Managing Performance Max Asset Groups and Campaign Overlap

They will all share users like crazy and they all will remarket everyone

else's traffic equally like crazy.

So that's what's nice is making a push here is good.

The conversions attributed to it is kind.

Anyone's game after that.

So when you're looking at Lowe's, multiple Pmax campaigns, one constant

that we can count on is, yes, I have this much ad spend marketing this product.

That's all we can count on in the beginning, and that's what

gets a little bit dangerous.

So here's what's really interesting about Pax, is they will all.

Share users like crazy and they all will remarket everyone else's

traffic equally like crazy.

The deciding factor for when one Pmax campaign takes over another pmax

campaign is simply by available ad spend.

That's actually something that's really, really dangerous in trying

to manage it from a structural per.

So when we're looking at campaigns that have, , massive amount of

clicks like sunglasses and speakers, , they're directionally accurate.

The right products are showing That's correct.

What happens kind of post first click and post, , kind of first

visit is then kind of a crapshoot.

So that's what's really interesting about, like speakers for example.

And sometimes you can identify that from insights cuz it is gonna

tell you that kind of inbound.

Where, this is what I'm saying is directly accurate from the

first click shower speaker.

Nailed it.

So when people come in from shower speaker six, two clicks,

what happened after that?

No one knows.

Hmm.

And that's what gets dangerous is the product that you're gonna market spot on

the the audience that you're gonna go.

Spot on.

This is kind of like d s, A and V combined to give the right products

based on what you're targeting as.

So that's why when you look at shower head, you're gonna get shower stuff.

Perfect.

When we look in the sunglasses, we should get sunglass search terms.

Right now they're branded.

That's okay.

, smart sunglasses, good.

Photochromatic good.

, dust, , electromatic Good.

, changing.

Okay.

All good.

So that's what's nice is making a push here is good.

The conversions attributed to it is kind of anyone's game

after that, and also the return.

Is gonna be shown a set of search category as well.

So when you're looking at those multiple PMM max campaigns, one constant that

we can count on is, yes, I have this much ad spend marketing this product.

That's all we can count on in the beginning, and that's what gets

a little bit dangerous because

what a pmax is gonna do, take one product, for example, one audience with one ad.

And earn users.

It will also remarket them, , remarket Google users.

They'll remarket, other sources will also show up for brand.

Sometimes it won't show up for brand specifically, like if you

have DU sunglasses, sometimes you'll just see amp here.

So you get kind of all brand traffic.

Did they show up before?

Yes.

Did they come from Google?

Yes.

Did they come from other sources also?

Yes.

So when you are looking at a campaign type that can do everything

and it has a specific purpose, a lot of times that is very good.

That's kind of the way that people buy.

, they need to see it a few times, click.

They need to be remarketed, they need to come back.

And you're seeing all that inside of performance.

So what's nice about this is it's a very, very efficient campaign

when it works well, like we see, good mix of cold traffic, some warm

traffic, probably a good mix of shopping, , clicks, some display or

marketing clicks, some search clicks.

The CPCs are probably, hopefully within one to $3 cuz that

should be a good average.

Nailed it.

, one and two and one, and , kind of some, some twos here.

So between one and three.

Good.

It's indicative that it's standard shopping and also dynamic marketing.

This is healthy.

This looks good.

Lovely.

Let's skip the, the conversion tracking and skip the, , , auction insights.

, actually and auction I, but the, the audience and just kind of

talk about this here for a moment.

Here's what we've also been seeing though inside of Pmax is, Depending upon the

asset groups and how many asset groups there are, it takes a total impressions

that it was going to give to that and divides it fairly equally between

each one of these, these asset groups.

So we've seen a shift.

The signals are not as directional as we all originally

were told and thought it was.

There are more polite suggestions that can be chosen to be.

I'm not sure why signals exist.

I've never really moved the needle that much.

It's been more just based off of like, how would I describe this?

, chance.

What AD showed that kind, the conversion.

Well, why did the show cause campaign had ad spend?

So here's what's interesting though, is the all converters, the

purchase intentions, the interest.

The remarketing and the wireless I Chargers, for example, this was paused,

so we can't use this so right now, but at a time, at a time, this was running.

When you look at the insights, the reason why Google shares a user with every

single one of those, , asset groups is because they all overlap enough

that Google says, eh, same person.

So what this means

is if you had a hundred users and one.

, those hundred users go to that asset group.

If you had a hundred users and two asset groups, 50 users go to

each asset group, not permanently.

They could bounce around, but the sum of the total will be divided by two,

three asset groups, 33 PE people in each asset group, and so and so forth.

So now we look at, okay, if that is the case, and that's what

we've seen so far when you add ad.

The groups all just equally go up some more than others.

But what I'm saying is all of 'em have a chance by saying that that was two

and then eight, and then one, now this is three, and then nine, and then two.

Like they all just kind of rise and fall with each other because

the S groups overlap so much.

So how, what do we do there?

An opportunity here that I would say is a good idea is inside the

asset groups, knowing that each one of these are essentially ads.

. What if we use different, we consolidated down a bit, maybe consolidated asset,

, , or sorry the signals down a bit maybe, but really divided up , our imagery

or change up our imagery and our ads.

, Glen, what's that?

You've pretty well preempted my question.

This structure.

And this pretty well universal across most of our accounts, we've

been breaking out asset groups and adding one signal per asset group.

Yep.

What I'm hearing from you is, rather than doing that, get all

of our asset, all of our signals, bundle them into one asset group.

So you might have six signals, put them into one, and then build out

asset group by creative and copy.

Yep.

That's it, that's what you're referring to.

That'd be the next thing I would try.

Yeah, because what Google's doing, it says, hold on, sorry.

Uh, so John, , in my part, actually we are trying this.

So all these new buildouts, the Pax Buildouts, we are sticking with

like max three asset grip per P max.

Yeah, per category.

Like, don't go above that because right now you showed in in source.

All the audiences are overlapping.

so yeah, that's why we are sticking with like max three asset groups.

I'm testing this with you right now.

Yep.

You'll see that my signals are the same.

Hmm.

But my ad copy's different.

So just on that, John, let's just say like we are looking at Ampire.

There's 35 asset groups.

There's one signal per asset group.

Would you pause all them and rebuild 'em out or start afresh

with a brand new Pax campaign?

Technically, for theory is true, we should be safe

. , assume all of 'em are on, we

37 and put all of the signals.

Yep.

Gotcha.

What's interesting is when we started looking at, and I look

at the insights all the same.

Four different asset groups.

I got four agers.

So now is our, what is our control here?

Well, nothing really, if we see.

One signal replicated four times four asset groups.

They're the same signal, four asset groups with different and copy.

That's the structure here.

So if we're gonna take our big bank of people divided by four,

shove all of 'em in different asset groups and see who wins is a signal.

No, it's the same exact end result.

Google sees that all of these people the same.

So, but now, but, but now, well actually, lemme just do this here.

General, ah, so general's got more clicks.

Okay, now , let's just keep that in our mind here in the beginning.

What about Paola?

Like, versions are on the general too.

Okay, interesting.

Doesn't have the most clicks though.

Good note.

Custom mail box, the asset group.

It's working better here.

Okay.

Product boxes, general wins again.

So what's interesting here is you see now that the asset groups are still the same,

but they're jumping around the users.

That's good.

Now I can possibly tie this back to what ad copy is being used and what,

what creative is being used in that asset group and what is the search term.

So we know that for custom cardboard boxes right now, it's

really working with general.

So Lake, ooh, what's in my general.

Custom cardboard boxes.

, let's edit the as verbs in general for your experience at space today.

How satisfied are you?

Very dissatisfied.

Google.

. , thank you for your feedback.

You're welcome.

So now we got cardboard boxes, is in the acid repairer.

do I have any cardboard box?

Not really.

Design your own box Custom.

Okay.

Maybe there's something here that can show, or maybe it has a quad

corrugated cardboard construction.

I don't know.

That's a search term that's winning right now in that asset.

Asset group.

Seems to be the line chair.

Would I change around the ad copy?

Would I change around the asset?

would I change around the energy?

Probably not.

But at least now I could look at the search term per ad group by,

based on that keyboard and my, , my responsive search ad and that

search campaign should reflect it.

That's kind of how I'm thinking about this old school tactic from

inbound search campaigns using search terms on the ad ad groups.

I'm kind of doing that with as groups.

I'm just letting Google dictate where he goes first, and I'll pause there.

Cause Glen looks like you wanna say something.

Theory using that data that you're getting from the Performance Max

campaign, you could say, well, cardboard boxes , is smashing it.

Would you create a search campaign just on targeting that keyword, like

with an ad group or a DSA to, well, I don't know cuz that's, you could