our optimization schedule is now three week sprints. Don't touch it for three weeks stack all of your changes at once. Don't touch it again for two to three weeks stack all of your changes at once and people that are thinking with their PAX campaign, they're just resetting, learning over and over again. They're never getting outta learning phase. Welcome to your daily Google news. My name's Kasim and I'm joined with our strategy team, most of our strategy team anyway and we're making this a weekly recurring theme where we bring in the best of the best, our best and our brightest. And we're gonna talk about topics that should be near and dear to your heart. And today we're going to throw our skunks on the table. We're gonna open the Kumo. And talk about every failing performance max campaign we've had and maybe more to the point and more productively, why we think they've failed and what we can do in the future to help mitigate that. And so a little context that I'll provide before we dive into the nitty gritty is performance. Max is new. They've been testing it since 2019, but they rolled it out publicly. It's this new thing that we have to contend with and Google's education. What they tell us to do is wrong. And so we're, seeing insane performance in a lot of instances, but we're also seeing some really. Troubling surprises and some difficulties, some things that are hard to contend with. And so we're gonna talk about those and what I'd love for everybody on this call to think about is not just like what failed, because sometimes it's fun to play the crash and burn game and just tell those stories, but more constructively. If somebody's watching this video, how can we protect them from making our mistakes? And, we were to build a beware of list, Which maybe I'll start to do, actually, while we're talking about this, let's build our beware of list. And by the end of it, maybe we can have something that people could actually have and take with them. So who wants to go first? Who has a crash and burn? Just fire and brimstone PAX. I'll start with just one. Don't run any PAX for any Legion. We have not been able to make it successful in two Legion accounts, but I will share why we feel that and why in one account it's working. So the two accounts that we were running PAX bar, they don't proper EO and, their website is not categorized in Google or proper keyword. So as we know, Google PAX works or advertise on seven channels, the channel is search channel. And it has a part of DSA. So PAX also runs on DSA. So dynamic Sur ads. If your website is not coming on proper or relevant keyboards, then PAX will not work for you. Because it'll go after keyword that your website is working on in organic EO or on that Google has categorized you for. So, that's why my two Legion account fail because the website was not on Google search console. They haven't categorized it or they haven't indexed. Properly. And then it worked on one because they have good se authority like do domain authority. Their website is ranking for good keywords. So the one thing that we use to check if the website is properly crawled or properly categorizes, simply put the website URL in keyword resource tool and just put the URL and you will get to know what keywords Google is, categorizing your website for. So if all the keyboards are relevant, then. It's not a hundred percent, but the PAX campaign will work for your account. And just to clarify, keyword research tool is in Google ads. You just go to tools and settings and you can find it there under keyword planning. Oh, Carl, that's a value bomb from hell right there. So if you're watching this go into Google ads, look at the Google's keyword research tool drop in your URL. And if the search terms, Google comes back and says, Hey, we're gonna advertise for this. If those are relevant, you're in good shape. If those aren't relevant, then you've got some work to do. And that's four seconds. Like there's no excuse not to do that. It's almost like seeing how the algorithm thinks about your website. Like yeah. Seeing how it sees that's exactly right. What would you do in situations where they are relevant, but all you're getting is bot traffic. That's what I was kinda going for. How'd you avoid all the bots on Carl? What's your secret? Caden? I have a thing that I'm testing right now and I will let you guys know. So we're using John's like team fields. Yeah, just ton of fields. So I'm doing that. And then I'm also parenting with a timer. So I'm seeing like how long people spend going through all the fields. And so. The theory is that the more fields you have, there's going to be a bigger discrepancy between someone that fills out something just by going through real quick and, spaming it. And someone that really puts time and effort through filling it out. That's genius. So you've got a lapse time added to what the conversion tracking. Yeah. To the conversion action for the lead forms. For example, if it's a page or a popup, you just have the click ID or the URL, and then just attach the timer to that. You can create that in talking to Yvonne about that. Actually you can create it in GTM. Yeah. How much time are you giving them per form field? Have we figured that out? So I'm testing 10 seconds 12 seconds, 15 seconds, 20 seconds and 30 seconds. So if they fill it out too quickly, are they going to a separate thank you page. That's just not tracked as a conversion. So not yet. Initially trying to figure out where the Delta is. So I don't know, for example, it takes someone and it's gonna vary based on every client. Right. I don't know if it's gonna take 'em like 10 seconds to fill out a form and then it takes a bot maybe like five seconds or further and further. And so the theory is the more questions you have, the easier it is to see the difference between like, spam traffic versus not spam traffic. That is absolutely brilliant. I don't think we're contending with bot traffic. I think we're contending with click farm traffic. That's for click farms. Yeah. Right. But click farms to your point, those people are gonna be way faster at the farm fields because they don't care about the inputs. So the note here is ADAPs time to the conversion action. And then before you decide what the line demarcation. Monitor it so that you can get a census to the, the relative average. And I'm assuming that it's actually not a bell curve in this instance, meaning it's not distributed evenly. It's probably gonna be heavily weighted to early stage traffic. Find out where that spike is. And then you'd put your elapsed time right on the right on what would we call that? What's the decline called? Any statisticians I just call it the Delta. But right. When it starts to slope down, that's where I would put my line demarcation. And you'd split the thank you pages and only give conversion credit to the latter. Thank you page. You could do that or just take off the conversion action or just put it as well, I guess you wouldn't wanna put a secondary, there could be reasons not to do that, but you could in theory, just take the conversion action off and have the timer. Just be the filter as well. But then people still form fill and you don't want that bad form fill. So you probably want to keep. The forms that are relevant, that you deemed relevant based off your conversion action, and then send the other ones to a non-existent and you paid like how we got, I wanna fight you though on that in at least early stage, because we don't know that our assumption is right. So I'd want to keep the form so that I could follow up and see like, okay. That is a better, that is a click form, as opposed. And then maybe, I guess what you're saying with some is end result is we would send them to the gotcha. Sand trap. Yeah. I see what you're saying. I had a question and I lost. It was gonna be brilliant too. It was gonna blow you all away. So I really like that, Katie, I we've got two leg gen accounts that I know of that I'm personally involved with that are working inside of PMAX and both of them have just wicked long forms. There's Southwest building solutions whose name we can use. And then there's a niche real estate lead generation. They're crushing life, but they have Southwest building solutions has 17 form fields. It's an aggressive form. It's very specific to. So I'm just gonna say use massive forms, caution, large cost relief forms. Yes. I would say it's easier though, for phone calls when it comes to PAX, just in general. It's a lot easier to catalyze but we also track record and score all phone calls. Most people don't do that. Exactly. Yeah. But it'll just stop bot phone fills. Right. It'll not stop bot traffic, like right. Bot traffic will come and Google will charges for that flick, but when you're optimized conversions, it would eventually push that away though. It'll we those out, but yeah. It's crazy how much a Google's display based traffic is either bought or click form. Like it's UN it's an open air conspiracy at this point, as far as I'm concerned, like, it feels like it's most of the traffic really we're getting charged for this crap. And then you come to find out that it's like a bunch of people with 50 iPhones lined up in front of them, all filling up forms professionally rolling their IP address. Okay. We've got PAX fill points. Lead gen is super dangerous. Your website has to be SEO. Go put your URL instead of Google's keyword research tool to see what Google thinks you're relevant for use AAPS time. Attach that to conversion action. Cadence. I almost wanna edit it outta this video. We won't though. We'll leave it cuz we're good people., why else did performance max campaigns fail? What else have we seen? Maybe that's surprising too. I know not enough asset groups, but I don't wanna do the easy rote and routine ones. People should maybe already know that. Well, not necessarily not enough asset groups. I've seen a fill where. The asset group doesn't support your budget. That's what I'll say. I audit accounts that come through for new prospects. And it's kind of crazy when someone's like, oh, we're spending, $60 a day and I have 99 asset groups and this is an impossible question to answer, but do we have a cost per asset group roughly? We said there are no SOS in there. We did it. We would have to industry is different. It's impossible because yeah, the industry, the geography, I think FA U responded saying it depends on your cost of traffic. Like if your CPC is fraction of a penny, then you can get away with anything. Oh, I was gonna say average order value. Like I keep imagining like a X, Y axis where we could say budget versus average order value and kind of decide like how many asset. Because average order value kind of indicates how frequently you're gonna get conversions, I think, is that right? I don't think so. I think AOV is independent of cost, cuz you can have super high AOVs and super low cost. Those are the industries that everybody wants. And I'm thinking about, remember our tape client, I guess they had a high LTB. But I know what you're saying, Regina, there's gotta be something that is kind of the predictive indicator. I would have to sit down and just make this and then iterate it for weeks. Yeah. Well, but I mean, to that point, I think that there could be like a Google sheet that we build out saying like, what's the cost of traffic. How competitive is the industry? What's your ceiling too. And then how much do you plan on scaling and those things added together. And we could guess roughly guess at what your budget needs to be per asset group mm-hmm. It's weird because we tell people, build as many asset groups as you possibly can, which is good advice until it's like, well, now you have to spend more money to test all these asset groups. And so the advice and the silo build as many asset groups as you possibly can is really good advice until now we're talking about budget too, and now we're like, well, you have to be able to fund all those asset groups. And now the logical question is, well, how much money do I need to fund the asset groups? And the answer is I Don. I think the way you wanna think about it is if each asset group can get about a conversion or two a week then it's gonna have the patterns and the frequency and the ability to go after remarketing for the person that it's following and all of that, you know what I mean? Like you want it to get frequent conversions. Two conversions a week, for example, would be good cause the data gets old. So, if each asset group only has enough money to get a conversion every three weeks. It's not frequent enough for it to be able to follow someone around, have the remarketing budget available when that person is back and they're searching and then it's not gonna have the patterns because the weather's gonna change or the competitor is gonna change. And then, three weeks is not frequent enough. So I think that's why I think about average order value, because you can kind of take your budget and divide it by. No. Okay. You're right. It's cost per conversion. Not average or valid. but I like what you're saying, the cost per conversion. You'd have to run first and test it, but once you've identified your cost per conversion and I like your line demarcation at two per week, I think that that's actually pretty valid. It's a valid minimum. Hopefully it's more, but it's a valid minimum because at two per week, you should start to see optimization adequate optimization for an early stage campaign. So let's say using round numbers that your cost per conversion is $10. Let's say, if you need two per week, that's 20 per asset group. And if you have 50 asset groups and that's easy to cascade across. Exactly. Yeah. That's as good a napkin test, as I can think of, we need kind of a guideline. Like there will be exceptions like that tape client, but I think we should build that and see how it goes. So I'm gonna ask the other strategists. I think the logic here's really sound. How do y'all feel about the number two, two conversions per week in an asset group? Should that be more, less? What's the number of asset groups you have cause overall your campaign is gonna be performing based off overall campaign performance, the calculator would spit out how much asset groups you can have. So you put in your budget, you put in your cost per lead, and then all it does is it divides the budget by the cost per lead and tells you how many asset groups you can have because you need each asset group to get two conversions a week. Well, it's very chicken and egg though, because now what we're telling people is build all the asset groups you can build. And our current SOP, which we're about to train people on is take all your product categories. Let's say you have five, take all your audiences. Let's say you have five and then multiply them. And so I have five product categories and five audiences I'll have 25 asset groups. So now if I have 25 asset groups to the point that Regina's making we'd wanna see. How many conversions per asset group we would want in order to properly optimize the campaign. But I guess what ends up happening though, is you end up with a Pretto distribution inside of your asset groups, where 20% of your asset groups are getting 80% of your conversions. So you would wanna arm your asset groups with enough funds to achieve at least two conversions per asset group, but also with the understanding that most of the asset groups aren't going to, but you'd wanna take money off of the table as to why they didn't you'd wanna have enough money there for them to achieve that. And then once you've reached that critical mass from a spend threshold perspective, when an asset group fails, you at least know it failed. Not because it wasn't well funded. Yeah. I think that's really sound logic there. Regina, I wanna back you on that. My only question again is whether or it's two conversions per week, one five, like, how do we feel? Two sounds good to me, by the way. I'm not fighting you. I just wanna see what the other strategies think. I don't know about that, cuz I'm also thinking about the CPC. So let's I'm for an example, let's say my average cost per acquisition is $70. So I need two converge. You're saying for an asset group, I need $140 a day, right. For a week, all right. So $20 for an asset group. So if I have five, it'll be 100. I don't know. That's a tough project to build upon are too low. I just base it off the total number of convergence. I get over like, 30 days. What to that point though, no matter what it is, something that we're backing into, it's not something that we're telling the client up front. Like, oh, you have to spend this much for this asset group. It's build it, run it. And then we will tell you generally speaking what we think you need to spend in order to, you wanna be able to cover all of your asset groups with enough spend. It's a weird paradigm to assume, because you actually wanna spend enough to let things. Because 80% of your asset groups are gonna fail. 80% of your ads are gonna fail. 80% of your offer are gonna fail, but you don't know what's gonna fail until you spend enough so we need to spend enough money to let things fail. I wanna say Google does not only optimize for conversion. So for an example, you can have two ads. One will have more conversions and better conversion rates, but two has better CTR. You can see that Google can sometimes spend more money on B. Even though the first one, it has a better conversion rate. So it doesn't only optimize for conversion. In my cases, I have two accounts for this example. We checked our asset groups. We checked the products we have. A hero product and let's say it was spending 50% of its budget and the conversion rate was fine. It was around two not to three, four but the role as was two and the client called street. So our hero product was spending half of our budget. And since the conversion rates are 3%, 4% of Google is thinking, Hey, I'm, I'm killing. So I'm gonna put more money into it. I'm gonna spend half of it. I'm gonna spend 60% the budget, but at the end of the month, you'll see that you're not paying your goal and actually you are losing money. So you have to be careful check your listing groups frequently and see if somebody is taking too much. You don't wanna have false positive., check your listing groups frequently. And what are we looking for there with listing groups specifically? We're looking for overspend like over commitments, categorically how much money did this asset group spend in shopping? Or you can also check your products and sent by clicks. So you'll see that let's say you spent a thousand dollars sometimes you'll see that 500 of those. Went to a single product and it has good conversion rate or bads. You have to be careful with that. What I did was I pulled that out. I created its own performance max campaign and gave it a budget that is reasonable. So let's say my daily spend was $200. It was the anything, two X. So I said, I'm gonna spend $50 in. So a I'll have value and the client will be happy that I'm still advertising for the hero product, but I have to spend 115 elsewhere and get that four X to have to meet the goal by the end of the month. So I love this FA as an SOP, this is really sharp. So if you have a product. That is monopolizing spend, we still wanna market that product. The general rule 80% rule is split that product out into its own PAX campaign. So you can dedicate, spend to that campaign. Now I'm gonna throw some grenades at you. I agree with what you're saying and I like it, but I also wanna stress test this a little bit. You'd have to make sure that product's conversion value. is actually the sale of that product, right? Because oftentimes somebody coming in on that product is buying other products. And if you take that product out of that PAX campaign, you kill that cross pollination capability. So in order for this SOP to be true, that product's wh when people are clicking on that product that have to be buying that product, because for those watching in case you don't know if somebody clicks on that product and buys another product, the product they click on still gets the Roaz credit. And if you took that product out of the, the PAX campaign, Then you'd kill that kind of incestuous relationship. It's a risk you have to take because like I say, Google is not going to optimize only for your conversion. So let's say USL dresses and shoes. Let's say dresses have 100,000 search volumes and shoes have 10,000. Since this will. A good CTR and a better value. You may not give a chance to law search items. So you have to be careful. Yes. Cain's point is right. If I remember correctly number of was. 65%, 35% of the time. Somebody's going to click on an ad and buy something else. So that's your odds, but 35% of the time. Is that your anecdotal experience? No. I think it was either HubSpot or hood suite, dude. That's great to know. Thank you for that. So which by the way, would make sense based on what I've seen, just poking around campaigns. There is a cross pollination thing that happens that advertisers don't realize happens. Everybody thinks it's oh, click and purchase people are dumb lemings and they click on an ad and they buy the thing and it's like, no, they don't do that at all. Ever. Like they click, they view, they watch, they bounce, they leave, they come back, they tell a friend, they use their wife's phone and then maybe they buy the other thing. And to Google's point. Yeah. I heard that when you click on performance, max, a, the click will not be shared with performance max B, but since as an agency, we don't believe anything. Google says but I still have thoughts about that. Cause the machine is smart and the AI is still learning. It happened in smart shopping where the smart shopping campaigns would not cross pollinate. And we learned that the hard way we advise people to split smart shopping campaigns and that advice hurt campaigns. So I'm inclined to believe them in this instance, just because I think that whatever mechanism they're using for cross pollination, I think if it appears to exist in both of those campaign types, Yeah, my thinking was just this, if performance max, a got to click and said, all right it got into the website. It looks at different stuff, but I tried this in remarketing is not clicking. So I'm done with this user. A B will not gonna follow that. I don't know. I see what you're saying. Here's a question is performance max doesn't remarket. Well at all. Anyway, so if I have performance max campaign, a performance max campaign B and a dynamic remarketing campaign, will that cover both of them agnostically? I'm assuming yes, right? Yeah. I have new metrics now, so it's more data and you're still getting that G clips, but. It's 80% new. So once team the smart shopping was 80% remarking. So it will take a long time. Smart shopping says, Hey, I'm done with you, right? I'm done chasing you. What about performance? Max is 20% returning users. So it may be done with a client with the user more quickly and B can follow it up. So it's just a theory. I cannot back it up with data, if some of your item is spending too much money and doesn't get the enough ROS, it's not hitting your goal and you are a small business, you don't have the luxury to wait around like three months, four months, five months, six months to let Google learn. You have to do something. People pick out what we say. A lot of our clients come to me and say, Hey, John said this, why aren't you doing this? And I wanna say, Hey, if you watch it fully, John also says, Hey, I'm spending $5,000 on this PX campaign, right. You're spending a hundred. You said something else FA that I want to keen on as far as why performance max campaigns fail is you have to leave em alone. Don't touch it. Yeah, build it. Let it run for six weeks. Unencumbered. And then even in, in an optimization, our optimization schedule is now three week sprints. Don't touch it for three weeks stack all of your changes at once. Don't touch it again for two to three weeks stack all of your changes at once and people that are thinking with their PAX campaign, they're just resetting, learning over and over again. They're never getting outta learning phase. Yeah. That's the other point? A lot of things changes in the agency like before it was more about specialists. Now it's more about client managers. They have to call the clients like, Hey, we're doing something just because you don't see any changes in the changes. History doesn't mean that we're not working, but a lot of the people think that you didn't change anything. You just build it. You just sit down. Well, I've actually said this on a sales call recently, and I think I'm gonna make it part of my standard narrative. As it is to optimize something, right? just because you're not pressing a button doesn't mean we still have to check. We have to open the account. We have to check the campaign. We have to monitor the data. You still need to go in and look, you're just the end of the day where we used to like, oh, I'd add a negative keyword or I'd make a better adjustment or whatever. Yeah. Now we're not doing that. We're casing all of that to come for the sprint. And when you case a change. You have to do an internal, what I would call an impact report. How does this change stack up against the 10 other changes that I plan on? So it didn't get easier. It's not like, oh, we get to optimize our time and make all of our changes. Every three weeks. We're not waiting for three open the hood and be like, all right, I'm gonna do these 15 things. Let's look at this, monitor it, decide what it is we're gonna do. It's so much harder because now there's a multi-variate problem. It's not like I get to make a change and see the impact, it's come up with every improvement that needs to happen on this campaign. And then. Think through how each of those improvements is gonna tie in with each other and themselves, this horribly complex web it's more difficult than it's ever been. And if people are paying us to make changes, go, there's so many software applications, PPC, samurai. We used to use PPC samurai. I've got friends, close friends who I love dearly by the way. So I'm not bashing them. They build in. Changes intentionally to show a high change history because they're tired of having this conversation with clients. So we could do that exact same thing I could automate arbitrary and worthless changes. Just so change history looks really robust. I could show you 500 changes a day. The changes don't matter at all. So if you're trying to hire somebody to make changes, go hire a nephew, put 'em through a, a weekend, Google ADSD course. He'll make all the changes you want. Like God bless you. I'm rooting for you. That got way more aggressive than I meant for it to be. Who else has reasons PAX campaigns fail? Yeah, I have one. No okay, so we are doing our job, right? We have one job to structure Google page, right. So we are doing our part right. We are not tracking how the user is behaving when he goes on the website. So my thought process is there is one user converts. It's 50% like we can give credit 50% goes to Google ads and then 50% goes to the website. But in most of our clients doesn't track or doesn't watch how user behaves on their website. So so right now, what I'm doing with my Legion clients or some E eco clients is that I'm trying or I'm convincing decline to install Microsoft clarity. So what Microsoft clarity does, it's a free software from Microsoft, which get. Scroll map, I think then recordings and then click tracking. So once a user lands on the site, it tracks the user behavior. It's important because. We get more data. Like where users leaving like on web stage? Let's say we have seen 30% of our visitors or our users when to add to card and then they leave the page. So we can go on, add to card and check what's the problem, like where people are leaving. What's the drop of and we can try to optimize that step or else. Using a landing page so we can check where people are. So where are the heat maps? Where they are spending their time, like on which part of this website. Then we can work on that, like our CTAs schedule free appointment. We will not get that data in Google ads or Google analytics, but in Microsoft clarity or other hate map softwares, we will get that data. And then we can work on improving our CTA. We can change the CT and we can, again, check what's the heat map for that section or for that CT. And then we can work on that. So, I think we should. And most of our clients do that, but we should track our user behavior, user behavior on our website. It's funny because Google ads, it used to be so bottom of the funnel, all you had to do was track conversions, and then you knew what was happening because performance max is full funnel. You actually have to track the efficacy of the funnel, which is something that Google ads agencies aren't used to doing. What the other thing I think is interesting is Zhar, you're one of the smartest people I've ever known. I truly respect and value you, but the fact that you're pushing a Microsoft product, man, it just makes me wonder, like, if you've crossed over to there's something wrong or it is free. That's why I'm promoting I think that's a really good point. I think that people tracking the efficacy of their funnel because performance max, if it dies on the entry level steps, and you're just tracking inside of Google ads, you're never gonna know. Yeah. And I've seen a lot of website. So that is one Rule we should follow. When we are like promoting a landing page or a website, we should optimize our above the fold. So above the fold means the section that you land on first or section that you see first on the website. So you should always optimize, or you should always match. The ad copy that you are using in the ad and what is on the landing page? So let's say on ad copy, we are saying 15% off on chainsaw and on website, we are on above the fold. We are not promoting it there is nothing mentioning 15%. On chainsaw, then that's a mismatch, right? So I think people don't look at it. It's important to match the value proposition in ads and on the landing. Can we push UTM parameters per asset group? No. Cause that would be cool. Cuz then you could just do dynamic text insertion, grab the UTM and then modify the promo by asset group. Why man, that doesn't make any damn sense to me. Why can't we push UTM by asset? That feels like an easy that's. I mean they already have mechanism problems. You can't see asset group performance. You can see the channel source for the convergence. You don't even know, where is your money going before that you have so many things to have. I wonder if we couldn't hack it, like if we used a really special string of characters in every headline description and I guess text for images doesn't apply. So just, but hear me through use a special string of characters for every headliner description of Emory. That gets pushed through on the UTM parameters, cuz you can include ads and UTM, as an include file. And then on page, you now capture that special line and then proxy it saying, oh it's whatever line it is. So this is the 15% off. Asset group, it's a wacky ass roundabout, but I think we could make it look if we wanted to. And actually what we could do there is we could start to see which asset group we could see. Asset group performance. You could build your own internal analytics tool around asset group performance. It would work way better with search and DSA. It probably wouldn't work for display. How are display assets working DSA? And this is you cannot define helpline. So you would only have the search, but you can define descriptions. Yeah. But I guess the description doesn't come through the UTM is what you're saying. So you're right. I'm wrong. I don't know. I don't know either if somebody's watching this who's way better at analytics than we are. Let us know if this is something that anyway yeah, I think this was great. Last words to the strategist. How do y'all feel PAX? There's a bunch of now here's the thing is there's a bunch of, we're trying to show you our PAX fill points. That doesn't mean don't use it. Like I still think it's the most powerful marketing mechanism I've ever seen in my entire life. It's like when you give a toddler a super soaker, you stuff's gonna get wet. You know what I mean? You have to know how to use it. So last words to the strategists going around, any, any closing words for our listeners or viewers. I was gonna say whatever you do, don't talk about. One product or service in one headline and a different product or service in a description because the algorithm is not great at matching up headlines and descriptions to the right product and service. That's a rookie mistake that we made in the early stages of PEX. Yeah. It's an easy one to make too. I think everybody who launches a new PAX campaign, they're just, it's just hard to get acclimated to how flexible all of those things are. Yeah. You wanna think about. What's being searched for, and then only talk about those things in that asset group. So unfortunately you do have to have asset groups that are divided by your products or product categories or services. Well said, phone, car, closing words yeah. Work on ad copy because performance max is full funneled, right? Doesn't mean that you have to work on ad copy. If you just use bottom of the funnel ad copy like product specific. If you write just product specific ad copy, it'll not give unaware audience. They don't know what you sell or what product you have. You can create classic groups or else try to. Combine headlines description and long headlines cater to all the audience, like a never audience, product aware audience, and then bottom of the funnel audience. That's a great note. What's the old cliche features tell benefit sell. Well said FA last word. Okay. It's it's mid June. You have 15 days whether you like it or not, PAX would be the only option you'll have. So, yeah, especially Freecom well, local too, right? PAX is replacing local. Is it in July or is it later? August. Okay. So smart. Shopping's going away. All local campaigns are going away and performance max, do you think they'll take DSA? Hope not, I don't know. I don't know the industry. Like I think they're gonna take away everything. So we should definitely be ready. The keyless future. Cool. Y'all did a great job. If you're watching this, we shoot a video every day, light comment, subscribe, and we'll see you tomorrow.