Creative Testing Campaigns May Be The Biggest Cost Center In Your Business transcript

[Music] a few months ago I released an episode of this show about my creative testing method for meta ads it is one of the most popular episodes I have released people continue to return to it and continue to argue with me about what I proposed there it is seems to have scratched somewhat of an itch amongst media buyers both in terms of positively helping some of them construct the method I've gotten some really good feedback about it I've also had people be really frustrated by it and suggest there's all kinds of reasons why it is wrong that episode which is linked in the show notes if you have not listened to or watched that episode you may want to go back and do that for this one lays out a positive case that what you want to do is launch all creative in a bid cap or a cost cap or minimum rows campaign some kind of manual bid and you ought to launch that creative against all of your other Evergreen creative you should not build a creative testing campaign in fact that's what the thumbnail of the episode says that's the title on it is don't build cre testing campaign that's the heart of my argument that basic basically what you ought to do is launch new creative against all of your old existing creative and you should not worry about forcing meta to spend money on new tests that's the heart of the argument I have heard since releasing that and have continued to hear lots and lots of push back on that from people who are frustrated for all kinds of reasons I'm going to get into those in this episode and what I want to do is take on the most common counterarguments to my position that I hear and explain to you why I think it is the case that creative testing campaign s for meta ads media buyers are probably some of the biggest cost centers in businesses and DDC businesses out there that maybe there is nowhere else where people are wasting more money in their business just generating a or creating massive drain on their profits and in Creative testing campaigns and I'm going to tell you why the most common arguments I hear in favor of that approach to create testing are wrong and lay out for you the case in the next few minutes so hello and welcome to the Andre FIS podcast thanks so much for joining me I hope this conversation is something this episode is something that makes sense to you that you are in the conversation enough to know what I'm talking about and to be locked in on this I think if you are if you've been in media buying for a little while and have been running creative testing campaigns it will be some use to you to think through the math of why creative testing campaigns just don't work well and why there's a much better way let's get into it all right so let me just recap really briefly my position here what I do when I launch new creative is I launch it within a cost within a bid cap for my Brands you can do it with a cost cap or minimum r as well I prefer bidc caps I won't get into that here I launched them in New adets in CBO campaigns against all of my evergreen creative and then I walk away and if meta spends on my new creative it's a winner if it doesn't it's a loser that's the basic idea of my approach to creative testing that by submitting my new creative to the to the machine learning of meta ads I am able to this is the massive advantage of it right I am able to suppress the bad ads and amplify the good ads without wasting money on bad creative this is the core idea that essentially if I'm going to run auto bid creative testing campaigns to to like quote unquote see which creative works then what I am forced to do in that case is spend a bunch of money on ads that by definition many of which will underperform nobody hits their creative testing perfect all the time like even the great creative people they're going to launch ads that sometimes are not going to work as well as the other ads they're already running and so you're going to spend if you launch an auto bid campaign I don't think anybody really disagrees with this you're going to spend some money on losing ads and my suggestion is you're probably going to spend a lot of money on losing ads especially if like me you care a lot about creative testing at scale in real volume if you are regularly testing new creative new Landers new off offers anything like that that can meaningfully change the performance of your account then if you're going to launch lots and lots of new creative then every creative test you launch if you were GNA Auto bid on that represents cost represents some cost now of course there's some sense in which every creative that you launch every piece of creative that you launch represents cost for everybody right that the cre has to be produced and created Etc and and this is a thing that people sometimes are frustrated with my method for because if you have run my method of creative testing before you have had the frustrating experience including I have had this of launching new creative that you are excited about and meta spends almost no money on it okay because there's a massive campaign with a bunch of adsets in it and meta just doesn't think that this new creative is going to perform as well as other creative and it spends very very little money this is a fairly common occurrence it doesn't doesn't really ever spend zero dollars but it will spend a couple dollars a day sometimes even less as it does not believe quote unquote it's a machine doesn't believe anything but and it doesn't forecast I should say that add to win and so if you've put real money and time and effort into generating new creative this can be a very frustrating thing can be really disheartening and you Al often will want to think okay I've got to give a fair shot to this new creative to see if it will win and this is this that experience is I think one of the things that really frustrates people I think the other thing that really frustrates people and this is like the core thing that I really hear is meta ads there actually there two two main things that I hear people say about this the first is if you do that you are not giving your new creative a quote unquote fair chance at getting real spend the logic is you actually may have a winner there but because meta has all this engagement and all this signal and all this social proof on your previous winning ads because of that Those ads come in with an advantage and therefore meta will not spend on your new ads which have none of those things has no pixel data on those ads it has no engagement signals on those ads and because of that because it doesn't have any of that information Those ads essentially would win if they got those things but they don't get those things to start and therefore they can't compete with these ads that already kind of have all that stuff built up that's part of the logic and proof of this is often explained by or proof of this is often trotted out by this experience I tried launching my new creative the way you said Andrew and meta didn't spend on it but then when I gave it a fair shot build some Auto bids spend against it Etc it became one of my top spending pieces of creative in my ad account over a significant period of time what do you say about that okay now I need to say something upfront here which is that I don't have access to your ad account so I don't know if you actually did my way of doing it right I don't know the rest of your account is a mess I don't know what your optimization method was I don't know what your measurement was I don't know a lot of things about those things so I can't speak to the specific circumstance that people bring up but I hear that story a lot I hear it all the time and somewhere in there too what I hear people also saying is something the effect of you can't learn what is working or not working if you're not getting data back on your ads and very often meta will suppress your new spend so much that it won't work and let's just let's just remember again to come back to the original Point here what I'm suggesting is that auto bidding creative testing campaigns forcing meta to spend lots and lots of money on your ads is generating a massive cost center in the business and in my view I think literally one of the biggest cost centers in businesses now what most people will say is who are running DC Brands who are running meta ads they'll say we're willing to pay that cost if on the back side of that it generates winners and so that's fine you're right we are paying money for that Andrew except that we're willing to do that because if that's what it costs for me to make sure to identify a bunch of winners and I can I can get those now those are going to push my business so far forward that that cost is going to be relatively low and that's just the cost of doing business and what I am saying is you can get all of that benefit of pushing winners forward in your ad account and yet completely reduce that cost in fact I'll take it a step further I think that you are very likely missing out on ads that are winners by doing it that way so not only are you paying the cost but you're actually not getting the thing that you're aiming at because and here's the core of my argument almost any methodology of creative testing that relies on auto bidding and then identifying winners and manually scaling them yourself is going to be so fraught with metal methodological messiness that it will actually be counterproductive so not only will it be cost costly but it will actually be counterproductive in the process of generating winners so so let me say also from the out outset that I do not believe meta's machine learning is perfect I do not believe that meta always gets the winner right and always suppresses losers correctly sometimes I watched it happen recently meta will spend too much on an ad that is actually a loser and there's a number of reasons that we do that sometimes I'm certain that meta will not spend on an ad that could win in my ad account I have no trouble believing that but the Baseline what I have to work from is not the idea of perfection but of probability I have to work from probability that I am getting the best thing in the most circumstances and if that's the standard and I believe that it is then ultimately I think my method works really really well of launching things I think it is going to on average give you a much higher probability of selecting your winners than than a manual method okay so that's the Baseline contention on the average in the aggregate I'm going to identify more winners faster which is why my method is built around the idea it's not really my method right it's like me it's just the what I do so I'm gonna call it my method but it's it's not just mine but it's the reason why I think that what I see happen all the time is that Brands as they switch over from Auto bids to manual bids and they test creativ this way they see increases in spend because what I think is that meta by by meta giving meta the freedom to scale your winners and suppress your losers the Blended outcome of the total account gets better there's two ways to win there's one way to win is to spend more on your best ads another way to win is to spend less on your worst ads both of those generate a better return than uh it's not just about spending more on your best ad so by accomplishing both of those and I think meta actually do both better than an auto auto bidding method then ends up giving you the best efficiency in your ad account and ultimately will push your spend forward and push your growth the growth of your business forward which is the thing you actually care about okay that's the basic concept that I'm getting out here and it's why over and over again I see Brands who uh test this way generate uh significantly more ad spend than they were before it's why I think media buying matters it's why I think media buying works and serves creative development and makes it so that all of your creative efforts get better and this is ultimately what I want for you right if you are listening or watching I want you as you go develop your best creative ideas to consistently identify the winners and get those all of your have all of your hard work and developing creative and offers and and great products and all those things have that all pay off as best as possible so that's the idea okay so I want to take on especially this idea of auto bidding your campaign so the first point I want to I want to make is that and this is crucial many people who suggest that you ought to autobid your creative tests I my my first response to them when they say that is okay how much do you spend per ad put another way how do you know when you have a winner okay how do you know when you have a winner and that I think is a question that that people sometimes answer in a way that is really really weak frankly you know you need great people to join your e-commerce team your e-commerce business and you can find them in the Philippines with my friends at more Staffing by going to more now.co more Staffing is a recruiting agency that helps you recruit train onboard coach continue to provide support to incredible talent in the Philippines for your e-commerce business people with deep e-commerce resumes don't just go look for a $5 an hour virtual assistant hire a virtual professional somebody who is trained who knows what they're doing and can bring tons of highlevel value to your business and we'll do that at a highly competitive salary in the Philippines but a salary that will be less than what you would pay somebody of equal experience in the US that allows you to attract incredible Talent bring them onto your team and you will do well because of it keep your offex lean it is a win for everybody across the board if you're not considering your next hire in the Philippines you need to and I am telling you I love working with more Staffing I have built my team in part with more Staffing I have four employees from the Philippines and an intern as well now and I love them they're awesome they provide a ton of value to me and to my clients and so I'm happy to recommend them to you they're a great partner to me have been working with for a while and I trust them so much go to more now.co to begin considering if you aren't considering it you need to at least consider it begin considering hiring incredible talent for your business this from the Philippines that's more now.co go do it right [Music] now I recently heard somebody suggest that well yeah I mean I've heard of I recently heard of a brand that spent a couple hundred bucks on a new ad and would say every ad we're gonna spend a couple hundred dollars on and we're GNA see if it gets enough purchases to show it's a winner at that spend there there's just no way in the world that you can tell if you have a winner or a loser within $200 an ad spend basically no matter what your average order value is ETC so that we'll just we'll just assume that kind of method a couple hundred a couple purchases that's maybe not the way that you are doing it I would instead in the spirit of thinking carefully and critically about things I want to instead take this on by taking on what I think is the best version of this argument and that is an argument it was voiced by actually by Zack stuck on Twitter Zack is the co-founder or founder of of Homestead the the marketing agency but Zach said he spends 90% of his time growing his own Brands now Zack is a smart guy who's built really good businesses and in my opinion is one of the better thinkers around Twitter at developing particularly the notion of like sort of direct response oriented great creative and Landers he's done some really good stuff there yeah so I have nothing bad to say about Zach's abilities here or something like that but I wanted to also not subtweet him in this particular comment here I just wanted to mention it because because Zack had voiced this opinion on Twitter in a way that I thought was relatively compelling which is that he was saying that for his Brands he is testing to statistical significance every creative he launches okay and I think that's the best version of this that many people are going to say we want statistical significance right stat Sig that in a in a test to say option A is better than option b and therefore is a winner okay and I think there are probably some advertisers out there who have some level of rigor doing this pretty well but what I want to point out and I I don't know I mean at this point I just mentioned Zack because I didn't want to sound like I was implying something about somebody when I was really talking about Zach or whatever but Zach had voice this in a way that was compelling like I said so so I don't know exactly how he does this all right but I think that is almost impossible to actually do this well and to actually do it without it being a massive amount of cost and I'll tell you why I think that's the case so what I'm going to do is I'm going to share my screen I've actually done the math to begin to think about how this would have to how how this would have to play out okay so let's pull up this sheet so I've begun to think about what the scenario would look like to try to run statistical significance tests on every piece of creative you run if you are launching a new creative and you want to test it relative to statistical significance if that's the standard you're going to hold it to then that means you need 95% certainty that your ad is go or that ad a is going to beat ad B that ad a is is actually that much better than ad B okay and so I began to think about what the math would have to be to make this work and so I used some round numbers here I actually asked I put out a poll on Twitter to ask what Brand's aovs were 25% of people said their their aov was less than $50 you was like 50% of people said their aov was between $50 $150 and then the other 25% were above $150 okay so I took a simple rounded average 100 bucks you could play this game yourself however you want to but let's assume a $100 aov for a lot of clients and let's assume that this client spends $500 getting a $150 CPC on their spend now at a meaningful amount of spend for meta ad accounts $150 click would be a pretty cheap price but this is going to vary a lot from account to account depending on CPM creative all kinds of different things okay so a $150 CPC you would get 333 clicks on $500 and spend okay at a $100 aov if you converted eight of those purchases you would have a 1.6 Ros and a 2.4% conversion rate okay I'm just going to tell you at a $100 aov within the optimization window that you would need one day click 7day click something like that a 2.4% conversion rate on $1 aov would be pretty high and a 1.6 row ass if you got that for most brands depending on their unit economics depending on their standards again this would change for a lot of Brands but I think this is reasonable they would look at that and say this creative at least has a chance this creative at least has a chance to work maybe they actually need it to be like a two two to one but we'll give them a lower standard here and say okay if you get a 1.6 we'll move it now people who who do this grave testing method they have all kinds of ways of approaching this this question and uh when they approach it they they think like okay I I need it to just give me a two: one or one: one or 1.5 you know I need it to be at my row Target you know it could be close to my row Target whatever so I'm just going to assume 1.6 is close enough and that would be eight purchases on $500 in spend so the reason that I do all that is because what I want to show you now is how much or how little that really tells you about the performance of your ad if you have eight purchases I'm going to use Neil Patel's handy dandy AB testing calculator online what I want to show you is that if you have eight purchases on $500 in spend you have learned nothing about the total performance of this ad you've learned nothing about it you've learned almost nothing in point of fact eight purchases on 500 clicks actually oh excuse me on 333 clicks okay eight purchases on 333 clicks at that point you only have even though it is twice as good if you only have 88% certainty according to statistical significant me according to frequent statistical significant methods of testing you only have 88% certainty when you need a standard of 95% certainty that your ad is as good as getting only four purchases okay so if you only get four purchases you cannot be sure or if you only get eight if you get eight purchases on your creative testing campaign at $500 in spend 333 clicks you can't be sure you do not have statistical significance to say that the roas of that perform of that ad scaled more actually wouldn't be half of what it is in that campaign right so let's see if we take it to three we're still not there we at 94% certainty even at three purchases so I hope you see what I'm saying here what I'm saying is you actually don't really know if your ad is as good or if is better or worse than an ad that has half as many purchases on the same spend and the reason that's so significant right is that if we go play this game again now if I spend $500 four purchases at the same aov now my row is 08 and what's very likely to happen in fact we'll take that all the way down to three purchases now in either case what's very likely to happen is that if my ad at that level of spend performs at that row ass what am I going to do if I'm making the decision I'm gonna say that's a loser that's a loser and I'm G to kill it and put it away right because that ad is not performing well enough and I don't have an infinite testing budget okay and so now and this is is so crucial the whole idea that people really love is that they are not going to miss out on their best performing ads but in fact what I'm saying is because we are bad at measuring this sort of thing and this does say nothing about the fact that these tests are typically not designed in a scientific way you know where it's given a a total budget topand over a total time period Etc but because we have not because we're bad at interpreting this kind of evidence in general and because our human lizard brains get in the way and don't do a good job of being unbiased in our prediction methods what we're going to do is very likely leave ads on the table that actually could be better than ads that look on some level of small spend better than than the ones that we turned off do you see what I'm saying we're going to leave one of these ads on the table even though it actually has a good chance of being a winner we haven't reached statistical significance in saying that it's a loser and now if you think about that not only is that the case but the actual cost of doing this gets incredibly High very very fast I'm assuming that you're spending $500 on every ad at this point and allowing it to run that much but if I'm going to do that now month the cost of testing if I launch 50 ads per month the cost of testing is $25,000 a month or $300,000 a year that is a significant line item and of course larger brands are actually testing quite a bit more than this right now smaller Brands simply do not have 12, if I if I drop that to 25 ads a month they simply do not have $112,500 a month lying around they can just blow on testing okay so there this High especially for testing that is not going to work that well as a test in general right they just aren't going to have that money lying around and so and of course larger Brands right if you get to a point where you're testing a 100 ads per month now you're talking about a $600,000 a year line item where you are going to blow a bunch of that money again that's assuming that you spend $500 and you may or may not get statistical significance still in in fact you're very likely not going to you really need like overwhelmingly great performance to even get to a point where you can really trust that the results are going to be really good I mean even if you take something like let's say we get a uh 10 purchases so you get a two to1 row ass on those first 333 clicks again you actually can't be sure that you have St you don't you do not have statistical significance that you have beaten an ad getting five purchases a one to1 row ass so a two to one row ass is not necessarily better than a one to one row ass in terms of predicting this at scale in terms of predicting which as a winner you don't have statistical significance yet and so that is that is the core point that I'm getting at so I hope that makes sense to you what I'm saying is there is so little methodological rigor and it's so Frau with errors that this whole idea of auto bidding to go locate winners just doesn't work that well because yes you may find a winner that meta may not find but in the process you're also very likely to be leaving aside Los losers that meta that meta would have continued to amplify based on its way of reading engagement signals to go find Winners so people are rejecting meta's machine learning because they want to give their creative a better chance and then they're making the decision about the Performance Based on methodological nonsense that's what I think it's methological nonsense it's just a way to feel like you're at is getting some spend in some cases okay now that may be critiquing the worst version of the other position and again I want to I think there's people who are starting who are thinking about this as well as possible and it is worth noting that like in the reality of the day-to-day here like none of us have phds and we're not running scientific tests we're not academics in a lab getting to design all these things perfectly part of the thing that we should all be baking into the way we approach this is the fact that all of us including me are going to test without real methodological rigor and we shouldn't take principles from an academic lab kind of setting and try to apply them to our media buying environments because we're just never going to be able to hold ourselves to those signals on top of that famously in in academic papers that do test things with randomized controlled trials and all that like there is a famous replication crisis in a bias towards results and all these things that are making it so that even at that level many tests are not actually proving replicable and workable in the ways that we would want and so this is the big point the methodology of so much of the testing that people want to give is actually really really really flawed and people often are going to leave not only are they yes they will find some winners here and there and of course that can work but they will also have plenty of times where they're going to leave winners on the table because they just can't see it now there's another element of all this okay which is that you and this is another crucial thing that people forget you actually don't care how your creative performs on its own you don't you don't care how your creative performs what you care about is how your creative performs relative to other options and this is so crucial as well okay because like the idea of forcing spend to some creative to see how it performs on its own it assumes the idea that the the target you care about is your row ass Target but the thing is what you actually want is meta to allocate your next dollar to the best most profitable place next so you care not just how it performs on its own you care how your creative performs relative to meta's other options okay so if meta can choose between AD a and AD B and they can both beat your Ro Target but ad a can beat it by a lot more you want meta to spend on ad a in fact you want to spend on ad a and take it as far as you can until until it reaches the same threshold of performance as ad B and then you want meta to spend more additionally that my friends is precisely what campaign budget optimization is set up to do it's set up to take all of your adsets and all of your ads or ASC is supposed to do this exact same thing they're set up to do exactly that to go find the best ad next and to spend your money in that place and so I would just do all of that in fact I'll tell you you don't even have to go with me all the way to bid Caps or to cost Caps or to manual bidding in general if you just launch your new creative in an auto bid campaign and you can go ahead and set the budget yourself if you really want to and you you just launch new creative against that that will have that same effect of allowing meta to go find the best ad next as opposed to you forcing spend to the ones that you want to run it to because maybe maybe your ad does work in an auto bid campaign where you force spend to it and yet maybe when you run it against your other creative your other creative actually is going to perform better and then it suppresses it right back down if that's the case then you've gotten what you wanted you want meta to allocate your money best place next as best as it can possibly do so won't it not spend what will happen won't your ad not spend if you launch your ad against all of your other creative okay and that's possible but in the vast majority of cases and this is another crucial thing if you launch your ad against all of your other ads if meta does not spend on it there is a reason why meta is not spending on it and it is not because it doesn't have enough signal it is not because it doesn't have enough social proof yet it is because meta's machine learning does not forecast that it is likely to be a winner relative to the other options that is meta believes quote unquote it is not your best option and therefore does not spend on it that's precisely what you want meta to do and there is no weird thing where the machine learning is breaking at least in the aggregate again it may miss one here and there I totally get that but on the whole it is very likely to be getting it right and what's more this is the most overstated concern people possibly could have about this way of doing it I constantly hear this idea that meta cannot find Winners against new creative that is just not going to run them at all Etc and it doesn't really know it's gonna happen but what I am here to tell you as a person who is currently managing five ad accounts with my team a person who's coaching other people and seeing their ad accounts as well is that every day I look into ad accounts where recent new creative test ads rise to the top of this pen because we have beaten other ads I see it constantly recently I just did this today I was looking at ads that yesterday we launched they just barely started to spend a little bit as meta just started to to go because you know day of just takes takes takes a couple couple a day or two sometimes to spend and today they are rising in spend against all these other established winners in fact winners that we had recently found a few weeks before that that did the exact same thing I watch it happen over and over and over again Add accounts get pushed forward spend goes up as new Winners Get located and meta is able to locate what those winners are to identify what those winners are and to spend more on them it's just not true that new creative gets no spend it is true that most times the creative doesn't get any spend because it will not perform but it's not true that it won't get any spend at all because it has some kind of built-in disadvantage it's just not true in the midst of all of this there's another implication one of the arguments that I hear often against my position here is that uh or or sometimes against the whole exercise is theide a that I'm too focused on media buying and actually the thing that really moves the needle is creative and really moves the needle is landing pages really moves the needle is offers and I understand why people think that because I do talk a fair amount about media buying but also think that this whole thing gets at the media buying setup gets at basically the the shared hypothesis between those two positions which is that creative is actually the thing that really matters most creative is what sells your product and offers or what sell your product and so I think that this method of testing actually allows you to test drastically more creative much faster without the cost of other people do because again if you go back to my screen share a second ago if you were to what happens if you want to test 300 ads a month well now your cost per month is going to go through the roof if you're going to test them all at $500 a day now you've got a $1.8 million line item on your p&l for the year in Creative testing what happens if you want to test 500 creatives per month my friend my friends at kinship talk about this that would be a $3 million wine item 250 Grand a month right my friends at kinship talk about this when they go run influencer content all over the place and they launch them all in cost caps they say we're trying to test 500 pieces creative per month we cannot run them all in autobid tests even if we thought that was the right way to do it and so because they are so oriented around creative testing as the way to push your ad account forward they launch everything basically in the way that I'm suggesting here slightly different but basically in the way that I'm doing it all cost gaps where the goal is spend at your target if it doesn't spend it's probably not going to hit your target it does spend because it probably is and that's that and so and they're able to do this and push and push and push on massively more creative than most people I also believe that creative pushes things forward and I want to be able to test it very very fast and yet without having to bear the cost of running things in a testing campaign this is actually magical I mean it is unbelievable that there is a tool out there that so completely suppresses losing ads while scaling winners without you having to incur the cost of the testing and it's crazy it's an incredible tool and I think to walk away from it is one of the bigger mistakes you can make in running a DC brand and running paid media in a DTC brand because it's just so magical that you can learn so much at so little spend again because of ultimately what I believe meta is doing here and to wrap this up I think it is very helpful to understand what meta is and is not doing in your spend okay what I am pretty sure I'll say 90% sure meta is doing in its machine learning is essentially basian probabilistic forecasting okay and the idea is that meta isn't trying to find statistical significance on and build randomized control trials off of all of your ads and then scale results from there instead meta is using basian principles which assume that there's a bunch of prior knowledge that meta has about what does and doesn't work and then it's going to apply that prior knowledge to an updated forecast at any given time of your ad and your ad set and take all that knowledge and say here's what we here's what the machine predicts will happen next at a probabilistic level and then the machine will respond to those signals okay and that is a much faster way to test things with high probabilities and with good forecasting methodology than having to submit everything toysical significance testing okay I'm not a statistician so there's going to be some details where this might get muddy for me okay but that basic concept is the way that prediction can work at scale and why meta can do this at low cost people ask this question all the time why is it that meta seems to be able to seems to suppress or amplify ads based on so few Impressions okay and my answer is because meta is reading way more engagement signals than you or I have access to and meta also knows that engagement signals in some way or another predict conversion rate and predict all these things further down the funnel and therefore and by predict what I mean is probabilistically predict this is why it's wrong sometimes because sometimes those engagement signals perhaps don't lad actually to the actual performance of an ad but on the whole the probability is that it it mostly does and So Meta is reading those signals and saying here's what is most likely to happen next based off all those signals and based off of the insane ridiculous giant treasure Trove of data that meta has access to and when you go and auto bid your test you're saying nah no thanks I don't need all that information I will read these signals myself I will handle it myself and to I say my lizard brain is not smart enough to do that and I don't have enough money to do that and so what I want to do instead is trust that while it's possibly the case that meta will miss here and there it is a much better much faster way to test much more creative many more offers than I could otherwise do on my own and that allows me to put my time effort and money into the best opportunities to grow my business push my spend forward grow my profits and accomplish the things that I want to in my business [Music] we're at a mouthful of an episode a whole lot of stuff there trying my very best to articulate what I believe about this please give me your feedback you can email me podcast a.com you J to me on Twitter at andrej feris I'd love to hear more what you have to say about this and anything else that I've been doing again that creative testing episode I did before if you felt lost there a little bit is in the show notes you go back and refer to that as always big thanks to my sponsor more Staffing who I absolutely love working with virtual can be helpful virtual professionals can change your business go work with more Staffing by going to more now.co and tell them I sent you I've got a bunch of good episodes coming up so many good episodes coming up I've got Barry hot coming up very soon who we're going to talk about creative I've been testing some things that Barry has said for a long time and spoiler alert he's right about a good amount of it so I want to talk to him about that and kind of what he's learned about his approach to creative I think there's a bunch of stuff there that can be really helpful to you also got some other episodes in the queue that are going to be really good so don't forget to subscribe wherever you are rate review and share this episode with somebody who disagrees with me and have them come yell at me via email via Twitter whatever way you best can yell at me about what I'm wrong about thanks so much as always for watching for listening I'll see you next [Music] time