Facebook Ad Interest Audience Testing Strategy For Getting The Best Results

Key Takeaways

incognita odor has an AdBlocker or something but I made two sales a $14 ad spent $27 each and the day after I made four sales three of them fired the pixel and those were five dollars and eight cents per purchased twenty seven dollar products and then what I did was on this campaign I turned off the two ads that weren't performing as well okay and I duplicated it with an identical budget thinking that that was better than just doubling the budget and making a significant change to make it have to be optimized and then for some reason yesterday they did way worse so wait a second yesterday you got zero but you did you say that you had purchases it actually wasn't zero it was two but they didn't fire the pixel I'm so technically how do we know if it's still way worse what's up what if it well I mean like what if those two purchases did come from this then you're still positive ROI yeah that's true but I would have thought that doubling bad spend and then turning off the ads that weren't doing well would have resulted in way more of an ROI yeah it would have but like you know like these things don't happen within like an hour of doing so you have to give it a few days for it to just oh really because I thought that it only had to go into re optimization again if you had to put it back in review but turning off at ads doesn't do that go to your ads on that one here are more requests to access if I can yeah sure oh they should be good yeah anything I have to do it in my system preferences why you know click unlock in the bottom-left yes okay cool well that's the only one purchases anyways right so that's why I turned off the other ones yeah so like you know that's fine but I thought that wouldn't make it have to be optimized it's not gonna redo anything because like you're not even out of this you're not even out of the optimization phase anyways because you haven't gotten 50 conversions oh really you know okay see you think it's you don't know yeah you're not on the new business manager yet so like on the new one you'll tell you if it's in learning limited it will always stand learning limits it if you haven't gotten 50 conversions or if you yeah or or if it's been longer than seven days so like you only have three purchases that's not you know that's nothing significant to be able to base any data off of within a day so like you know no purchases today or yesterday but the thing is you said you're you're getting purchases and you're almost positive it's coming from this because there's not really another source that I should be coming from yeah I mean it's possible that yesterday's sales came from the one right above it which is a duplication but I forget why I thought it wasn't from that I think I looked at the number of link licks when the when the sales happened and it was significantly less so I just assumed it was the other one okay all right let's check your pixel real quick John because it could be depending on how you have your purchase conversion set up it's down what is your site on leadpages you know so you need those on so you can match people better see that these information your customers have provided to your business like email address phone number to match your website visitors to people on Facebook if they spoke basically can't match them it's gonna have a harder time basically attributing better for you wow that's way that those two thoughts of being off yeah so we just turn those on for you huh and not only like it's recommended to turn this on but we'll leave it off for now because I think yeah your event should be fine yeah that's the out of it so let's see okay yeah so that's that should be good to go okay sweet and then switch over to the ads manager back to that you know why someone wouldn't fire the pixel is it kinda what I said because it seems like 50% of them don't why it's not firing yeah it could just be delayed attribution the pixel might not be the purchase conversion might not be set up correctly they might have saw the ad and googled you and like went through Google and like purchased so it's going to tribute it to last clicking from Google interesting yeah so like there's cases like that that's why it's important to be you know mindful of those situations because if you like completely judge this off of like oh no purchases well it's like hona and like we got like three purchases yesterday but we you know where would those have came from because the only place were running ads from its Facebook and we got a ton of traffic from there yesterday anyways mmm like what is the like what order how likely is it depending on like you know somebody would make a purchase outside of this like our people naturally finding you as often as you know you I think there's I haven't done any sort of SEO optimization or anything exactly so like it's just if there's no other way of really knowing if somebody's gonna come from another direction and you just know you're not building up other flows right now then like it's gonna be a Facebook thing okay you know so like how many purchases have you gotten overall I think six okay that's about 50% all right yeah we'll leave that running I would turn this one off red yeah there's the only gun purchases it looks like yeah it's the exact same thing that's a copy oh then you know definitely turn it off oh really because I thought you're supposed to duplicate it instead of like doing a big change like doubling the ad spend because that's what I kind of do no turn that off and just increase this by 10 to 15 percent daily okay and then you need more ad sets let's go in this yeah create some more ad sets of different audiences to test okay what is this audience I was testing another audience at one point um oh Jesus what no yeah that's crazy all right I would duplicate this and start testing different interests but only do one interest for each okay oh really you know and then be fine done yeah let's choose one yeah interesting yeah because then then you can know where the purchases are coming from because I saw like 60 different artists in that thing and it's like you know where did those three purchases come from what if they all came from one artist mmm then you're spending money on all those other ones okay and you don't even know where the purchase came from yeah the group so like break it out like it hit duplicate and then add one new interest you know each time can you define goodbye what keep on clicking it again real quick to edit yeah so you would literally so you would literally take each of those interests and then just put it on new ad set yeah five dollars each how come so literally every single last five dollars you don't have to do every single on of these just choose the like the vet choose the ones that you know that would be the best okay gotcha like just based off your experience which one do you think is most likely and then just break it out by that would you leave something like the engaged shopper behavior now only one interest for each okay interesting okay yeah that makes sense mm-hmm because I want Spotify you know like those those are big one SoundCloud like the known stuff I'll rat them out and then like you know what I mean you'd know better in this industry just break them out by individual then set it to five dollars daily and if it doesn't get a purchase for I'd say if it doesn't get a purchase by how much is your how much does it cost to order that 27 all right if it doesn't get a purchase by the $15 mark turn it off okay so like if you have an ad set running and it has no purchases up until 15 turn it off and then if you're getting if you're finding ads that started getting like five dollar purchases then anything that doesn't get a five dollar purchase turn it off mmm okay I hope you would set the the budget for the ad set not the campaign you can do it on the campaign level necessarily put five dollars into that no yeah so like you can break it out by ad set five dollars each and then like whenever you have like three or four audiences that are actually like producing like good cost results like five dollars per purchase or you know that are consistent ten dollars whatever the case is whatever ends up being consistent they're getting good results just put those in a CBL together so that way you basically the campaign budget okay how about you just go into a CDO so that way you have like your best ad sets and a campaign budget optimization and then your testing will just be done on the outside level hmm yeah I've never done that before so if I go into the ad set I can actually set a budget for the ad set yeah honestly just stick to CBO just stick the CBO in like just for every ad set you add just add five bucks to it okay so right now at 15 I could do like three exactly okay I'll probably boost it to like oh he said let's get more than like 10% no you can take that well if you're gonna launch a bunch of ad sets you can go up as many as you need don't worry it's not gonna like make it finicky or anything no did well it's gonna be testing so it's gonna be all over the place anyways your goal is just to basically find out which one's working okay gotcha dad that's crazy cuz I thought it was working type in CBO all right I got it yeah you know what what you could do is this CBO proven because if this is working to you and it's positive ROI it is working yeah you're right and then just to placate this and then you'll launch all of your new ad sets into this one which is going to be your CBO testing however and just put all your individual ad sets in this one that's a good idea is it and now launch all your testing ad sets in here is it normal to have campaigns fluctuate pretty significantly yeah literally stay guys yeah oh god okay then you should see a big fluctuation in cost okay cool whether the algorithm just does its job yeah exactly but I mean like you can micromanage it as well like if you know some ad sets that are like spending money and not getting purchases by the $15 March's turn them off okay you know save save the save the testing yeah I've trouble just like waiting for it to do its thing mm-hmm so just if it reaches 15 without no purchases turn it off turn off each individual ad set that hits that number is that yeah that makes sense it makes sense but I'm also wondering like is it optimizing within that 15 dollar period like it would work but if you waited till I hit like 20 dollars or something then you do you know perfect dude you're already watching this one I was just about to tell you to come watch this oh yeah yeah yeah so like this is exactly this is a good advice in this one oh god yeah watch the whole thing oh you've seen this entire one no I haven't I'm going to I'm like but you see how these interests are broken out one by one mm-hm exactly oh I see oh that's right he did that yeah okay sweet I had today with a Facebook representative via chat just about like the learning the learning phase I'm pretty sure that robots actually I don't know if you know do you know if it's actually people who are typing in an answer everybody has a different access to support we don't have the East chat we have like actual like person we get to talk to right yeah but someone like me do you know if I'm talking to a person or a robot I don't know I wouldn't be able to tell you I mean it's possible they're getting really sophisticated systems now so those things are pretty accurate yeah you know they're coming off is like they're passing the little I forget what that the test is called the Turing test or something hmm where it's like you know it can be perceived as an actual person because it's so accurate now yeah right right there's things that seem legitimate and they pick like an AA generic name like Wayne I think is who I was talking to today oh yeah I have the craziest names too it just seems like a name I haven't heard in a while right right yeah I don't think I know a way nin real life exactly yeah nobody using names like that that I'm just like I've never ran into somebody with that name he was telling me though that um he's I mean it could just be the but not really understanding what I'm trying to ask but he was saying that we can actually load up the conversations I don't want to get what is your point where was your question I was asking what the learning phase and basically if turning off ads within my campaign will put the campaign back in the learning phase there's no it's campaigns if I turn off ads within the campaign cuz this this at one point had no three unless you significantly shut off a ton of ads so this wouldn't do that no okay cuz that's like he was saying to Wayne the the robot mm-hmm but he was also saying that not every campaign even goes into the learning phase and he says even though the algorithm is always learning it's not necessarily going into the quote unquote learning phase and he said it would show me in the delivery if it did you're not on the new manager so it won't so I took that is me not having ah you don't have access to that feature yet having the learning face yeah you just don't have the learning phase feature yet that'd be nice like that yeah you'll get in cuz this isn't that an advantage that all my competition has against me oh I don't know I mean it could be it depends on how you use it and how it's you know I mean it's just it might be a slight advantage but I mean like really the advantage it's gonna be like how good you're setting up you're targeting in your ads and all that your product and all those things yeah true cool okay sweet so I just let this run pretty much more launch for testing audiences so what do I do with Cpl proven spends 15 dollars today and doesn't get any sales check yourselves let see if it does on your actual store well yeah I'm always monitoring that but if it doesn't get any sales on Facebook or on the actual store and it goes up to $15 but it's already made sick sales what would you do then I'm not anticipating that happening I'm just curious the thing is what happening as this thing is going through and testing and it hit one audience interest that you had and boom I got three purchases then it went on to the next interest to see if it was worth you know you sent spending budget on and it got nothing and then it went on to the next one for today and it got nothing because it goes through audience clusters that's why you like to break it out by interest so that way if it's not working you turn it off because you know you're not getting anything out of that period holy I did not know that yeah so like on the back end it would be too complicated if they showed a visual of that nobody would ever understand it but that's the test the clusters within each audience you have the watchers that will go through spend money on that cluster and then we'll go to the next one and basically once it's tested all of them and it find it out which one it's going to start optimizing towards that's when it exits the learning phase and it starts putting all of the spin towards the one that was getting the best purchases but when that point happens usually everybody would have turned off their ad set by that time right right but it will eventually hit that point is what you're saying yeah it should and if it doesn't like turn it off like I mean like but if it doesn't get another if it goes two days in a row and you don't see anything on yours with sales and yeah turn off that's why you need to like you can't Bank on wine like you really think you film like go a winner like the very first day I guess pause yeah I mean like you should launch like you know a ton of ad sets to find out what's gonna work okay yeah Dan that's crazy because I I've hired Facebook Ads coaches in the past and that's how they told me to do it like make the ads that's super complicated like that and having all these exclusions and no way man okay I would never do that yeah then there you go it makes a lot more sense to me it's just like putting fog on your window you can't even see what's going on in there yeah that's right I didn't realize that they are with the algorithm tested in clusters yeah that's you so you can learn what clusters yeah it goes like think about it this way everybody who's interested in Spotify not all those people were alike there's going to be different clusters within that interest of people there's gonna be your daily listeners there's gonna be your mom-and-pop shops that have you know that are playing the music when people are getting there of course yeah there's gonna be you know artists in there there's gonna be just weekly listeners I'd you know like to test each little group of them and once it finds out which one is gonna be the line and zeros in hmm okay yeah that makes sense okay