Expert Tips for Managing Campaign Overlap in Google Ads | Google Ads Basics Part 4

Campaign overlap is something that I think that

will need to get much better at.

At performance.

Max, , becomes more of a , permanent foothold in Google.

Some ways that they overlap omnichannel.

Is regardless of what your performance Max is targeting, running, how

it's set up, it, does not matter.

You cannot stop this if you wanted to.

We've seen it before.

We've actually seen it kill our performance Max campaigns when no

changes are taking place in our accounts.

Let's talk about, campaign overlap.

Not all our clients have, , but not even nor Beam is fully accurate.

Campaign overlap is something that I think that will need to get much better at.

At performance.

Max , becomes more of a, permanent foothold in Google.

I think it's good performance.

Max works.

, go ahead.

It's allowing me in this campaign to spend, you know, 275 grand a month at a

1.73 return where I need to be over 1.49.

It works, works well.

Campaign overlap , , is insane and I really want to discuss how PAX

works at a higher level because that needs to become our basic.

performance max overlap, how it overlaps, what's being represented.

, some common understandings of things have changed and some new developments

have come alive that I think it's worthy of having all of us get a very

good understanding as How the campaigns overlap and why, and, all the good stuff.

Here's something that we're all going to need to get very familiar about is

the asset groups and then also dynamic remarketing inside of Performance

Max and how it works within perform Google Ads campaigns and how it works

Omnichannel, so traffic from other sources, Performance Max is gonna

overlap more so than we know initially.

Some ways that they overlap.

Omnichannel is regardless of what your performance Max is targeting, running,

how it's set up, , it does not matter.

You cannot stop this if you wanted to.

It will dynamically remarket all of the users that come to

the site, regardless of source.

So if you have a Performance Max campaign set up for a hundred percent cold traffic

and you have a campaign that is not using T C P A or tira, so it's learning

its own audiences and you have Euro URL expansion off, doesn't matter if a

massive amount of influx from Facebook comes in, your performance Max, that's

brand excluded, your extension off.

Cold traffic only targeting will remarket that Facebook traffic.

We've seen it before.

We've actually seen it kill our performance Max campaigns when no change

has been taking place in our accounts.

I forget exactly the client.

I know who it is.

, I think on car.

You're with me on it.

But it came out in a risk meeting a few weeks ago, and it was kind an interesting

point because it was something that we all kind of found out together.

But I wanna make sure that it's kind of , , part of our foundational

understanding of performance Max.

So, as Google ads experts live and die instead of Google ads,

and sometimes we don't think about outside influences to our campaigns.

If you have a Performance Max campaign on, and it's running, it's been running

for let's say six months, getting a 300% row ads and everything is hunky dory,

and someone launches a Facebook campaign to a bunch of uninterested users and

they drive themselves to the site and they look at some products and they

leave, your Performance Max campaign is going to reduce its efficiency.

It's not going to hit 300% roas.

it's gonna spend the budget, but you're gonna see the same budget.

Your click, the rate's probably gonna be fairly similar.

Your impressions will be a little bit higher.

start to tank performance.

it's nothing that we did, but we have to understand the ebbs and flows.

The rises, and.

of Google Ads and how it, optimizes in other campaign types.

So I'm gonna give you all a quick example and I'll share with you how your campaigns

can, immediately benefit or immediately, , retract, I guess in performance.

So this, the company we've been marketing for, about seven months.

Performance Max works really well.

I have,

, 18 Performance Max campaigns running

that anytime we'll see Facebook added increasing and this is clicks.

This isn't click and views.

Anytime we see, let's say Facebook increase, we see Google increase.

Anytime you see Facebook decrease, Google decreases and I haven't made any changes.

Is that a Google?

What is going on?

Is that, The quality of the user coming from those outside

sources directly affects how your campaign is going to convert.

Google is a self-governing, self targeting system, and it has hierarchy levels, and

one of those hierarchy levels that it likes to attract first, or likes to market

to first is the warmest audience, which is who the people currently on the site.

That's why if you add in a whole bunch of ad spend, you don't necessarily

get a whole bunch of new clicks and conversions from cold traffic.

You simply start to potentially remarket them harder.

So you don't really necessarily see that that conversion coming back.

You hit a point of diminishing returns quickly, especially if you're not

in control , of your audience and.

Okay.

You're saying if you're using conversion base bidding, when I'm talking about

performance Max, since we're kind of stuck with maximize conversions,

value maximize conversions, that's kind of what I'm thinking of.

, something different that you were thinking of?

, I was just saying just in general terms like can happen to any campaign, but yeah.

Yeah, like things like inbound search, not too much.

, cuz it's not gonna remarket that kind of cold traffic on the site.

So like other channels or other campaigns will be less susceptible to it for sure.

Though.

You will have like display, if you're using optimized targeting

as a list, it's gonna include those people in there for sure.

, and Google just use this as an ecosystem just in, in and of itself.

Any, anytime you're using Smart Mini that has its own kind of directional

governance, even if it's just optimized targeting turned on, it's gonna go

where it wants, which could be those new traffic to the site where they come from,

doesn't matter, it's just new traffic to.

So knowing this, this is something we have to remember.

Foundationally sometimes ask this on a Wednesday call and it's like, Hey, our

performance Max did, we don't know why.

And I'm like, well, what'd they do on Facebook?

We're like I don't know.

. That's something that we'll need to investigate first

before looking at Google ads.

If things just started to decay out of nowhere, and what I mean,

out of nowhere do the work, e-com product is still working.

Convergent tracking's, still working product that is doing

amazing, didn't go out of stock.

Whatever.

We didn't launch a new YouTube campaign at five grand per day and all of a

sudden just flood our PMM Max campaign with a whole bunch of remarketing

traffic that's not yet ready to convert.

Like unless you did something that you know, could have an.

Before we look into saying, well, I'm not sure why we need to get with a client

and we need to say what has been changed , inside of your, Google Ads account.

There's also not even a paid channel source.

This also can be heavily, heavily affected.

I'll share with you an amazing, . Like, dang, that I wish we would've

captured but wasn't due to us.

And this is a client , that I was talking about on this morning's call where I said,

I'm running a feed only, . I'm running a feed only and a YouTube at the same time.

This is what I'm doing.

It's a feed only on YouTube outbound and I'm scaling this up and it's

working pretty re pretty well.

, this here, , I saw my feed only campaign in my YouTube campaign.

We have all of a sudden like a good spike on like really good traffic and then

it died off, and then it started to go, go against the spike that I saw here.

It went from like, Two, then, four, and then two, and then five, and

then crash hard at 1, 2, 3, 4, 5, and it started earning its way back.

What actually happened here?

I pinpointed it.

It was down to a client, this client here, sending out a massive tens of thousands

of emails to their own subscriber base.

People clicked on those emails.

They went to the website.

They didn't buy.

They left, we remarketed to them.

They came back and purchased what was a traffic source,

their own customers on email.

Crazy.

That's how touchy pmax can be, especially when remarketing built in.

And that is built into every single Pmax campaign.

So diagnosing this, we have to think about within and outside of

Google Ads, what is happening on all traffic referral, organic, direct

social, influencer, email, all of.

Is going to affect our campaigns and if we're not understanding this and if

we're not investigating this, we'll be chasing our own tail from something that

we never catalyzed in the first place.

Does that make sense?

Any questions there so far, Regina, what's up?

Can you clarify that?

Drop in conversions is because people were sent to the site from an email,

and why did the conversions drop?

Nope.

This day, right here on October 1st where I was like, oh,

this is, this is going good.

And then why did drop, and then why is it slowly come back up?

The first was their email, campaign.

on that day was the day that the email was sent out.

Yeah.

So it's like, here's my traffic, my traffic, my traffic, a bunch of email

traffic, and then we convert a bunch.

It says, I'm gonna remark all this email traffic.

They didn't buy.

Oh, okay.

Well we, we off that email traffic.

Go back to my traffic.

I see.

It's an interpretation.

It's my best educated guess.

But that's when I look inside of their, their campaign, I'm like, oh,

they're back end of their account.

I'm like, you had a 66% increase in value on one day and 10,000 users come to

the site on one day and 44% increase in conversion value in one day and like 16

grand when we usually make 50 in one day.

Like it was an anomaly.

But then you see that same anomaly in Google ads.

Why?

Because I'm remarketing with a feed only dynamically.

Everybody that clicked on the email went to the site and did not buy.

Even if they did buy, it's still remarketing them

because I have no exclusions.

You can't exclude Pmax.

So that's the thing too, is if you have a massive amount of traffic

that purchase, we may see a dip in our conversions because it's

like these people are on the site.

Well, they did.

They buy her, did not buy.

I don't care.

These people are on our site.

That's what Google will.

So outside influencing is really susceptible to our campaigns.

So that's a good understanding.

Now, it's also susceptible within our campaigns.

That's another danger.

So when I'm running things like , this campaign here, you can see the

overlap in how crazy it is just from a search category where it's

like, you know, rice, water, shamp.

This was found in all four of my, my asset groups internally.

Why Google's matching so broad that if anybody looks similar to

anybody , it's like close enough.

So here's what I mean by that, is what we're talking about.

People that may have been Googling, , a shampoo bar and then you have a person

that may Google Rice water shampoo.

Those are two different campaigns.

Those are two different campaigns in my account, but it will share the same

user, which means that person from day one could have been targeted by

either campaign, even though it's a different product and a click on one

campaign and a click on a different campaign that resulted in a conversion.

Is going to be split between our campaigns, meaning my Roaz and my cpa,

my roaz is half and my CPA is double.

So when running multiple, multiple campaigns, we all have to be very, very

cognizant of the facts that if we're using data driven, a conversion that comes in

may look terrible on a day by day basis.

Here's what I mean by this.

If we have conversions that are split, watch this.

was that by time?

Okay.

. Check this out.

This is an extreme case, but it's perfect cuz it makes my point.

I spent $39 that day.

Anybody explain how

conversion look?

At distribution, the point we only attributed 0.01 of a

conversion, so that's ridiculous.

Exactly.

It should just be zero and then it would be a zero cpa.

What happens if your cpa, your T C P A is set to a hundred dollars today?

Did you, did you miss your goal?

Oh yeah.

You way missed the goal.

What is it gonna do?

Is it spend the next day probably.

, here's a less extreme example.

I spent $125.

We know that 2.34 means three.

We all know that to be true.

That third one was just split.

, we absolutely know that anything that's over a 0.00 is a conversion.

So you spend $125, right?

You divide that by three, what's my cpa?

41 66.

What's my CPA to Google?

53.

So I spent 125, actually got three conversions, but I have

a $53 cost per conversion.

In reality, it's 41 cuz the other campaign, we have to take that out from

when you split a conversion, you double or triple or whatever it is, your cpa.

Same as true with conversion value.

If I make 1.1 sale, what's my conversion value?

, 43 cent.

Because I got 0.01 of a conversion and I actually probably got,

you know, $4 and something in a purchase, but I got 43 cents.

So what's my target return on ad spend?

That's a pretty bad ad spend.

You spent, bucks to make 43 cents.

Is that true?

No.

What we're talking about here is this the overlap between the two.

If I take a look at something and said, man, , I'm getting $29 c p A here.

This is looking good.

Ooh, this one's getting a 59.

That's bad.

Pause this.

And then this thing tanks.

Why?

Well, half my conversions that were here were also half of my conversions here.

All I did was just shut off half my count.

So knowing that these are going to be in line with each other when

they're splitting conversions, this is not real and that's not real.

Why?

Because that 0.3 8.9 1.1 0.4 7.05, 0.46.

. Well, at the end of the day when you go down to something, this is gonna

equal out to a hole minus one or two because it came in through maybe,

let's say a few day, few days ago.

But this is splitting two things.

It's splitting the time and it's splitting the campaigns.

A click on day one that converts on day two, we'll have 0.5

on day two and 0.5 on day.

So both those days I overshot my CPA A by double and I cut my conversion

value in half those two days.

This is also why T CCP a T row as are a little bit more dangerous when running

multiple P M max campaigns because your T Row as a T CPA A is measured against

Hals and thirds and quarters not holes.

So if you wanna say, Hey, I need 300% roaz.

Awesome.

Set your targets to one 50.

Why half of those are split?

Very odd, right?

So when I look here, when I look at my roaz, . so it's like 1, 5, 1 3.

One four.

One, two.

Two 13.

1 13, 1 3.

It's my higher spends are lower, . They all at the bottom equal 1.72.

That's cool.

I'm not gonna shut up my 107 cuz I have another half of that

campaign probably at a two 14.

And then that's equally this.

So just know that when you're measuring individual p m max

campaigns, and that's why when it's like, Hey, should we shut off this?

We shut off that one unless it.

There's no conversion paths.

The first and last click attribution show the same.

There's no way that these two are the same products.

Person that's buying a dog bed is not gonna buy a toilet.

okay, these have nothing to do with each other, but the more similar that

these products overlap, which this one does amazingly, your eye masks

pillowcase your hair clips your hair shampoo, your scrunchies, your headbands.

these are all products that are just from like the top eight inches of your body.

Like everything overlaps here by person that wants a scrunchy

is gonna buy a hair tie.

I can't really care or tell the difference, but I'm measuring

at MER and overall count ros.

and then you're splitting it by day.

Clicks on day one and clicks on day two, and then you're splitting

it by clicks on day one and campaign one and clicks on day two.

By campaign two.

Now you have 25, 25, 25, 25.

All four campaigns now are missing the C P A targets by three quarters

and have all been had their conversion values cut by one quarter.

That's bad when you're looking at what do you do in the account.

So you have to look at those overlaps and then know that also in addition

to this, it's also remarketing the other channels like Facebook

So this got real super deep, really fast, and it's going to

make sure that as we optimize.

Don't take this for face value.

There's three things you're not taking into consideration when

you take it at this screenshot.

Here is how did the previous clicks from the previous month compare to this month?

How much campaign overlap in there and what happened?

Omnichannel, unless you investigate and know and kind of keep up with that, you

can't really make educated decisions.

You're just making reactions that are going to be half right and half wrong.

That's bad.

The other part that I would like to talk about, Is, and then do we

have any,, questions on that so far?

I know it's got like really crazy.

There were some comments in the chat earlier regarding fractions, but

I think that one is gone already.

Okay.

One good way to do this too, if you wanted to, is use first click.

, if you wanted to use first click, you can, or last click.

If you're not running brand, I don't know.

It's gonna be up to that individual campaign and how you wanna measure it.

Attribution models are for you.

They're not for Google.

It doesn't matter to Google.

, it only matter how you want to count it.

versus how you wanted to show up.

One quick way, this is gonna be an advanced technique, but if you wanted to,

you can actually replicate a conversion action, counted as a custom to campaign

level conversion tracking and not account level, and then change the attribution.

So , here's a ninja one.

This purchase is running on data-driven.

This data-driven, I want it on data-driven because I'm getting

view engaged conversions.

First click does not count.

View engaged conversions well.

So know that you miss your view.

Engage conversions, because that's not a click attributed conversion.

Data driven is not a like attributed data driven conversion, or not

a click attributed conversion.

It is a,

, View and click.

So if you change the first click, know that you will miss

the view engage conversion.

But here's something we can try if you would like to.

Replicate this conversion action.

You can't just like copy and paste in here.

You have to actually go through and set up a new conversion action,

counting the same exact thing.

This is , in your measuring measure as a custom goal, and then take

that custom goal and count it as a different conversion, , attribution.

So if this was the shopping app purchase and I say, you know, , this is, name

your custom goal, that's cool, , . And then this custom goal here, you can

actually, when I can't do it now, when you actually set it up, you're gonna

be counting that conversion action as a different attribution model.

Let's say this was first click.

, and if you, can count it as a secondary and then attach it to a campaign so you

can say, okay, this campaign brought me 101st clicks and 25 last clicks.

Actually, no, what?

I did this already, sorry, I already have an example of this.

Here's something awesome now saying to do this, but this is.

A high level trick that makes us all ultra ninjas, , and is very impressive.

So inside of this account I did, forgive the nomenclature.

I'm using other people's and changing it around, but you'll

see stuff like this here.

data-driven test, 90 day counting every.

And then this is website orders, a 15 day only counting one, and then this one

is a purchase 90 day also counting one.

It's actually not that these involvements switched around, so

just ignore what these are counting.

But you can actually see the differences between the conversion

actions a 90 day versus a 15 day.

This is actually the 90 day, I think, , , this is still the 15 day.

I forget exactly where it was.

Long story short, this is all messed up.

, it's me just kind of modifying it to try this, but you can actually see the

different conversions and how they're attributed, , for the different campaigns.

So segmenting by conversion and by conversion action.

I'm only accounting this one, one, but I can see, okay, there's

data driven would count, , 4,600.

New customers are 2,800, the purchase and one count for 90 days.

Okay, so it's 463 and 43 97.

There's 200 returning here because this was 90 day, 90 day one count every count.

There's 300 more here.

Perfect.

So now I can see what the difference is between them.

You can do this if you wanted to kind of see different attribution models

for the same kind of cool stuff.

Not saying that we need to do that, but just know , that's the thing.

I think that's it so far.

There's a lot of basics to cover, but the basics I think are the

way that we've run our campaigns.

This is kind of like the basics, convergent tracking.

How the budgets work, how the daily ad spends changes actually

equate to future changes, the restrictive bidding strategies and

how they will kill your campaigns.

The, kind of overlapping between omnichannel and inter channel, the

fractional conversions and how they mess up your goal tracking and what

you should be tracking as a goal.

You can actually use this model here to build a brand, but we'll talk about kind

of brand building techniques next week.

, this I think is something that very important and is going to give us a

templatized structure to how we should be building out the e-commerce accounts.