how to a/b test facebook ads

Facebook Ads A/B Testing

many of you are probably familiar with,the google ads campaign experiments and,all that they can do either because,you've used them or because you've seen,our video on them if neither of those,apply to you and you're interested in,learning more about them you can check,out this video right here,but not as many people know about the,facebook ads campaign experiments and,all that they can do they've been around,for quite a while but over the past,couple of years there have been a change,in how you set up campaign experiments,on facebook ads so in this video i want,to talk about what you can test with,facebook campaign experiments and then,we'll run through an example of how you,can set one up in your account,there are three different ways that you,can set up campaign experiments within,facebook ads and it all depends on what,you already have existing in the account,you can create just a slight variance on,a campaign or ad set that you already,have active or you can set something up,completely from scratch if you don't,have anything existing that you want to,test just yet for this video we're going,to go in the reverse order of what i,just said so we're going to start off,with an example that we don't have,anything set up yet that we want to test,so let's go ahead and create a new,campaign and then this account for some,reason has the six simplified campaign,objectives so for now just to keep,things easy we're gonna click traffic,and we're gonna click continue over the,course of this setup i'm gonna be pretty,quick about skimming through all the,different settings all that sort of,thing but the one that you cannot miss,at the campaign level is going to be to,opt into this create a b test you'll see,that when i hit the toggle we have a,little window that popped up that says,this campaign will be version a in your,a b test after you publish this campaign,you'll be prompted to edit a duplicate,version to test against it so you have,to publish your campaign before you can,then duplicate a version of it to run in,your a b test you cannot just leave a,campaign in draft and set up the,experiment just an fyi that you'll have,to publish to create the a b test so,let's go ahead and click next there will,be a number of different variables that,we get to test for facebook a b tests,for this specific example i'm going to,show you how to test a different,audience rather than go through all the,different settings on this ad set tab,all i'm going to do is come down here,and customize the audience just a little,bit i'm going to add in a couple,detailed targeting options that are,about ppc that sort of thing,overall that's good enough for me so,we're setting up to target search engine,marketing and digital marketing within,this first ad set that we have now that,i have my audience set up i'm going to,click next and just so we can keep,things moving i'm going to use an,existing post that we have from our,facebook page rather than creating,something new from scratch so let's just,go ahead and use the latest video and,we'll click continue now that we have,this campaign set up all ready to go,it's got a proper ad set proper ad we're,going to click publish now we need to,get started with the a b test portion of,the setup so here you can see it says,that we can test different image,audiences and other settings to see how,the impact the performance of the ad so,we're just going to click create a b,test now the language on this page is a,little bit misleading start by setting,up an ad to test against the one you,selected we can decide either to make a,copy of this ad or pick another existing,ad they keep using the word add when,what they probably really mean is,variable or campaign or something like,that just trust that this will set up,what we want it to for now i'm going to,say make a copy of this ad which is the,default rather than pick an existing ad,we'll talk about that in a little bit so,for now i'm just going to click next and,now we get to pick the variable that we,want to use for this new version to test,again it keeps saying add but it'll,encompass the entire campaign so let's,choose from this drop down and as i,mentioned earlier there are a number of,different variables you can test,creative audience placement or a custom,variable but for this example i'm going,to use audience you're then prompted to,choose which ad set you want to copy so,new traffic ad set is the one i set up,just a little bit ago that's the one i,want to use because that's the only one,in this campaign but if you had,published multiple different ad sets you,could open this drop down and choose the,one that you wanted to use that'll be,even more important when we get into the,other setups of existing audiences and,ad sets but for now let's go ahead and,click next and in this portion of the,test builder all we're doing is giving,this test a name and then setting up,different settings around this test,itself we're not updating the audience,just yet so first thing we would do here,is want to give this a name that would,describe to us what we're trying to test,in this experiment again i'm going to,skip over most of these for right now,but then the next thing we need to do is,tell facebook how we want to determine a,winner what are the metrics that we're,looking at to try and optimize on for,this traffic campaign experiment we get,to choose from a number of different,metrics cost per result which is going,to be linked clicks is recommended which,is a pretty good metric to try and,improve upon or we can use cost per link,click we can use cost per thousand,people reached or cpm cost per purchase,or you can choose from a number of,different standard events that are down,here so whichever metric you're trying,to improve upon choose that one from,this drop down if you are in a,conversions campaign you might say that,you want your cost per lead or in a lead,generation campaign you might want your,cost per lead form to go down whichever,metric you're trying to impact choose,that from this key metric drop down,you'll see that this toggle here is,switched on that says include upper,funnel metrics in this test report if we,click on this little informational piece,here it's basically saying that metrics,can help gauge user activities that,precede the metric that you're,optimizing for it gives an example and,says a test may not find a winner using,cost per purchase as a key metric but if,the test was able to find a top,performer based on cost per link click,you could use that information to make,decisions in my opinion there's no harm,in utilizing this function so i'm just,going to leave it on for now you can,also add additional metrics if you want,to add in different pieces into your,experiment you get up to nine additional,metrics to include within the report so,if there are other things that you want,to see alongside the cost per result,maybe we decide that we do want to see,that cost per lead something along those,lines you can include that in here as,well if you then decide for whatever,reason that you change your mind you,don't want to include that before you,hit duplicate you can just click the x,and it goes away the last piece that we,have in terms of the settings is going,to be this test duration we can tell,facebook how long we want this test to,run now choosing the right test duration,is mostly going to be a function of,volume and how much data you want to get,and i say that because there is this,little metric down here that's estimated,test power basically what this is saying,that at 65 the way i have the test,currently set up with all the different,settings and the different options,there's only really a 65 chance that,we're going to improve upon the metrics,that we have the longer the test runs or,the more specific the test is or the,different metrics you have the better,chance your test is going to end up,yielding a better result if i come up,here and change this day's duration to,something longer let's say 14 days,you'll notice that the estimated test,power went up to 71 percent because in,theory we are going to have more data to,work with over a period of time that is,longer and with more data leads to a,better chance that we are going to have,confidence in a test the last thing we,get to do is choose whether or not we,want to end the test early if a winner,is found effectively facebook is going,to run this test for the duration that,we have it in place the 14 days which is,currently set at unless it decides that,it has a high confidence in one of the,different audiences based on the cost,per result that we're trying to hit,overall this is one of those things i,will leave up to you as the advertiser i,personally will always turn these off,because i want my test to run for the,full 14 days or however long i have it,in place but i know a number of people,that when you run a lot of tests or you,have a lot of things going on it can,help to have some automation in there to,end your test when a winner is found,once you have all the settings in place,you just need to click duplicate ad set,now you can see in the breadcrumb over,here that we are still in the new,campaign that i use to create this,campaign experiment but we have the a,version which is new traffic ad set and,then we have the b version or the copy,of it which is the new traffic ad set,dash copy which isn't super helpful so,i'll just name this one version b so now,it might be a little easier to look at,at this point we would just scroll down,make the changes to the audience that we,wanted to have in place so let's say my,test is that i really only wanted to see,if detailed targeting expansion works so,i'll check the box here now my targeting,options are the two interest that i had,originally and detailed targeting,expansion so let's see how that works,now that everything is set up i would,just click publish,now back on the campaigns tab you can,see that we have the new traffic,campaign set up and it now has this,little icon next to it that talks about,it being in a test it's got that little,chemistry bottle next to it that's an,easy way to identify when something is,set up as a test if i click into it we,can see that we've got the two ad sets,the first one that doesn't have a,version name and then the second one,that has version b and then if i go over,to the ads tab you can see that they,both have the same ad in place,as this test progresses we can see,performance of the different audiences,at the ad set level if we want to go,there or we can head into the experiment,section of the interface which is going,to be heading up here into all tools and,then i already have it up here in the,shortcuts but if you scroll down the,experiments are going to be in the,analyze and report section under,experiments,the home page of the experiments tab is,going to launch you onto tests where you,can set up new tests and we'll get to,that in a second but to see the tests,that you currently have running you need,to click results and then here you'll,see scheduled i have the a b test that,starts on seven nine it is an audience,test and then i can view the progress of,this test,since i just set this up and i'm going,to turn it off here in a minute because,i don't actually want it to run because,it's a terrible example you'll be able,to see all of the different metrics that,are available we've got all the,different insights around the key,metrics the objective the type of test,what's running and then there will be,results down here that actually show you,the performance of a test but for right,now i'm just going to abandon this one,and let's move on to the second type of,setup the second way we can set up an a,b test in facebook is by duplicating an,existing variant within the account and,utilizing it as the a version and the,new duplicated version as the b option,so let's head into this other stage,traffic campaign and let's say that i,wanted to test against this us18 plus ad,set the easiest way to set this up is,going to be to come up here and click,duplicate and then under select a,campaign for your ad set you just click,new a b test then let's go to continue,to test setup that drops us into one of,the later stages of the a b test setup,process where we get to pick the,variable we want to change for the new,version of our ad if we choose from the,drop down the options are going to be,the same creative audience and placement,and the setup process for each of these,is going to be very similar to what we,did for audience basically we would go,in we would make an adjustment to the,audience itself and then have it set up,that way for the sake of trying to make,this as impactful a video as i can let's,see what it looks like when you do the,custom option because effectively all,you get is to choose which ad set you,want to duplicate and then you get to,make all of the changes that you want,you're not opting into any specific type,of test whether it's creative placement,or audience you basically get to do,whatever you want change anything to,your heart's content to set up your,experiment so here i've chosen the ad,set that i want to duplicate and then i,just get to click next i would then go,through set up all of the same types of,settings that i would want around the,key metrics the test duration all that,good stuff so right now i'm just going,to go ahead and click duplicate ad set,you can see in the upper right portion,here that it shows that this is in the a,b test it tells you the ad sets that are,in this test and it gives you the,information around the test itself the,test name the test schedule the key,metrics all that good stuff if you want,to change any of those settings after,the fact you can click edit test,settings right here but then with this,custom a b test setup you could change,any of the other variables that you,wanted to in this ad set and then once,you were ready all you had to do is,click publish and you would then have an,a b test running between this original,us18 plus and the copy version of it,with all the different changes that you,had in place so if you already have an,ad set or a campaign running that you,want to use as the a version of your,test this would be the way to do it,there is one final option which is,assuming that you already have,everything built out in your account,that you want to test against each other,you can set up a campaign experiment,with existing assets already but we have,to do it from the experiments tab so i'm,going to head back in there i told you,we'd come back to this page we're back,on the test section and then for this,option we just come over to the a b test,and click get started this option is a,bit different because again everything,is already existing that we are testing,against each other we can decide if we,want to choose campaign groups campaigns,or ad sets i'm pretty sure you know what,campaigns and ad sets are but campaign,groups are a little bit different let's,go ahead and click that one and here,you'll see that the campaign groups,option basically lets you group,campaigns it's pretty aptly named using,campaign groups would be a good idea if,you have a series of campaigns that has,similar settings similar objectives that,you want to test against another group,of campaigns that has a group of similar,settings and objectives maybe you want,to test a bunch of your video views,campaigns against your traffic campaigns,and see which ones have the better cost,per link click or perhaps you have a,number of different campaigns set up,based on different targeting types and,you want to use an experiment to see if,lookalike audiences are performing,better than your interest audiences,whichever one makes sense to you but,i'll be honest for myself i'm probably,almost always going to stick to the,campaigns or ad sets so in this instance,let's just assume that we want to test,two different campaigns against each,other i would then come down and choose,campaign a which is going to be from the,account that we had set up already i,know i have these names blurred out i'm,sorry for that but i'll just choose the,ones that have the most generic names so,let's say we want to choose the new,traffic campaign which says it's in,review and then as the b version we want,to choose that stage traffic campaign,that i mentioned earlier there are going,to be some errors here which will tell,you some of the things that you need to,know when you're setting these up,obviously the first is that your,campaigns need to be active so the fact,that i have one that's in review it,doesn't like that so much but let's,assume that you have all your campaigns,active so we'll close that then the,other pieces says your attribution,settings do not match to continue,setting up your test make sure all ad,sets in your test use the same,attribution setting just as a note you,cannot set up a test amongst different,campaigns that have ad sets with,different attribution settings but once,you had all those issues resolved you,would then come down here and set up the,remainder of the test so you get to have,the schedule we can schedule it to start,ahead of time or effectively right away,we then get to determine if it has an,end date so you kind of get to determine,the duration but you don't get to tell,it duration you just a start and end,date you can still opt in to end the,test early if a winner is found give it,a test name you get to choose the key,metric and you do still get the option,to choose up to nine additional metrics,as well and then at the bottom it'll,give you the estimated test power of all,the different campaign settings and test,settings that you have in place i went,ahead and made a quick tweak to it just,so i can go ahead and review the test,details,effectively you get to review all the,different settings you have here you see,that it does give you the duration in,bold available at four days because,that's where i set it up at you get to,see all the other options that are in,here and then lastly you would just,click create test once you were done i'm,not going to do that because i don't,want to run this test but that's what,you would do and then again any and all,tests that you have running would show,up in the results portion of the,experiments tab within facebook setting,up a b tests across different ad sets or,campaigns can be a really powerful tool,in facebook to get you data backed by,statistics that tells you which of your,creatives audiences placements or any,custom combination thereof can perform,best across your account for the key,metrics that you have in place they're,relatively easy to set up and it doesn't,matter if you want to test things that,are already existing in your account or,if you're trying to create something,from scratch there are different options,for how you can set those up hopefully,with this rundown you feel a lot more,confident in going in and setting up,your next a b test on facebook but if,you have any questions feel free to,leave us a note in the comments below,thanks for watching our video if you,thought it was useful give us a thumbs,up below we release a new video at least,once a week so if you want to get,notified of when a new one comes out be,sure to subscribe to the paid media pros,channel

The above is a brief introduction to how to a/b test facebook ads

Let's move on to the first section of how to a/b test facebook ads