Yuma 4×4

Media and Communications

[Facebook Ads For E-Commerce] CBO Testing & Scaling Strategy For 2019

[Facebook Ads For E-Commerce] CBO Testing & Scaling Strategy For 2019


– What’s up guys? Tyler Garner hear from Standout Social. And in this video today, I’m really excited to bring to you, our step-by-step strategy
of how we are launching, our CBO campaigns to profitably, how we test, how we scale, the rules that we use and
everything else in between. So one thing that we found within the communities that we’re in, with a lot of people that
reached out to us with questions, of their biggest blocks, is that they don’t know how to
use CBO and use it correctly, because it’s got a lot of strong features, but it’s also got a lot of issues and so, we thought it we worthwhile, to bring to you how we’re doing it, because everything that
we do is within a process and within a set structure. And just to give some context over, like, how much testing
we’ve done with this, we’ve spent it over hundreds
of thousands of dollars, just on this process, to get it right. Step by step what we do. It’s very similar to the ad set budget, process that we use with a
few tweaks to incorporate, minimums and maximum spends, and if you have no idea
what I’m talking about, then let me just go ahead
and get straight into it, to show with you. This might take a little while
so make sure you strap in. I’ll make sure that there’s captions, so you can watch it wherever you want. Save this video, comment below, do whatever you need
to do to make sure that you can see this video again and again, because I guarantee you that, this would be one of the
most valuable videos, you’ll ever see. All right, let me go ahead
and shoot myself down here. Move myself to the side
and move myself here. And let’s kick this thing off. So, I’ll give you an overview first, and I’ll show you what it looks like, from a bird’s eye view. As you can see there’s quite a lot to it. All right? Now it might
seem a bit complex, but this is the biggest issue I find, before anything else is that
people don’t understand that, in order to scale, you
actually have to do a lot. You got to get a lot of things right, and a lot of things tested, in order to get to five, 10,
15, thousands dollars a day. It doesn’t just happen miraculously. Most people don’t even
have the right systems to get to a thousand dollars a day. And if that’s you, it’s perfectly fine, it is honestly one of
the most normal things. The biggest block that people
seem to face is around that, $1000 a day mark, they
don’t know what to do and it just ends up either being, extremely profitable one day, or extremely unprofitable the next. And so this will help you, overcome that. Now, the first things to
first, let’s get straight into, how do we test? How can you test with CBO and let it’s effectively
test or properly without, allocating all of the
budget to one ad set, because if you’ve ever
tried using CBO in the past, the chances are you’ve had
it blow out your budgets, blow out your spend, and it literally, you know,
prioritize one or two ad sets and you might have like five to 10. So how do we do it and how
can you also make it work? So let me go to zoom right on in here, so I can share it with you. The first thing that we do with CBO is, we break out our tests per campaign. So if you’ve seen the ad set video, if you haven’t, once again,
you’ll see it some ways, because I’m guaranteeing if this is an ad, you’re gonna get very
targeted with those videos, so it’s gonna be really viable, check it out. So, the way it works is this. We have a minimum of around
three to four campaigns, typically we’ll start there, if the budgets don’t really permit it, maybe we’ll go two, but
there’s always at least, more than one campaign. And so this is all top
of funnel testing, right? So the minimum is always
around $100 per ad set and the way it works is this. We will have five ad sets, each of these ad sets
are different audiences. So, number one, is one audience,
number two, is one audience number three, is a different audience, number four is a different audience. And for this example, just to
keep it really, really easy, let’s say number one is a cat, number two is dog interest,
number three is a bird interest, and let’s say that these
two look alike audiences. Okay, very, very easy to understand. Inside each campaign,
the way that we test, in order to get the most accurate results, is we do not let Facebook
decide which ad performs best or which creative performs best, because, although Facebook has an extremely intelligent machine, it doesn’t always get it right and if early signals
can change day by day, you could get the wrong creative, getting the wrong amount of spend, and the wrong prioritization. Okay? So the way we do it, like I said, is by per campaign, will
be a different creative, to be able to segment out, each individual creative, so that we can identify very
quickly from a campaign level, which performs best. Also why we’d use the
exact same audiences, throughout every single campaign
because ad set one here, is the same as ad set one here, is the same as ad set one here. These are the exact same audiences. Because we wanna also test, at the same time is testing creatives, we wanna know which
audiences perform best, but not just which audiences perform best, which perform best, for what creative. Because if you’ve tested
it thoroughly enough, you should already know, that, each audience can respond differently, to a different creative, depending on age, depending on gender,
depending on interests, and it’s your duty as a
marketer or a business owner, or whoever you might
be watching this video, to test that, to get that understanding, so that you know how to
best serve your audience. Because the better you
serve your audience, the better the sales you’re gonna get. All right. So here’s the next step. Once we’ve tested all the
creatives, we set it up. We’ll always have two duplicates in there. They’re all sharing the same post ID, there’s only one original
which gets set up, in a creative hub campaign, so
we have no originals in here, they’re all duplicates of
sharing the same post ID, just so we don’t have any mistakes or any mistaken duplicates that are not all stacking the same social proof. Once again if you’re not
familiar how to share post IDs, that’s gonna have to be for another video, because that’s not for this one. So, the creative one, is all creative one here, here, here and here. It can be an image, a
video, whatever it might be, that’s up to our creative team to decide, or our media buyers to decide
and we just plop it in there. And we always have three. Now the reason we have three is, the audience’s Facebook optimizes, or enters the auction at the ad level, meaning before bids,
before everything else, it’s gonna assess your ad and enter you into different auctions, and it never gonna compete with itself. So for us to be able to run three ads, nine times out of ten,
all three will get spent, and all three will get different results. Some may get great results,
some might get bad results, and some might just get no results at all, or a very average kind of result. Now the reason we want three,
is because Facebook will see, which one’s perform best
and prioritize those, and if it doesn’t, we’re able to go in and turn the ads off that are bad, and keep the one that is
getting the best possible result running, because it’s
in a different auction. It’s in a better auction, and we wanna reward that
one and keep it running. Okay? So that’s how we do it. And that’s once again, if
you’re not currently doing that, just this is a small hack, you wanna go and do that
as soon as possible. Don’t run just one
creative, go and run three. Or if you’re the very least, go and test the difference
and I guarantee you, you’re gonna see a better result. So then like I said, each next campaign has
a different creative, different variable that we’re testing. The audiences stay the same,
but this is a creative one, this is a creative two and
this is a creative three. It can be carousels, it can be videos, it can be collection ads, it really just depends on our client and also what our media buyers decide. Okay, well that’s the
setup and the process. Now, all good and well to
have a minimum spend of $100, but how can we effectively and correctly get an even distribution. Because obviously with ad set, we can just go ahead and select the budget that they wanna spend. With testing, you want every
variable to be controlled and CBO gets in the way of that. Well, at least it did. And this is how we control it. So if you come down here, I’ll explain this one in a second, if you come down here,
what you’re gonna see is our maximum budget and our minimum budget. Let me go ahead and zoom in, just to make sure that everyone can see. So the way this calculation works, if we’re spending a
minimum of $100 per CBO, what we’re going to be doing.
is setting the max budget, with this equation. So it’s gonna be the max
budget equals the total budget that we’re spending on the CBO, divided by the total number of ad sets. So it’s gonna pull an average. So if we know that we’re spending $100, and we know that we’ve got five ad sets, we know that our maximum spent per ad set, has to be $20. Okay? And then, unfortunately, your minimums
can’t be the exact same. Because otherwise it’d
be super controlled. But then by setting a maximum that it cannot go over that cap, it’s gonna distribute the spend elsewhere, in order to make sure that everything hits a maximum of twenty bucks. So here’s how it works. For minimum budgets, Facebook is set, that you have to have at least a 10% differential between
the max and the minimum, meaning that in order to determine, what your minimum budget needs to be, because you always wanna set a minimum and a maximum when doing this, you need to determine that,
your minimum budget is, whatever the max budget equals, and then you times that by 0.9, and that’s gonna give you a 90% statistic and then you use that one. Okay? So that’s what you stats gonna be. So from this example,
your maximum would be 20, your minimum will be 18, and you run a per each exact ad set. Okay? If you can’t keep track
of that, just save that, and so the next time you
go to setup a CBO campaign, you can use that calculation, because, that is the only way that
you can set it so that, no ad set is gonna spend over that amount and everything’s gonna get an
even distribution of ad spend. It’s the only way to control it. Okay? It’s also how you
can use it for retargeting or middle of model, to make sure that your CBOs will spend a controlled amount, it gives you full control, which is usually what CBO doesn’t allow. Now in order to get to scaling, which is further down the track, but just a pre-find it now, we change these rules up a little bit, which is really, really
cool and I’ll show you that, so it gives it a little bit control, but then we let go once
it’s got a bit of data, and I’ll show you how that
works in a little bit. Now here is some different
audiences that we use, just to give you some understanding. Our rule of thumb is that
we always want to run, a minimum of three
look-alikes in these tests, whatever it might be, whatever
we might have available. So, three stars is high-priority,
two stars is medium, one star is low. So we have a customer list,
a top 20% or 20% page viewers and let me just pause here and say, if you aren’t using
that audience right now, then you need to go ahead
and go do that right now. It’s in analytics, I
can post another video to share with this. If you haven’t already seen
it, just go check the page, on my page, you’ll find it somewhere. And so what this does is, following the Pareto rule principle of, you know, the 80/20,
that 20% of your results, or 20% of people will
bring 80% of your results and vice-versa. Essentially, what this does is, it not only follows the 80/20 rule, it actually goes a little bit further and applies it twice, meaning that there are
specific customers out there, who are not only frequent visitors, but they’re ongoing visitors as well. So, what that means is essentially, they’re coming back to your
page and visiting your website and visiting your website
and visiting other pages, their pageviews are higher
than your average customer. So when you’re building look-alikes, yeah, sure, you can build
look-alikes out of page viewers, over, you know, someone who
hits your page once or twice and maybe they initiate checkout but, what this allows you to
do is go deeper than that. It allows you to find the
top 4% of page viewers, who might have hit your
site 10 or 20 times. And if you’ve got an audience size, of somewhere between
500 to a 1000 minimum, that’s gonna allow you to find the cream of the crop of your audience. And if you’re not already doing that, go ahead, but that video
is not for this right now. So I’m gonna stop there, if you want more of that, once again just go find it yourself, it’ll be on the Standout Social’s page, so you can click above to go find that, or search in our social to go find it, but it will be there. Now, next thing. We’ll use then two times interest, so it might be competitive
brands, obviously interest, or we might use audience insights, to be able to find some key audiences. Okay? Now, one thing I didn’t touch on, is how we’re also using Dynamic Creatives. So, on Facebook obviously unveiled a thing that they call the Power5. Now I’m not gonna sit here
and talk to you about Power5, if you wanna look it up, it’s Google’s job to be
able to show you that. Now, the way it works is the
exact same principles apply, the only difference is, the variables that we’ve got in here, meaning the creatives and
also maybe even a potentially, a couple of different
copy variables that mean, we might want to test, we’re actually gonna include it here. So in here there is only one copy variable being tested at a time. Meaning, that if you use something
very basic, very generic, that’s going to go across
the board to everybody, while still using best
practice in order to try and to have as many sales as possible, the copy testing comes next, so we’ll get to that in a second. Okay? This all has the same copy. In here you might add a
couple of more variables so long as it ties in and you’re not using too many variables, with minimal budget. Once again, there’s another video about how to set up the right budgets for you dynamic credit. If you haven’t seen it, go and find it. And that’s essentially how
the Dynamic Creative works. And so if we don’t have
enough budget for these three, will simply start here. Most of the time, our clients do have a decent
enough budget to work with, but I do wanna be able to show you, how to incorporate them both. So if you’re watching this and you just don’t have the
biggest budget in the world, but you still want to
apply these principles, Dynamic Creative is gonna
allow you to do that. All right? So go ahead and add in, all of your variables in there, go and add a couple of different
text variables as well, or copy variables and go from there. Now, let’s move on to stage two, of how it looks. Okay? So once we’ve done all of that, okay, once that’s done, what we’re gonna do is, we’re gonna pull those winning
audiences and creatives, after around 48 to 72 hours. So two to three days is
when we pull the winners. Okay? From there, we’re gonna start stage one testing next week. So, we give it a week, and then we’re gonna do this again. And so our media buyers
are constantly testing. Our clients will spend on average between, 20% to 30% just on testing, not just to pool winners consistently, but because it actually
works and it gets results. So the results is what’s most important. So if we can help our clients get results, whilst testing and worrying
about scaling at the same time, and also doing all the retargeting
and everything that we do not only does it allow
us to scale horizontally, but it’s allowing us to pull more assets, to move into our scaling
campaigns at a later date, so we’re always able to
scale and grow and continue and we never run out of creatives to spend and our results never
really seem to dip too much. All right. So we’re gonna move on. Now, the next part is we’ll add them into our winning audiences. Adding winning audiences into our winning audience
spreadsheet which we’ve created, which I can’t show you, this got a lot of confidential
information of our clients, and then we also do it
again to winning creatives, for a winning creative
spreadsheet for our clients. And once again, this is for
our internal media buyers and may share this in
another video, I may not, but that’s what we do, we store
everything off of Facebook, so that we can keep track of what works, if we ever need to go pull winners, we’ll go ahead and do that. Then what we do is we move out
winning ads into what’s known as an Incubator Campaign. So it looks a little bit like this. Now Incubator Campaign is a PB campaign, we’ll use our winning audiences, all combined together so long as it works, obviously if you’re using look-alikes, they get wine and interest get another because you can’t use the two together, it doesn’t really work, we’ll spend minimal, like
five to ten bucks a day, it really doesn’t matter if
we use CBO or ad set here, obviously following the CBO principles, you would just set the
budget at CBO level, and if you’ve got two ad sets, just once again, set the
minimums and maximums exactly how we explained earlier. And then what we do is we add
all of the winning creatives into these ad sets alone. Because we don’t care who it
distributes the budget towards and typically PPA campaigns
distribute the budget, relatively evenly. Okay, so we don’t really mind. And essentially, all
we’re trying to do here is keep our winners cooking, what we call cooking or
incubating in another phrase. And the reason being
is we wanna be able to pull these winners at any
time, to go and deploy them into our scaling campaign, so that we can keep rotating our winners, without burning everything
out at the same time. Because a massive mistake people make, is that they end up using, all of their winning
creatives at one time, meaning that they’re all in circulation, meaning that your ideal audience, is seeing all of your best creatives, and all of your best copy
all at the same time. And so if they’re all
running at the same time, typically, the length
of time that it takes for something to die out, which is anywhere between
depending on the audience size, three days to two weeks to three, four weeks, if you’ve
got a big enough audience, what’s gonna end up happening is, it’s gonna start dying out. Okay? All of them will die, and then you’re constantly
scrambling to finding a winner, again and again and again, and that’s a massive mistake. You wanna actually incubate
your winners and hold on to them Okay? So we’ll keep moving. Now, this is stage two, so
that was only stage one. So congratulations for making it this far. Now move on to stage two. Okay? The next part is, so this is for our winning copy, right? So stage two testing, is all about copy. Right? So this is how it works. So our image test is still going on, but what we’ve done now is,
we’ve pulled those winners, we know what creatives work, and now we wanna see, can we improve our results, by tweaking what our copy says? And so let’s get stuck into it. We follow the same principle. Meaning, campaign one, campaign
two and campaign three. So we still split it by campaigns, we still split based on the
variables that do or don’t work. Okay? And now what I’ve
done for ease of use, is I’ve color-coded the creatives, so that you can understand
that each creative, or each color, is a
different type of creative. Many obviously image,
video, carousel, whatever, we’ve pull those winners, and we’ve put them in here. Now we might pull those post IDs, obviously if this is the copy one, that we used in the initial test, we’re gonna pull that and put it in here. Right? These are just
all our previous winners, and these are our past winning audiences. So we don’t use all of the audiences, if they’re not all winners. But if they are winners, then
we’ll move all of them across, or we only move across
two, one, three, five, however many audiences we’ve found that are actually working. Now, into audience or campaign two. Now, campaign two, same thing, same audiences, depends on how many winners that we have, but this time we’ve got
all the winning creatives. It could be one creative might
have been the only winner. Or it might have had two or you
could have had three or five depending on how many tests we’re running. So you have your one, two and three. From here, what we’re gonna do is, this is gonna be a different copy. This copy might belong to the lines of, well, I love this, can
and get this now, link, this might be more along
the lines of informative and say this revolutionary
product does XYZ and solves these problems,
click here to get it now. Completely different copy and we wanna see what’s gonna work better than the other. Okay? And then once
again, self explanatory, ad copy three is gonna
do the exact same thing. So what we’re trying to really
refine and go deep on is, what copy works, with what creative, and how can we constantly
improve our results, even before we get to scaling. Because if, once again, if you can improve the gap between, profitability and breakeven
and stretch that gap, meaning that your results
are better by reducing the the cost per purchase or reducing other metrics
like your CTR, etc, then it’s gonna make it
easier for us to scale and battle Facebook volatility. Okay, we’ll touch on that
and another video I’m sure. But this is all just copy testing. And once again, following the same principles
we followed before, in order to control our spends, and maximum budget is gonna be 20 bucks, and minimum budget is gonna be 18, and our spend is still
gonna be $100 a day. Okay, now depending on client’s budgets, we’ll determine how
aggressive we go with this or, whether we keep our image testing or not. Most of the time we
keep everything running and that’s usually because
everything’s working or we only keep running what is working, and so it’s still bringing
profitable results. So our client become happy
to spend more on testing before we even get to the scaling phase. Okay? And so that’s how that works. Once again hopefully you’ve
worked out that calculation. And if at any time throughout
this video you have questions, by all means, type them below, and I’ll either personally
get to your questions or send us a message and
I can get to that as well, and if you ever, you know,
want some help or have, you know, any advice about
what you’re currently doing, then, once again drop us a message and we’ll see if we can help. Now, once again, the next phase from here is identify the leading poor performance indicators and we’ll touch on that in just a second, but once we’ve done that, we go, we take our winners, and add to a winning creative sheet and the
incubator campaign once again which is in here. So, we wanna know what copy is working, with what audiences, once again, we probably already have
our winning audiences stashed in here we’ll also want to add them into these if we’ve got any new
variables or creatives. So for example, our winning
creative would go in here, meaning it’s the new copy. Right? Not so much images, but the
copy being a new variable, that’s now working, we’re gonna add it into
our spreadsheet here. Right? So, the next phase. Identify leading poor
performance indicators. This means that other metrics
in the funnel besides, just your cost per purchase
and your return on ad spend, day by day Facebook changes
and there’s no denying that. And so what we do here,
is we wanna identify, what is the leading indicator for us? Is that poor CTR? Is it expensive cost per view content? Is it add to cart’ ratios? Is an initiate checkout cost? All of these elements, start playing into, how we’re going to set up our rules. Our rules aren’t set on a
one-size-fits-all approach, it’s actually based on, what’s going to allow
us to scale the best. Or, based on this account, what some leading poor perform indicators and we’re gonna get to that. So let me go ahead and show you this one. Stage three is all about
rule implementation because by this point, we’re looking to spend around, $1000 a day or more, and that is in my opinion the only time you start setting up rules, sometimes a little bit
less, like, 700 or 600, more times than not though
around a $1000 a day is the only time you should
start playing with rules, because that’s where you’re gonna start seeing some consistent volatility, and also the only time where
it’s actually worthwhile. Otherwise, your rules are
showing off too early, or you’re not spending enough for rules to actually be quantifiable. That is a conversation that
we can have it another time. And as you can see, we’ve got
ad rules and ad set rules. We’re starting to slow
down on our ad rules, because we find that it’s actually getting too much in the way, but that’s a conversation
for another time. And I just wanna quickly
touch on the kinds of rules that we still follow and the
principles we follow for them. So we have all day rules
based on cost per purchase, being greater than our KPI goals, and also our return on ad spend is less than what we want it to be. So we switched them off. Then we have all day rules. Once again, for spending
grading and cost per purchase. And then, you know, we don’t
have any purchases at all, we’re gonna turn them off. If at any point that our ads are off, but our return on ad spend is
actually better than the KPI, so somethings kind of
come through with light, or if we had email
marketing, bring up a sale, we wanna turn those ones back on, because they’re actually doing better than we initially expected. And then obviously, if
something’s had spent yesterday, that’s greater than, say, $5, just depending on what rules
you wanna set on your account, we’re gonna turn those ads back on. If it’s a bad ad, and it’s
been turned off early, or we’ve even manually gone
in and turn something off, that $5 metric, means that we can intervene sooner, or we can set other rules
like further down the track that we don’t let them spend too much, so that they can be turned back on. It’s kind of like a kill switch. Which once again, kind of technical, we won’t get into that right now. Now, at sunset last three days, if the return on ad spend
is less than the KPI, then switch off at 1:00am. So that’s where the whole
spend thing will balance out. That it’s not gonna spend
more than $5 before 1:00am, ideally, so that’s going to go ahead and turn off the bad ones. If that doesn’t make sense. Once again, the rules not for that, I just want to kind of, explain to you everything
that we’ve got going on. The next step, the ad set rules. So it’s like midnight rules, spend greater than $1 yesterday, turn it off or turn on. Oh, actually, no, this is correct. So we do something a little bit different, we turn these off, and
I’ll get to that why. We’ve got a thing that we call a winning time curve indicator. So, every accounts different. There is a curve in performance
where you can start seeing, at what time of day, things
work better than the others. And there’s a video on that,
if it’s not yet released, it will be released soon,
and you’ll see that soon. But what we use as a winning
time curve indicator, as to the average of your performance, average of your account performance, what time of day does it go up, what time of days does it come down, and we set rules in between that. So we know when to turn ads on, and we know when to turn the ads off, based on title rules
or more relaxed rules, and it works really well, and our rule, we don’t have
the time to go into detail. So I’ll explain that in another video which you will get access
to and you will see. So we’ll turnouts it’s only
if the spend you say was $1.00 at the right time and once again, we’ve got a training video for that, but you’ll get access to that anyway. This is more for our
media buyers to access. The next one is all day, if the cost per purchase
is less than the KPI, then you wanna turn it back on, so that’s just like a revive rule. Now we’ve got leading
indicators switch off, identify what events
are leading indicators and set rules based on
that, based on account KPIs. This is more custom, once again, there’s a training video on that, it’s not as simple to explain, it’s just depends on
each individual account, what works, what doesn’t and what kind of the maximum spends, we can see we’ll save an account per day, than others and once again, it’s a little bit
complicated to get into here, maybe I’ll shoot another video for that, at another time. The next one, is all day. So if it’s spent greater than
two times the purchase price and the purchases are less
than one is typically, no saving that one especially
at certain times a today, so we’re gonna go ahead
and turn that sucker off. Right? The next one is, all day, if the cost per purchase
is greater than two times, and the return on ad spend
isn’t where we want it to be, we’ll also turn that off, as it starts to creep up there is a very, low chance that it’s actually
going to be able to be saved especially based on
different KPI metrics, etc. All day, return on adspend,
if it’s greater than KPIs, then we will turn it back on and then also once again,
if the cost per purchase is less than our KPIs we’ll
also gonna turn that back on because if it’s cheap purchases, ideally it’s going to keep running up and it can only take one
sale to order a few items, whatever it might be, in order to bring the return on adspend back to where we want it to be as well. Now let’s get into the juicy part. Let’s get into scaling 1.1, and then we’ll get into scaling 1.2. This is probably what a lot
of you have come here for, that’s all good and well to do testing which is actually more important
in my personal opinion, but now I wanna share with
you how we actually scale it. So, similar kind of rule. As you can see, our winning audiences, now this can keep going, depending on how many
winning audiences we have, so long as that we’ve
got the budget to do so. And we actually have an
algorithmic,(mumbles) We have a calculation that we
use every time that is dynamic based on clients, based on ad sets and how much we’re gonna spend
on each individual account. And it’s really, really cool
and it works incredibly well, and it just, it ties in, it’s not
anything we’ve just invented, it’s essentially it just
follows best practice on how CBO wants to work, and it’s just simple math. So let me go ahead and explain. The minimum spend on the
CBO needs to be two times, the minimum spend, of
all ad sets combined. So the minimum spend is 2x,
the average cost per purchase between each individual ad set. So if our average cost
per purchase is around, let’s say 50 bucks, then each ad set is gonna
have $100 minimum, okay? 10 ad sets, $50. Here’s an example here. 10 ad sets, $50 minimum spends is
gonna equal $1,000 budget. Right? As you can see
the minimum CBO budget, is two times the minimum
spend of all ad sets combined. So if the minimum spend
on the ad sets is $50, that’s $500 then we needed two times that, to get the maximum spend, for the total CBO budget. Just let that sink in, if you need to kind of like summarize that or like look at this, either
screenshot or pause this and just draw it out so you’ve
got some understanding of it. But hold on to that
because it’s important. Now, what you’ll see down here is that, there’s a minimum budget set
on each individual ad set. The minimum budget on
each individual ad set is two times the average cost per purchase. Here’s how this all works. We have no MAPS, meaning that these
scaling budgets can spend as much as they like, as
much as they possibly like. He’s how it works. So we can continue adding
audiences and like, all our winning audiences etc, so long as that the budgets
is there for the client. And then we’ve got four
creatives only inside that, keep other creatives for rotations. It’s like as if we rotate our winners, we don’t burn them all at the same time. So we’re winning creative,
one, two, three, four. Now these could be the same
image or same video or carousel or whatever with different
copy variations or vice versa really it’s just a matter of
knowing what’s in use right now and what isn’t in use bla bla bla. Now, he’s how it works. And let me explain this
in detail how this works from a mathematical position and how it works within
what CBO’s are meant to do. We you control the spend on these ad sets, up until it’s got at
least two times purchases. And why it’s two times is
actually really really important. These two times per purchase, or this two time purchase spends minimum, is based on our rules. And so our rules will kick in, once it spent that minimum, meaning if they aren’t currently
performing to the average, of what we know our account performs, and how many purchases
they’re supposed to have, they’re gonna be turned off, before they get the chance
to spend any more money. So if we’re spending a thousand bucks, the only ad sets that gonna
continue to scale after that and continue to spend more
money than the others, are the ad sets that have
performed either on par, with good performance or
better than good performance. so we might have all four
turned off and one turned on. But we’ve got these rules
turned on the entire day, to make sure that these are controlled, so they’re never gonna be turned off, or they’re never going to
overspend and lose results. And the minimum spends, are in there set that they
all spend an even amount before we let CBO run its course, or let it be controlled. Because CBO will just wanna spend, what it decides is the
best possible ad set. But typically it doesn’t always know best. And if anyone’s run CBO before, you kind of to probably
know that yourself. And so what we do is we control
it so it forces data in. So it forces spend in so it
gets enough data for the day before we let the algorithm
in this machine learning, run its course. But once it’s done that,
the rules are turned off, and then it’s kind of like, if we’ve got five or ten ad sets in there and 30% or 50% of those
have all been turned off, we’re left with those
that are performing well and at that point we just release
the floodgates essentially or open the gates so that
Facebook can start spending what it decides performs best. If you need a like just recap on that, because I understand it’s
actually quite a lot to understand it’s, once again, it’s simple math, but if you need to understand
the logic behind it, rewind this and just go through it again, I find that writing something out, if I don’t understand it works best. But CBO is no joke. So that’s the biggest thing. A CBO is no joke and this is how we do it, this is the only way that it makes sense to let this machine learning and there’s like what Facebook is forcing us to do, either in September or
February depending on how you’ve been running
your account so far. It’s the only way that it’s gonna work. Now let’s move on, because
obviously there’s no Q&A in this. So we’re gonna move on. If you have any questions
about that section, once again, feel free to just contact me, but it’s pretty self-explanatory I’d hope. Now at the next part. When creative fatigue
from scaling campaigns, keep ad sets running and rotate creatives, keep other ads in ad set
but just keep turned off. So, that’s something for another time. Let me go ahead and move on. So now we’ve got a couple
of juicy CBO scaling methods that we can introduce at all different times, and I’m really excited to be able to share these with you as well. These have all worked at
different times for us really really well. So let me go ahead and explain these as easy as possible. The way the rules work and
the minimum spends, etc are all the same. It never changes. So that principle that you saw before, once again, I’ve never seen, out of all people I know, out of all the big spenders
that I’ve spoken to, no one’s doing that, and like I said, it’s the only way that I
have personally discovered, that allows us to controllably spend CBOs, and then give it the power once it’s actually got decent data. Because Facebook’s not dump, but sometimes they just
need to be controlled before we let it run its course. So, the first way we run
it is at CBO campaign, it’s all the one winning audience. It’s all the one winning
audience meaning that, if it’s dog interests,
this is dog, this is dog, this is dog, this is dog, this is dog. And it’s all the same. Same age, same placements, everything. Then we have four winning creatives and rather than just go ahead and draw everything out massively, we want we follow the
same principle where, whatever the winning audiences are here, are the winning creatives, we follow the same winning creatives here. And we do that all the way through. So it’s all the same creatives,
following the same thing, that when they die out, we turn them off, we don’t delete them, and then we’ll rotate new ones in. Minimum budget, once again, two times the average cost per purchase, the maximum budget is, no
max at this point as well because we just want Facebook to spend as good as it sees it fit. And there’s other ways
you can scale with that like adjusting the minimum, if you only want one ad
set per day to spend, just increase the minimum, which kind of forces it
to have to spend the most, especially if, you know, that it’s doing better and you wanna manually get involved, just control the minimum, not the maximum. The next one, CBO method two. Notes;Manual bid lowest cost bid cap. No notes on this one,
pretty straightforward. This one here, CBO campaign, once again, only the winning audience, so, only the one type of
audience in this one again, we just wanna duplicate it
because we want Facebook to find the best pocket
of the winning audience, through the audiences’ selection, as well as through the creative selection, and also through the best bid, because these are all manual bids. Same minimums and maximums
don’t even go into that, the only difference is if I can go into here is we’ve got, winning audience one, big cap two times our
CPA 2.5 times our CPA, three times 3.5 four
and five times our CPA. Hopefully, that’s pretty straightforward. The next one. Repeat with all winning audiences. So CBO, method three. The next stage is this. We’ll go as you can see here, oh wait, no ,so this one applies here. So, we repeat this whole process, with all of our winning audiences, which might mean we have
quite a few campaigns, but it’s all about winning audiences, and it’s just another way to horizontally spread our budgets across, while still knowing that we’ve
got the right assets in there in order to allow us to scale. And once again, if we follow all of these, so if we follow this, if we follow this, if we follow this and just those, and we’re still testing, chances are my friend that
you’re gonna be spending in excess of three to five, most likely around 10,000 bucks a day, and typically its profitable. More times(mumbles) either
front end profitable or by week two or however
you run your systems, you will be profitable and
it will be scalable at that and that’s how we control everything. Now let’s move on a CBO method 3 because this is a bit different. So CBO campaign, once again,
pretty self-explanatory, the only difference is,
is rather than running up here, which is all of
our winning audiences once, all we’re doing is we’re doing it twice. So this is gonna jack the
budget up a little bit higher, but essentially what it’s
gonna allow us to do, is it’s going to allow us
to make sure and guarantee that the winning audience, we
know is a winning audience, is getting the best possible chance, to perform well today, not tomorrow, not another day. It’s allowing us to make sure that if the one audience or winning audience one doesn’t do well in this ad set, it gives us the ability to
know that winning audience one is gonna do well on this one, or at least by chance it’s gonna do well because once again day-by-day volatility, comes and goes, one day it might do extremely well, the next day it might perform
like a dog’s breakfast, which is just rubbish. Right? And so, that is essentially how we do it. It’s a bit of a marathon
but we finally got there. There is not much more to it guys, there’s other ways you can do it, like split out placements
with the one winning audiences which we haven’t added in here, because we’re still playing with that. What you’ve seen in this spreadsheet, guys just so you understand, is this is what we’re doing right now. Meaning that this is not theory, this is not something
we’ve just thought of, this is this process as we follow through our clients right now. We spend around over 500k a month, the
moment for our clients, if not more depending on
obviously who were working with and who’s scaling and who’s
got different events coming up etc. And so this is what works for us. It allows us to use CBO, we still use ad sets as well and I’m going to share another video with that. If you haven’t seen it already, you will see it soon. And this is our mosaic marketing method or mosaic testing matrix, that we’ve gone on to name and so if you’ve got any questions about this, you can comment below. There is a comment section. Let me go ahead and make
myself a little bit bigger now, and yeah, that’s
essentially in a nutshell, how we do it. I wish I could say that
there’s more to it but I’m sure that’s enough anyway. If you’ve got any questions about this, let me know. And fortunately if
you’ve made it this far, or even if you haven’t
but you’ve got questions, I’m going to give you guys
access to this framework so you can take it, you can download it, you can print it, you can study it, and if you’ve got any questions at all, you can let me know. But, the only thing that
you won’t get access to is our creative audience spreadsheet, because once again, kind of confidential. But if you’ve got any
questions other than that, comment below, look in the comments, there should be a link, there might even be a link above depending on where you’re seeing this, a little description or whatever
but yeah, guys, that’s it. And I’m glad I could share this with you, and like I said if you’ve
watched to the end, thank you for coming
on the journey with me. Anything else, just let me know. Drop a comment, send me a message, reach out to me, but this is the number one way we
found to controllably scale with CBO and control the volatility without letting the machine take control, until it has enough quantifiable data, and then tie rules into this as well, so only good performers can keep running and keep spending the way CBO wants it to. Yeah, feel free to watch this
as many times as you like, send it to your media buyers. Tag your media buyers,
do whatever you need to, and obviously if you want our help or you wanna
know if we can help you, then I’m happy to have that conversation but at the end of the day
you’ve got our processes, have a look and if it seems like too much, then hey, we might be
able to be a good fit. But yeah guys, until then, thanks for tuning in and
catch you in the next video.

4 thoughts on “[Facebook Ads For E-Commerce] CBO Testing & Scaling Strategy For 2019

  1. Sick video, have been testing in a somewhat similar way with cbo. The top 20% of 20% of page viewers is a mad idea mate, how do you expert this data out of your google analytics?

  2. So To finally understand please, if in phase 1 having a CBO campaign, and after 12 hours, i have spent $ 20 on the first adset that don't generate any sales, and the second adset with spent $ 12 yes it generated 2 sales, Only that Facebook is focused on spending the budget on the first adset, can I decide to turn off the first adset correct adset? but on day 2 should I turn on all the original adsets again? just keep the remaining adsets? or replace with a new audience?, thank you very much for the value, currently doing 300 / day with cold interests without cbo

Leave comment

Your email address will not be published. Required fields are marked with *.