The Dos & Don'ts of The FACEBOOK LEARNING PHASE

Ben Heath· 18 min· 3,723 words· 17 min read· English ·Watch on YouTube

This is the full transcript of The Dos & Don'ts of The FACEBOOK LEARNING PHASE, published on YouTube by Ben Heath. Every paragraph carries the moment it was spoken, so you can click any line to jump straight to that point in the video, search the whole thing for a word, or copy it out.

0:00hi guys it's ben heath from lead guru and in this video i'm going to talk about the do's and don'ts of the learning phase facebook's learning phase there are an enormous amount of misconception when it comes to facebook's learning phase what you should do to your facebook ad campaigns what you shouldn't do how it works i'm going to clear those up and then explain the do's and don'ts what you should be doing during the learning phase what you shouldn't be doing so you can better navigate that part of facebook advertising before we're going to actually very quickly ask you to smash that like button and subscribe to my

0:30channel if you are new haven't done so already and are interested in seeing more high quality facebook and instagram advertising content okay um so i've got a list here of all the things i want to talk about but before i get into that i'm going to very quickly explain what the learning phase is and i've got facebook support doc up here which you can obviously come and look at i'm just going to read out the very first part and and offer some context around that so this is what facebook say about the learning phase each time an ad is shown our ads delivery system learns more

1:00about the best people and places to show the ad the more an ad is shown the better the delivery system becomes at optimizing the ads performance the learning phase is the period when the delivery system still has a lot to learn about an ad set during the learning phase the delivery system is exploring the best way to deliver your ad set so performance is less stable and cost per action is usually worse the learning phase occurs when you create a new ad or ad set or make a significant edit to an existing one and there's more information that they give there but that's the uh the nuts and bolts of it

1:30so basically the learning phase is something a lot of facebook advertisers fear you shouldn't and explain more about why that's the case in a minute but the learning phase is a great thing it's basically facebook has taken the information that you provided so you've said to facebook look i want to target these people using these ads i want you to spend this much all that sort of stuff right and facebook goes okay within those parameters how do we get this business the best results possible and don't forget facebook is highly motivated to get you the best shots possible because they know if they get you the best results possible you're

2:02going to spend more on the platform and that helps facebook right so their system is going to go okay we've got a million people to advertise to let's say all right let's try and find within those million which are the best 50 000 or best 150 000 depending on your budget to advertise to um we've got this budget to spend on let's say a daily basis okay are we better off spending that at 8 a.m or 5 p.m or uh middle a day are we better off putting these ads on instagram are we better off putting them in the stories placements either within facebook or instagram are they better

2:37off going on the facebook feed um should we advertise to everyone within this audience just once or twice or is it is a an offer that sells better if we advertise to each person five six seven eight times so we have a lower reach but more impressions and there's all these sorts of calculations that facebook is making during the learning phase and they are doing that to optimize performance that's what the learning phase is now the learning process never ends okay there's a more intense period at the beginning known as the learning phase so right when you launch a new ad set or make a big

3:12significant edit to an ad set it's going to go into the learning phase so the learning is there's more experimentation going on on facebook's uh machine learning process but in that the ad set continues to learn the longer and longer it runs so performance um can often and we've certainly seen this happen with our clients continue to improve it's just more intense right at the beginning okay so that's what the learning phase is i want to quickly show you the effect that it has in a perfect example so here's a campaign that we run it's a traffic campaign promoting a blog post of ours um to a mixture of warm and

3:46cold audiences okay and you can see um that we've got the data here so we've rotated this over what i don't know six month period spent just under a thousand pounds on it didn't spend a lot i've talked about this um this particular example before but this is such a good learning phase learning process demonstration so you can see what happens to your campaigns in the learning phase so i want you to ignore this topic because this is results we started off with a much higher budget that we then drop back down so so much of your results is going to be dictated by the budget what i want you to focus

4:16on is the cost per result it's probably the most important thing that um that you want to focus on as a as a facebook advertiser it's going to be the make or break of your campaign okay how much does it cost me to generate a purchase how much does it cost me to generate a lead all that sort of thing so when we first launched this campaign we're looking at a cost per result of about nine p again this is a traffic campaign so results are you know um gonna be very inexpensive plus we know we're doing it so nine p cost per result okay on day one day two nine p

4:48nines in the nines nines nines nines goes up to ten up to eleven to 10 up to 13. so now we're what a week just over a week in and actually our results are a bit more expensive okay then the results come back down again 9p9p shoot up to 11 16 down to 8 up to 14 down to 7 back up to 10 down to 8 9 7 10 ten eight six seven seven okay so we're now about a month just over a month into this campaign and you can see that the average here is probably around seven whereas the average right at the beginning was a nine now we went through

5:38this period here where things got really volatile so we were up at down eight p up at 14 p and again imagine that that's your cost per purchase and you're down at 18 cost per purchase up to 32 cost per purchase down to 12. these sorts of things happen they drive people crazy as facebook advertisers because they want consistency but this is just part of the process so you can see the process but look as we continue going as we continue going we're 11 there and there's always going to be an element of volatility you're not going to get rid of this but look the longer we go we're down to

6:07seven averages six averages down to five seven five spikes up to nine there's always gonna be variation um but look five four five four five so around this point when's that december campaign's been running now for about three months we are less than half what we average say in the first two or three weeks less than half that's an enormous improvement again if you've launched a facebook ad campaign and you've got a cost per conversion of you know 30 and you think we need this to come down big time for our campaign to work getting it down to say 15 might be the difference between you not being able to

6:48advertise and you being able to spend you know hundreds of thousands maybe even millions on a campaign successfully profitably um and you can see that not only is the cost significantly lower say three months in but it's also much more consistent yes there's still variation there's always going to be but it's much more consistent we hit a patch of okay well this variation here is purely calendar because that's the 31st of december 1st january obviously that's going to have an impact but look we then come back down to another period of very stable lovely low cost five four p um etc etc spikes a little bit for a

7:22while but still stays relatively low so this is what the learning and by the way we barely changed so these in these changes and improvement is not our doing this is what facebook's doing because this is just traffic campaign promoting uh blog posts so it's not like we were creating tons of new ads for it like we would with a conversions campaign and testing things this is very much what facebook's optimization process looks like so we know the learning phase is more intense at the beginning and facebook will say that the average learning phase is about 48 hours now it depends how much you're spending if you're not

7:52spending very much and you're generating less results then that process that can be extended um it could be much longer than that if you've got a massive budget we've seen this you get out of the learning phase within a few hours um this we weren't spending a huge amount on as i said it was just a traffic campaign um you know over what period of six months we spend less than a thousand pounds so we're talking 150 to 200 pound a month like not not really a lot at all so you can see that not only we more expensive during the learning phase but even as an extended learning phase like a

8:26week to two weeks in we're seeing really volatile results and that's what you'd like to see during the learning phase more expensive results cost per conversion whatever at the beginning combined with volatility up down up down up down the longer you go and they continue so you would say for example that at this point when the results are all over things are really volatile that's a month into the campaign still moving the the the learning process the facebook campaigns goes through is going to get us all the way down to this section down here not only massive improvements but also much more stability that happens over time so

9:00that's why yes you're going to see a much more significant learning process and things all over the place at the beginning because i said if you've got a million person audience and face was trying to work out the 100 000 of the best they might be testing all 10 segments after a while they'll just card a few of them they'll be narrowing down onto the few so you get better results so it's going to be more volatile at the beginning results likely to be worse but it still happens and it still keeps going for much longer than people traditionally associate with the learning phase of just being those first

9:27few days a lot of people think that the learning phase is just like three days then the campaign never learns again it's all done during those period and you can just crack on that's not how it goes so i thought this is a really good thing to show you i've got a bunch of other points i want to mention um i think it's important okay so average can be 48 hours for the strict learning phase but the process continues and as i said that 48 hour variation is really dependent on budget how many results you generate if you make it a significant adjustment you will reiterate we made no

9:57significant adjustments to this campaign to this um campaign as far as i'm aware maybe one or two but i i don't think so um so what does that mean if you can see from this data that stability helps improve performance that would lead you to believe that you should not be making any adjustment or could lead you believe you're not should not be making any adjustments to facebook campaigns that's not correct you have to adjust your facebook ad campaigns you have to improve them if this is a conversions campaign that we're really optimizing for performance we're trying to get the best possible results we're going to be

10:33introducing new ads all sorts of things again particularly at a larger budget and we're running into ad fatigue issues i think that's where a lot of people go wrong with the learning phase and one of the points i really want to make make strongly here is that you should not be afraid of the learning phase if you need to make an adjustment to your campaigns to help improve things like introducing new ads like testing new creative like testing new targeting options do it the learning phase is not going to be feared but it is something to be understood so what you can see from this data is that yes

11:05you should be a joke making adjustments to your campaigns but don't do it all the time don't be making a tweak on monday morning and then again on tuesday afternoon and then again on thursday morning because you just can't quite leave it alone and have to do it we like to make adjustments usually it can change once every seven days or so and if we're going to make a number of adjustments we're going to do them all in one go because of the learning phase i'm not afraid of going back into the learning phase but i don't want to live there if you make a significant

11:34adjustment to your campaign every two days you're in the learning phase all the time you never get stability you never get the best results so i don't want to live in the learning phase but i'm happy to go back into it because i appreciate that a necessary evil it's like a necessary thing that you have to do in order to have the massive benefit that comes with testing new audiences testing new ads which is super important something you really want to do and you can check out i've got other video content on how to test new ads and how often to introduce new ad creative and

12:03all that sort of stuff okay and just be prepared that once you make an adjustment your results might be erratic for a little bit and then they should start to settle and the process goes again and again and again so i think people tend to fall into one of two camps right they tweak things all the time because they just can't leave it alone or they're so afraid of learning phase they never change anything so once every say seven days is it's good for us now we're usually operating with reasonable size budgets if you've got a smaller budget i would usually be looking to extend that out it

12:32might be once every 14 days you make you introduce new ads or test a new targeting option and not more often than that that might be quite hard for some people but i think it's quite important and particularly after ios 14 when we've got a reporting delay when we know it's taking longer for data to come in you need to give it longer for facebook to optimize but also for you to be able to assess the results that's quite important so give your campaigns time to get out of the learning phase and then assess the results after that that's taken place so a balanced approach like a lot of facebook

13:03advertising strategy is the best way to go now one thing that you may have seen that's all wrapped up in the learning phase is learning limited and facebook's saying you're not able to get out of the learning phase because you're not generating enough results in fact i'm going to jump back over to our support doc and scroll down here where we can see the learning limited and facebook will say here look this is what learning limited is if your adsense isn't getting enough optimization events so that's whatever it is you're going for in the campaign i just showed you that was link clicks it could be

13:32purchases it could be leads depending on your campaign objective and then within that could be your conversion event and if your answer isn't getting enough optimization events to exit the learning phase the delivery column status reads learning limited learn more about ways to fix learning limited now i've got a whole video about learning limited what it is what you could do about it i'll include a link to that in the video description below so i strongly recommend checking that out if you are seeing learning limited alongside your adsense if you are seeing learning limited the first thing to assess is whether or not you think it's accurate

14:04we very often see learning limited on ad sets that clearly aren't learning limited they are clearly are generating tons of optimization events i mean i've seen learning limits alongside assets that have generated thousands of conversions for example you're thinking well there's no way that's learning limited and often it's next to an ad set that's generated far less conversions that is just saying active not saying learning limited so the first thing is go okay is this actually accurate now if you're not generating many conversions or results many optimization events you might be thinking okay yeah this is accurate um i've got a big problem with my facebook

14:35ad campaign what do i do about it the first thing i want to to talk about and you can go and check out that other video for more information on learning limited but the first thing i want to to talk about with learning limited is not to panic ad sets campaigns in learning limited can perform very well we do it all the time and often the the measures you take to get out of learning limited so basically to complete the learning phase could do more harm than getting out of um the learning phase and getting out of learning limited you'd be better off just just operating within learning

15:09limited so for example i've seen people triple their budgets to get out of learning limited it worked but tripping their budget overnight also spiked their cost per conversion not something you want to do i've seen people go from a nice targeted audience to just an enormous over-inflated audience again they got out of learning limited but the campaign didn't perform as well now sometimes you want enormous audiences but not always so you have to rank everything to do with the learning phase is is a is a you have to net it out against what else you're going to do so going back to the previous example constantly adjusting your campaigns will

15:48keep you in the learning phase all the time that's a net negative the constant adjustments is not going to more than make up for always being in the learning phase however go the other side never adjusting your campaigns to avoid the learning phase is also a net negative you're better off the advantages you get from introducing new ads from testing new targeting options more than outweighs the negative that comes with being in the learning phase so it's a net positive to make the adjustments not all the time 7 10 14 day rolling cycles something like that um but that's something to be aware of so i said go

16:28ahead and check out learning limited i'll include a link and we'll pop up somewhere around this video too to that video on learning limited if you want more info on that but that is the learning phase right don't adjust your campaigns or all the time once every 7 to 14 days something like that depending on your budget don't be afraid of the learning phase it's a great thing it helps all our campaigns if you want to test new ad creative introduce new targeting options do it and that is how you should operate with the learning phase something to be aware of something to be respected not

16:58something to be feared something that helps your campaigns and yeah hopefully that's cleared up some questions around that hopefully that's helpful and if this video has been helpful comment below to let me know click the thumbs up button that always really helps me out it's much appreciated subscribe to my channel if you're new and you haven't done so already um three thing i want to quickly mention is my facebook ads mastermind group so i've got a free facebook group um it's a hundred and eighteen thousand members as of the recording this video one of the best communities of facebook advertisers in the world we'd love to

17:27have you part of it completely free to join link in the video description um full of face you know it's full of facebook advertisers just like you and me they're trying to get better results asking questions getting them answered i'm in there all the time i do free live trainings and things um so yeah we'd love to have you in there as i said if this is useful thumbs up don't forget to subscribe and i'll talk to you soon thanks guys bye for now

Where these words come from. This is the caption track YouTube holds for this video, written automatically by YouTube rather than by the creator. We read it, tidied the line breaks and laid it out so it can be read. The plain text version is at https://viewrankai.com/tools/youtube-transcript/OOczkVOSsYQ.txt.

All rights in this video belong to Ben Heath. Watch it on YouTube. If this is your video and you would rather this page did not exist, tell us and we will remove it.