pollination of asset groups, it made me think about, number one, if people are gonna segment categorically, they're actually not gonna be able to see that it's successful. Mm-hmm. So, and that's probably why everybody's going back towards, oh, everything in one asset group. Because even though it's helping, they can't see that it's helping because like no per asset group attribution. And so they think like, oh, just put it all into one. It does help, but you can't see that. So that's just a comment, actually a question. Do you agree with that or not? No, I would I agree that the lack of visibility usually makes people just throw their hands in the air and say, well, I guess we'll just put it all back into one. And I would say that's, that's a reaction to the unknown that I think is not a, fair assessment. not from what people, who purchase products online, their purchase habit. I think we can get a little bit smarter, with our, segmentation. I'll share with you what I mean inside of these asset groups, because I've grouped them categorically. I've actually capitalizing on a normal humans buying process in order to replicate that same buying process at scale. When you are looking. Board shorts, for example. Very rarely do you look at a board short, section on a website and then say, well, I'm buying the, bow tie. You know, it doesn't happen. You're gonna buy a bore shorts. It just may not be the exact board short that you clicked. From the ad, you may find a different one. So the, what I'm doing here is I'm essentially trying to bring Google, to the forefront of finding repeat traffic the way that they purchase. So people that are looking for polos 80% of the time by polos, people that are look for board shorts 80% of the time by board shores. People that are looking for henley's or sweaters, most of the time they're, purchasing within that. So what's nice is when I look at Instead of this pmax campaign here, I can look at the insights and I should be able to see something like this If I look for a golf shirt and then I click on the 44 golf shirt and then they bought a polo. That's a golf shirt essentially. That's the insane with the button downs. So there is all products here. They may have bought a, a polo or, essentially a button down from here. Okay, so let's look at like, Did they buy hats? So they bought a zip up and converted, the all products board charts. They bought everything here. That didn't happen at all. Okay, so we gotta look at the next thing here. grateful that, I'm sorry. That didn't happen at all. What do you mean? Oh, they didn't buy any hats? The asset group did not serve for any hats. They bought bunch of other products within the hat asset group. They bought everything but hats. Yep. And then we see grateful. That button down shirt here. This one. What do they buy? do the conversion value. Fish robes. What is a fish robe? It's a robe. It's a big, yeah, it's a, it's a big robe. so they, I must have one right? they bought a button down shirt and then they also bought a fish robe. You know what's funny is this actually wasn't a fish rob fish robes a $400. they bought a button down shirt. They just, they probably came in through search. Mm-hmm. So here's what's crazy though, is a $29, that's our budget. Why? It just blew my mind for a second., I'd never even thought about. Huh? Search attribution is based off of the phrase clicked. Yeah. But if you take that and carry it into Pmax, it means that asset group attribution is going to be variable based off of the channel they came in on. Yeah. Oh, that's a nightmare. It is. So that's why when you group things categorically, at least I can show the listing group that's proper. So here's what's funny is that $29 is gonna go into all products. What do they click on? A search or a shopping ad? Not sure, but I know they bought a button down shirt. Fish robes. I know they didn't buy a robe. It's the button down shirt price. And then they have the button down. Then they, we probably clicked on a button down shopping ad. So when I go watch this, if I search for Grateful Dead button down shirt, I see my ad here, here, here, here. And. What am I clicking on? Well, this one right here. What asset group in Pmax was that? Hmm. No idea. Whatever asset group's working well. Okay. But I also see that my button down shirts here are top dead and center. This is coming from by button down asset group. I know that for sure. So what I'm telling Google is people that Google these, usually people look to know that because we're running pmax and not standard shopping. Say again? You only know that because we're running pmax and not standard shopping. If we were running pmax and standard shopping, couldn't that also just be a standard shopping ad? Yes, but my listing group only has my, button down. shirts turned on in those asset groups. Got it. Okay. Mm-hmm. So I'm old, but that's what's nice to, and if I go into, my asset groups here and I look at like my board shorts and I look at my listing groups, these here are all my board short. Dude, this is a pmax is a data nightmare. It is. There's no consistency in terms of traffic acquisition. No, it isn't. But what's nice I can set it up for a success that as long as Google sees that people that are doing something over here most often do something over here, which is when they search. Grab that button down. Shirts, they always buy the button down shirts inside this asset group. So now when people are gonna be Googling button down, shirts, hopefully the listing group that has the button down is gonna shield the shopping app, but also the search ads. So I can try to make sure that it's most often from the first time click legitimate. The bad part is, is then you're gonna do this. People are gonna come back and search the brand name and now I have no idea. Is this all products? Is this, you know the robe? Sure. Is it a bomber jacket? Yeah. Did anybody buy a bomber jacket? No. They Googled, they clicked on this ad here, and then they, they went wherever they wanted to after that. They're like, oh, that's not what I wanted. I'm, I was looking for men's tops and now nothing in Google makes sense anymore. Instead of the Google Ads dashboard. it's a. Complete nightmare. You cannot track a return user. Yeah. Well, and if you think about your own buying experience, dude, I don't remember the last time I clicked on an ad and bought right then and there. Never do. It's always click and return. Click and research. Click and look. Click and Google. Yeah. That's why landing pages don't work. Or squeeze pages, I should say. Mm-hmm. if you look at the last seven days, half of my conversions came from more than two app clicks. So Pmax is halfway right. That's why it's like, and then the half that came from one ad click are just things that couldn't have been tracked, cuz it's probably multiple ad clicks. Right, right. And then what's funny is after, half of my data inside of Pmax is incorrect, then you have to look at the other area here, which is the conversion lag and know that Oh yeah. By the way, it also takes 18 days to get a user to convert. So, it's not half had two, half had more than. So we had a, and then the people that had usually one that's gonna be your branded Right. Which means that they probably had 30 and now we've just lost tracking for them. Yeah. Or they've been visiting you 18 times on Facebook and now they're ready to buy. So they Google the brand name, click once and buy once. So this is a nightmare So what I, what I can say though is what is some, takeaways from this though. What can you control? You can control the amount of ad spend that are being shown to the products that you want to make the most aware. That's something that we know is a constant, is a truth. We know that if we take all of our asset groups and we, well, diversify. And we give them signals that are halfway relevant. We don't use our data too much. You can use your data. That's okay if, if you have a really good L T V or you want those return users and you're okay with paying for those return users, that's okay, but just know that this is going to capture a little bit more than potentially what you may. And so I'm gonna give you a little kind of a status update behind the scenes. Here's what's crazy is in the back end of things, you're gonna see this, see the last 30 days we spent 19,000, may 82, we have a roaz of 436. Good. This looks great when you hop into Pax. majority of the time is good. Non-brand cold traffic, good, everything looks great. It's not that great in the back end. It's good. But here's what we have to always just be aware of, and I'm okay with this because. It's helping us gather an L T V. What I'm okay with is, what I'm gonna share with you now is how much new and returning users is there in that 436 row eyes. No matter how you set up pax, this will happen. Sometimes you definitely don't want this to happen. Sometimes you're absolutely okay with it happening, having the return. So this is the backend I'm using Nor Beam, just to share with you. This data here last 28 days, performance max. I have a 2.4 roaz on new traffic, and I have a 3.48 on returning. Can I just pause you to remind the viewer Google reported what, a four 80 Google reported? Yeah, and it's, it's actually in the last 30, I'm comparing the last 30 to the last 28, so it's gonna be, what was Google's? I just wanted apples to apple. Yeah. So let's do this. That's the 16th to the 12th. Can I say to remind the viewer, I mean to remind myself, So six 16th to the 12th is showing a four 16. Four 16, okay. Yep. And actually, if I just combine this here, performance Max is actually getting a 5 89. Wow. So it's getting a little bit better than what Google's showing. this is the Roaz on that p m max campaign. So we spent 17 actually made 1 0 1 and Google says we. 17 make 80. So it's missing attribution. Google's just horrible at attribution by the way. You can see there's worse, there's 20 grand missing. So, but all right, so let's just stay. that's okay. So we're gonna miss some attribution. However, dude, it's not, I, I wanna say that differently. Yeah, yeah. I know there's 20 grand, there's 20% missing. It's 20%. Like that 20 grand is, you know, depending on your spend, it's like, okay,, like margin, error, whatever. But like 20% of your gross sales, like, holy crap. That's insane. I know. It's huge. Yeah. And what's funny too is like, it's, this is where a lot of times people will have a disagreement where it's like a vendor. things are looking good on one side or things are looking bad on the other other side, but people are not really jiving because it's like, now it's Google versus Shopify. Like Google says it's good, Shopify says it's doing bad. And it's like, well, I don't really care what your Shopify, well, I don't really care what your IT attribution is getting worse and worse and worse, and it's causing more and more headaches for a lot of people. And then, but that's where Pmax actually starts to get a little bit dangerous. It does. It gets real dangerous. Why? With the loss of attribution. Attribution comes when there's more recent activity that it can track in a short time period. What is that return traffic brand? Google's gonna save its own roas, however it wants to. There's only two buttons that you can press inside of pax. Maximize conversions, maximize conversion value. There is no, bring me more non-brand cold traffic or, give me a good top placement or gimme the best impression. It is how find sales. However you possibly can. And if Google says, well that's not brand gold traffic. I don't see sales. Why? Well, it's probably too long. They change devices, they come back directly and Google misses it. How do you save your own row as Google go after the warmer audiences? And that's what we see. So I'll share with you here cuz this is, what is actually happening. So our pmax campaign, we know we spent 17 grand. We only made, or not only, but it's good, we made 41,000 on first time. Good. That's a 2.4 roaz on first time sales. What does this client need? What's your goal? Uh, they wanna be above four. Okay. Mm-hmm. it's $40 cost per required. New customer. Good. The problem though is on the other side. That's four, sorry. 423. First time sales this month. Good stuff. Now what Google's still taking credit for though, is the $60,000 in second, third, fourth, fifth, sixth, seventh, eighth time sales. So I made an additional 60 grand that I just, the campaign woke up today and said, oh, I'm 60 K positive already. Don't try too hard. You already made 60 grand. What does that say? Why doesn't, going into Performance Max and saying New customers only fix. We've done this many, many times. I'll, I can share a few case studies if you wanted to. I have to use a very, very, very large client that I can't share publicly. Yeah. you can just give us a sound bite. Yeah. So what hap what ended up happening is we uploaded millions of customers. We turned on new customer only. The campaign traffic dropped by 60% and Google said, yeah, that's fine. It's only going after the new. You did have a lot of returning in there. Ran it for four days, and we found out that our divisional split between new and returning was just halved, but they didn't skew one way or the other. What should have happened is like 50 new 50 returning, it should just been like 50 new. It went from 50 new 50 returning to 25 new, 25 returning. It just simply reduced a match error. It couldn't match the customer list. I don't know. I've tried it four times in four different client accounts. Same thing happened every single time. I actually have right now a campaign inside of Google where I have excluded 4.1 million customers and Google says it's working. And Google says, because I've excluded those as much as I can through some campaigns, I will only do 90% negative because it won't let me do a hundred percent cuz it's just not possible size standard shopping. But it said last week, I only had eight returning customers. And then, In the back head of the client's website, I add about 900 returning customers. So Google says you upload your customer list, you negated it out and it's working, and then you check the actual data and it doesn't work at all. I had a one week period before and after where the week after I added all those customers and negated them. I actually got a little bit more. We're returning customers so, we actually,, it's an open ticket with Google right now. ticket I know, I know it's an open ticket, but that's what's so funny too. This is the part in the sitcom where you're like, will you flash to their support ticket q and it's just like some idiot sitting there on the paddle ball, like just not care. It's like monkeys with typewriters like My paper's flying around. Monkeys jump around. Yeah. It's like, oh, it's getting fixed. We're looking into it now. yeah, so, but that's, here's what's crazy though. So I got 60 K in returning. I made a 3.48 on my returning already. My cost per returning customer still cost me $26, and then I had 660 returning. So Google said, Hey, I'm getting 400 new 600 returning. So that's the danger. Try to scale this. Yeah, you just buy all your own customers back. Right. Well let me ask you this, at a certain point, especially given, you know, the market size for this particular prospect, for instance, is, is narrower than your average apparel brand. Right? Cause they have to like specific bands. Yep. If you scale past Google's ability to saturate your current customer list, would it then default back to cold? You think you could outrun it? Yeah, I you can. It just takes a long time and actually that's expensive. Yeah. And actually sometimes you want this, like this is still profit. I mean that cash is coming in. That's okay. Yeah. Well, esp I mean, dude, there's some brands that actually have a difficult time recapturing the return customers, you know, if you're in a hyper competitive space, consumable, absolutely typical. Like this would make a lot of sense for this particular brand though. You're like, eh, I probably would've got most of those. Yeah. Well, that's the thing too, this, we're like, we're pretty much the only ones in this space, like really high. There's a couple little knockoffs, but they're, I mean, they're really, horrible. so we own this space. we're the household brand in there because of this issue though, this particular client, we're moving into standard shopping next week. we're moving outta pmax as we're gonna see how this works. and so that was interesting is p m. It's, it's good and it's bad. This is not a bad scenario. I can run this. I haven't, I have been running this forever, uh, and it works really, really well. If I look at this performance Max here, if I look at all time, I mean, not forever. I only started it whenever first quarter 2022 came out. Um, so I've been running it for, you know, over a year or about a year, and it's working great. I, I just am stuck. I guess I would say I am stuck here. You'll see this like, I can't ever scale out of this. Like I, I've had ceiling as if you're not a, I'm not thinking about Google Ad Managers right now. I'm thinking about clients. Like if you are a person who owns an e-commerce store, your Google ad team could be saying, yeah, 500 row eyes. you're welcome. Mm-hmm. and it'd be really hard for you to manage your way out of that. Yeah. Or Well, and if you're an ads manager too, the other part that kind of stinks is you say, Hey, I got a 500 row. They're like, great. How much more money could we put into it? It's like, oh, nothing. No, we're tapped out. Yeah, we're tapped out. My work here is done. I cash my check now. Please. we've captured. Oh yeah. So some extreme cases that we've actually seen this go really bad is we have a client that sells, very, very expensive bulk items, b2b, and they have a lot of returning traffic. Returning customers will come back and the first time order is like under a hundred. The second order is like four grand. So what happens ends up happening on performance Max was as soon as those people start to return, I would see spend spike during those times. So it would know that these people are gonna make reorders. And my spender would go from a thousand dollars a day to like $3,000 a day and I make like 20 grand. And that would go dormant until the, and then looking at the back end of Nor Beam, we saw that, oh, that's when the returning customers would come in. So the row I looked at, as soon as those guys are online and Google's like, I am going to melt your face. Right? Yeah. And so any which way, shape, and form that they can take attribution for something which isn't bad, but that's what it does. that's what the system's trying to. So any which way, shape, form, we can take attribution for something, it will, and then it will start to show good roaz. And that's where people are just like, well, if Roaz is good, everything's good to a degree. Yeah. But is it scalable? Not really. When six like that, the climate, we're saying six out of 10 sales are returning. Okay. were you gonna get those if you didn't go after em with Pmax? I don't know. We're gonna find out. But yeah, that's, I can't scale it. So that's the good and bad is it will hunt. It will find everybody. It worked. It worked for a year and it still works well right now that cash is sitting in their bank. Did I overpay for about half of those? A little bit, probably. Yeah. And that's the bad part. So all that to be said is, I think that's a learning lesson for pmax. The good news is you don't have to really, hound these people on email. Google will do it for you. They'll find them and they'll make them. That's so funny. They'll make them come back, they'll find new, but the limitations from that is. it's hard to scale. sometimes you can pull back, you can pull back ad spend and just simply get a better roas. The sales will still come in because they're gonna be coming in anyway. and that's some of the things that we kind of saw is we actually made a slight pullback. I said, Hey, I'm gonna just gonna start to drop cost because I don't need to spend this much. so check this out. And here's where things get a little bit funny with Pax. So a week of January 23, I dropped. here it was, I think, yeah, right here you'll see it was 1300 a day. I dropped it down to 700 a day. So I have my, ad spent and that was on the 23rd. So I'll go the 24th compared to the 24th to like the 11th that compared to the previous period. and what you'll see, Is because there is a lot of returning. You can't scale, but you don't have to spend as much. I spent 39%, less, and I made 64% more sales. 42% down and 57% more sales. Well, that's kind of kind of weird. Like you shouldn't be able to do that with the campaign. And that's when, when happened. We also have what, a 29 day conversion lag too, right? Wouldn't that factor into this? Well, this is, uh, a little bit, yeah, this is conversion by top time. Ah, I just can't get you I just, yeah. The conversion's here. this one right here, this is the conversions. So not by time that went 2% down. I, I took 42% of my cost out and I made 2% less sales. Why? Well, half of those are gonna happen anyway. Bro, you know who you are. You're, you're Marissa Toman, my cousin Vinny You remember Marissa Toman? My Oh yeah. Yeah, bro. Just, just a killer. Just I could not be got right? It's got psy, right? The, the, the psy track in the rear end. It's the possible question. No, it's not possible to happen. I love that movie. I gotta watch that movie again. But that's, that's what's kind of interesting. So, and in the, in the, in the areas where you can pull back and not lose sales, you are going to push ad spend and not make more. They're just gonna kind of do whatever it wants. Mm-hmm. So that's the good news about Pmax, is you can turn it on, it will find the customers, it will get them to buy, it'll get them to return. That's good enough for like 80% of our Yeah, it really is a good set and forget it. It is for, it's quite a few folks that are like, yeah, I actually want both of those things and I'm not too worried about maximizing efficiency. I'd rather know I'm getting my return customers anyway. Great. Go do it. Yeah. This is a fantastic campaign. If you're just like, Hey, I wanna sell whatever sells. I want my brand to be top placement. I want cold traffic. I want them to return. Yeah. I mean, it is a great, great, great tool for it. Just know you'll have a lot less control and sometimes you'll overspend on it. Now when you get up to the areas where you're talking like 50 K a day, some clients we're doing 50 K a day for that doesn't fly that because they're like, Hey, it's 50 grand. I need a cost per acquisition of a new customer under X. And so. You can't do that. so that, that's the area where I say is good and bad, but for 80% of the advertisers out there, this is great. This is gonna be amazing. You're, you're gonna even make your repeat sales and you can, you can have a very healthy business for a long time. So, yeah, that's kind of our structure that I say is it works the best. Well diversified asset groups. Use your own audience as. Sparingly as you can. We only have, I think, six out of 60 that use our data, and that's by design. Um, it's not something that we, uh, we have a whole bunch of people. I do want those people to return. That's why we did this. And it just got a little bit too much returning. But it just go down here and then you'll see how it's, it's not, Not using all of the, our data. So sometimes it's not, not even hitting all best practices, we're just diversifying everything that we possibly could. So we have everything repeated, so hoodies to a converters, hoodies to custom hoodies to interest demo. Like you'll see. It's all, it's all diversified. But we found out, that this model does work pretty well, well diversified if I was to do this one over again. What I would probably do is stick just all of my converted or all of my signals, into one big signal, like a really, really, really big signal without any sort of the, my data. And I just replicate that with asset groups each time. that'd be another test. So hyper segmentation from a product perspective. No segmentation from a signal perspective. Yeah, just lump everything into one. That's what kind of seems to be doing. Yeah. You're basically just saying like, What's the 10 BMA search terms? It's like, well, this one's only gonna have one BMA search term. That one's gonna have a different broma search terms. No time. We spent segmenting audiences too. You know what I used to say? Like you take all your product categories, you take all your audiences, and then you multiply 'em. That's how many asset groups you get and like that. If that segmentation was all worthless, like just blah. Well, there should never be a time Google shouldn't do this. This is the bad part, is there should never be a time where I have a like Grateful Dead golf shirt. I have five asset groups that match up to that. There's 44 here. What this is, Google's telling you, Hey, this search term and these search terms, we wrote everything you told. Right. And they're right right It's like, Hey, you sell Grateful Dead Products. Well, did you know that your Grateful Dead Golf shirt and your Grateful Dead Socks and your Grateful Dead Hoodies and you're Grateful Dead Hats and you're Grateful Dead Shoes, that's why that matches. Oh, basically golf shirts. Right? Right. It's a Grateful Dead golf shirt. We just ignored the golf shirt. They said, grateful Dead, that matches all your products. And it's like, so, it's so broad and so vague and so, uncontrollable that, yeah, I would just stick every one of my signals into every single of my master group and. Let the people sort it out. Like click on the products you want my search campaigns, let the people decide. choose your destiny. It's like going back to like, you know, Zelda, it's like choose your destiny. Like where do you want to go? Like just, just pretty much just non-controllable. It's really interesting. Two early nineties references. We've had Zelda and my cousin Vinny, we're getting old. We're talking about dead. Have a midlife crisis right here on YouTube. Yeah, it's awesome. I'm going bald. You're going gray. consumer's gonna be disgruntled old men. It's like he's whipper snappers. You don't know what it's like to have standard shopping Monday. You had to set your own manual bids and you changed them daily. No, this max I convers. I was maximizing my conversion That was it. We did it. Pex. Yeah. Yeah. Thank you for watching the Ultimate Guide to Google Ads Part five. If you haven't already watched parts 1, 2, 3, and four, that probably means you can't count Go back, check those out. Part six is coming. It's more of a wrap everything up, some of the little. Things on the, the, the mortar to the bricks. How would you Yeah, we're just gonna kind of fill in some gaps. We did a lot. Um, we talked about like, you know, standard shopping search and did remarketing and YouTube and pmax. And so we're gonna kind of have a video just kinda like wraps everything up. Um, so I think that this one may have gotten you kind of all set off in an area where it's like ohoh new and returning now wide. So we're just gonna do like a little bit of a recap, basically how to structure a setup that is going to give you the most well-rounded growth. Stay tuned for that. In the meantime, please see you later.