Facebook Ad Optimization Strategy 2020 | How To Fix Sudden Increases In Cost Per Results (eCommerce)

Key Takeaways

you could take a look at my ads they have been doing so good and then over the weekend they just kind of took a little turn okay that was good take a look at it yeah let's look all right I think I shared my screen right yes perfect okay so this is today's we look at yesterday's it was still kind of high okay if you want to go ahead and click on that the one with the porn teen purchases got it and then do last seven days okay okay can you go add a column that adds your return announcement in here oh yeah I always forget to do that there we go perfect okay and then can you do yesterday again sure okay hover over the control audience and then hen inspect and then du Vieux last seven days where it says that yeah then select first time impression ratio it's that fifth one down the line okay now hover over the May 3rd it's right below ok May 3rd so the first time impression ratio means that 72% of your people of the audience have seen it for the first time seen your ad for the first time ok so about 27.5% have not have already seen the ad and you can see your customer purchase went up significantly on this day if you want to go ahead and do your amount spin as well click on that one so it looks like now go ahead and remove first time compression ratio and then hover over April 30th and go day by day so on that day sorry I can't just go a little bit slower 1277 cause pers old so that's good right yeah that's great and then amount spin 266 then it goes up by $30.00 just about yeah okay twelve dollars that make sense yeah we increase the budget so any yeah okay next then fourteen so it's gone up about two more dollars and your first time impression ratio went down pretty fast okay next and then your purchases have gone out but that's May third when was that that was yesterday got it that was a Sunday so your daily on this audience might make more sense at go to April 29th okay yeah I probably good April 30th yeah so about a 160 170 170 amount spent per day would make more sense on this okay and the auction competition oh I'm gonna go back to that yeah if you don't mind let's look at one more thing at the bottom scroll down just a little bit I also think Sunday was an anomaly in this case the chart doesn't make any sense just from how graphs and data works that Sunday what you would see is you see a gradual increase okay but that is just to sleep steep of a jump that just makes no sense okay so it's probably just an anomaly if it if the cost doesn't go down today what's it out today today I know it's only half the day but alright not even yeah okay so if it doesn't go below 20 today then pull back your budget on that one specifically by about 40 about $40 okay because you might have hit like your peak point for what the total number of purchases you can get per day with that audience okay yeah that makes sense because I raised the budget on Friday okay let's do the same thing for your page view thirty days okay do the first ratio so this one's pretty clear as your first time impression ratio goes down your cost per all goes up so hyper over May first that's about the equilibrium point right there okay that's anything below seventy even on the other audience we saw seventy was like this for two should go because we saw the moment antique light goes past that point your cost per result just jumps and we see it identical right here like we know that's like the equilibrium point meaning that's like the perfect balance at that stage so about a hundred and eighty-five amounts spent anything over two hundred isn't gonna make sense for you anything number 185 is gonna be risky 180 in below your perfect on a daily spend now it just comes down to finding different audiences that can handle scale okay so how do I do that by testing more audiences okay this control audience and then just the look-alike off of that mmm how do I go into it and reduce paste you could yeah let's go to me second on this one just to be sure so it continues going down costs go up go again and then it looks like 63 costs go up so yeah you're gonna even though the audience size is big her you're likely reaching your people over and over again in that audience because of the Davis thing and the amount of people that are active at the time okay cool okay so you don't even have to go to interest audiences what you could do is do a lot of a look-alikes with higher percentages that can handle more scale okay so let's say okay go ahead and duplicate this audience into a new campaign this page view yep and then do a non CBO CBO and then it's a non campaign budget optimization and then do about one to ten percent look like alright duplicate so one percent can handle the in your case is about a 2.2 million audience size so we're going to look to essentially expand that so what you'll end up doing is you end up creating a custom audience inside this one okay hit great nail custom audience and then you'll do it which audience has been working really well that you found as far as they look alike yes the pageviews was the one that was doing the best okay let's stick to that website and then you'll do page for you Oh better yeah you already have this audience so just go ahead and X out and then do to look like yeah they're just sick right Neal look alive since when use the same line and then you'll just choose page view 30 days and then you'll do 0 to 2 and then create audience see how it's 5.1 90 now okay thanks out yep yeah and then just go yep remove that yeah there you go and then now you can name it 2% appear at the top and then you just publish this and then of that audience sizes a little bit bigger and it can handle or scale and keep that daily budget the same 207 oh no because the 207 isn't working right right gonna scale the neck up there so I'd move it down what would you want to move it down to for testing purposes um I mean I I don't mind a risk but I also don't want to be foolish yeah and well we're gonna create ten of these huh okay so what like maybe 35 40 a day okay yeah that's gonna be a piece okay and then you'll just cut them if they don't work so then you're just gonna rinse and repeat that process you'll duplicate again three duplicate I don't know what the exact number will come out to be I'm pretty sure it should be ten because you already had the 1% maybe it's 9 audiences since those are Neal he'll go 0-3 duplicate then you'll go 0 to 4 Q look it and just keep repeating okay I'll probably lower the pants a little bit so then now for this other page view do I just go ahead and move the budget down on these to that sweet spot of 180 yeah I would say the sweet spots more about 170 just to play it on the right a safe side perfect some audiences are 680 summer 100 daily summer 20 daily summer zero they never work a couple thousand daily okay for the ones that can scale I love it okay helps cuz I got a little discouraged I was like I don't know what to do but now this makes sense this is great so then I just did that I'm gonna copy all those four page views and then if there's another look-alike audience that's doing really well would you suggest doing the same thing yeah exactly okay okay very good but I can handle it awesome that works all right thank you so much tape appreciate it yeah no problem bye