Latest / Scalability School / Heroes, Sidekicks, Villains & Zombies: Collin Slattery on Google Shopping

Transcript

Heroes, Sidekicks, Villains & Zombies: Collin Slattery on Google Shopping · Scalability School

0:00
The relationship that you have with
0:01
Google is inherently adversarial,
0:03
where it is not necessarily
0:04
adversarial with Meta.
0:05
And the reason it's adversarial with
0:06
Google is because Google is a
0:08
monopoly, and they don't really have
0:11
a direct impact on
0:11
search query volume, right?
0:13
Like the amount of people who are
0:14
going to search for tennis shoes in
0:15
any given month is
0:16
relatively set, right?
0:18
Like they can't really impact that.
0:20
And so brand is like this
0:21
stereotypical example, and we see it
0:23
with brands of all sizes, even from,
0:26
you know, seven-figure brands to
0:28
ten-figure brands.
0:29
You go into their Google Ads
0:30
account, and it's very rare that
0:33
brand and non-brand is
0:34
truly fully segmented apart.
0:37
And it has to be.
0:38
YouTube today is kind of like Meta
0:40
in 2015, where, you know, you do the
0:43
1% lookalike audience, right?
0:45
Like we on this call all old enough
0:47
to remember the old lookalike
0:49
audience stacks
0:49
and things like that.
0:51
So you do have to do
0:52
audience targeting there.
0:53
But that's kind of like the lowest
0:56
form of YouTube advertising.
0:59
And now let's take a listen to the
1:07
Scalability School podcast.
1:10
All right.
1:11
Welcome to another episode of
1:12
Scalability School.
1:13
Incredible guest today.
1:16
Longtime friend of the podcast.
1:18
Longtime friend of mine and somebody
1:20
who I deeply respect and love.
1:22
Colin Slattery, the Google Ads
1:24
expert of D2C
1:25
Twitter, I would say, right?
1:26
I mean, can we name you that?
1:28
What do we think about that?
1:29
I'll take it.
1:29
It's a moniker I'll
1:30
happily wear for sure.
1:32
The person that people are willing
1:34
to fight with about
1:35
Google Ads on the Internet.
1:37
Whether they are
1:38
employed by Google or not.
1:40
Yeah, exactly.
1:40
Who actually is smart and can
1:42
respond back to them.
1:44
Colin, so glad to have you.
1:46
Colin's been an expert in the
1:47
Founders community on Google Ads for
1:49
a really long time.
1:50
And then a couple of years ago,
1:52
hired the best in
1:53
the game, Phil Kiel.
1:55
Previous Scalability School guest.
1:57
Now has an absolute empire that is
2:01
being built at Tyken Digital.
2:03
So we're stoked for you about that.
2:05
Good job.
2:05
Thanks.
2:06
And thanks for having me on.
2:07
Excited to check Google.
2:10
Excited to hopefully dispel some
2:13
myths and share a little bit of
2:15
knowledge in the process.
2:15
Google is kind of like the ugly
2:17
stepchild of DTC Twitter.
2:19
People don't want to talk about it.
2:20
It seems like you're one of the few
2:21
people that want to talk about it.
2:23
And I feel like it needs to be
2:24
talked about more because I
2:24
personally have a ton of questions.
2:26
Because I'm curious
2:27
how you think through it.
2:29
I know you have strong
2:29
opinions, which is good.
2:31
Because I don't think enough people
2:32
have strong enough
2:33
opinions that are also thoughtful.
2:35
And you have both.
2:37
You have both.
2:37
So I'm excited to dig in.
2:39
I feel like there's a whole other
2:40
episode at some point.
2:41
We just talk about everything that
2:42
you're building agency-wise.
2:43
Maybe we've got to get Phil and the
2:45
gang back together and
2:46
just talk about that.
2:47
Get Jake on here too.
2:48
Yeah.
2:48
Yeah.
2:48
Yeah.
2:49
Might as well.
2:49
Yeah.
2:49
Jake the ad nerd.
2:50
The OG.
2:51
He's in the game now too.
2:52
Yeah.
2:52
We'll have to get the whole crew on.
2:54
But yeah.
2:54
Excited to jam Google today.
2:56
Yeah.
2:57
I think it's going
2:57
to be a lot of fun.
2:58
And excited to dive in.
3:01
Maybe some hot takes.
3:03
Or lukewarm takes anyway.
3:05
Well, we can kick it off with the
3:07
nice original great starting point.
3:11
I thought it would
3:11
be interesting too.
3:12
I know Zach, I was looking back on
3:14
this tweet from like almost a year.
3:15
It's almost a year ago now.
3:16
It was actually June of last year.
3:17
It feels like it was just yesterday
3:18
that Zach was shitting on
3:20
all Google ads agencies.
3:21
But I thought this was
3:22
a good starting point.
3:23
And so Zach's tweet, I'll
3:24
probably summarize this.
3:25
But Zach kind of started by saying,
3:27
if you're a DTC brand, being your
3:28
Google ads agency more
3:29
than 10K a month, you are very
3:31
likely getting robbed.
3:33
Zach is known to
3:35
take strong positions.
3:36
Usually with good reason.
3:38
Usually with good reason.
3:39
But sometimes he has strong takes.
3:41
And so we don't have to spend a
3:43
bunch of time dispelling this.
3:45
But just curious if you have a kind
3:47
of quick take on your
3:49
feeling of, I guess, maybe the
3:51
average Google ad agency and what
3:53
you guys do differently
3:54
that helps set you apart.
3:55
Well, obviously, I
3:57
think that's bullshit.
3:58
And I said as much around that time
4:00
and the engagement was fun.
4:02
And I know Zach loves to throw bombs
4:04
out there and then
4:05
turn around and walk away.
4:08
But yeah, no, I mean,
4:10
that's clearly not true.
4:11
Right.
4:11
And I actually would say that Google
4:14
in particular is one of the places
4:17
where investment in top
4:20
quality talent is even more
4:22
paramount because of the difficulty
4:25
and the number of dark patterns
4:27
that exist in the platform.
4:30
This gets to a foundational concept
4:32
around the differences
4:35
between Google and Meta.
4:37
And one of those is that the
4:39
relationship that you have with
4:40
Google is inherently adversarial,
4:42
where it is not necessarily
4:44
adversarial with Meta.
4:45
And the reason it's adversarial with
4:47
Google is because
4:48
Google is a monopoly.
4:50
And they don't really have a direct
4:52
impact on search query volume.
4:54
Right.
4:54
Like the amount of people who are
4:56
going to search for tennis shoes in
4:57
any given month is
4:59
relatively set.
5:01
Right.
5:01
Like they can't really impact that.
5:03
And they already
5:05
have the full market.
5:06
They can't get any
5:07
more market share.
5:08
And so the only thing that they can
5:10
do and speaking primarily around
5:12
search and shopping and
5:13
things like that is increase the
5:15
amount of money that they get paid
5:17
for the queries that
5:19
already exist.
5:20
That's the only lever that they have
5:22
to pull to increase
5:23
how much money they make.
5:24
They're a publicly traded company.
5:25
They're in the
5:26
business of making money.
5:27
Right.
5:28
And when they are increasing the
5:30
costs, that is born
5:32
by the advertiser.
5:34
Meta, on the other hand, has the
5:36
ability to impact their inventory in
5:39
a way that Google
5:40
does not.
5:40
Right.
5:41
They can increase the frequency of
5:44
ads in reels or your feed.
5:47
They can make the
5:48
product more addictive.
5:49
So people scroll deeper so they can
5:51
be exposed to more ads.
5:53
Right.
5:54
And so fundamentally, Google is not
5:56
playing nicely with advertisers
5:59
because it is in their
6:01
best interest to.
6:03
Take money from advertisers.
6:07
And so if you have someone who
6:08
doesn't know what they're doing and
6:10
doesn't know Google's
6:11
dark patterns, it is very easy for
6:14
Google to be an absolute waste of
6:17
money because that's
6:19
good for them.
6:20
Right.
6:20
Because they need to make money.
6:23
So if you don't have someone who's
6:25
excellent, and admittedly, there are
6:27
not that many excellent
6:28
people out there, it's going to be a
6:31
non-incremental waste fest.
6:34
And I think that's what most people
6:36
run into because they're working
6:39
with people who don't
6:40
know how to do it or
6:41
they're working with people who are
6:42
meta first and the rules between
6:44
meta and Google are completely
6:45
different.
6:46
And then they're
6:47
like, oh, Google's trash.
6:48
Everyone who does Google is trash.
6:50
And screw this.
6:53
I'm known for taking notes.
6:54
And the first thing I wrote down is
6:55
Zach equals skill issue.
6:57
So that's what I heard
6:58
from section number one here.
7:00
It is, in fact, a skill issue.
7:03
A hundred percent.
7:04
Good.
7:04
Good.
7:04
Well, I'm sure Zach will hear this
7:06
and we'll remind him.
7:07
But that segues nicely into like,
7:09
OK, so then what can brands that are
7:13
listening to this do and
7:15
take away to actually start to make
7:17
Google a channel that
7:19
is effective for them?
7:20
Because I think a
7:20
lot of people will.
7:21
I mean, our average, the average
7:23
client that we work with has Google
7:25
and there's brand set up in
7:28
some capacity.
7:28
Whether or not it's incremental,
7:30
there's a bunch of
7:30
questions around that.
7:31
And yeah, right.
7:33
And you mentioned you mentioned
7:34
there's there's probably different
7:36
things to consider, certainly
7:38
across industries and how you should
7:39
consider across industries, but even
7:41
account spend and
7:42
size.
7:42
And so I think if you could just
7:45
start to kind of take people through
7:46
like what what is the
7:47
actual account setup
7:48
and structure look like?
7:49
How does it change as you scale?
7:51
I think that could be a really
7:52
interesting starting point to just
7:53
kick off, you know, where people
7:54
can start to get
7:55
some tactical takeaways.
7:56
Yeah, and I think this is this gets
7:59
back to the size and
8:02
complexity of Google, which is
8:03
material.
8:04
It's a materially more complex
8:05
platform than meta is.
8:07
There's a lot more stuff going on
8:08
and there's a lot of
8:09
stuff going on that.
8:12
Is beneficial to Google, but
8:14
detrimental to the advertiser.
8:15
And I think if people are.
8:20
Looking to succeed on Google, I
8:22
think that is a mindset shift that
8:24
needs to happen where
8:26
when you are in there, the.
8:30
Everything you look at, you need to
8:31
think this is set up in order to
8:33
benefit Google and
8:34
not benefit me.
8:36
So you should be very skeptical of
8:38
anything that Google recommends.
8:39
Right.
8:39
And we're kind of that way with meta
8:41
when they're like auto apply all of
8:43
these AI enhancements
8:45
and shit like that.
8:46
But those are much more visible in
8:49
meta than they are in Google.
8:51
And so you absolutely have to be
8:54
approaching it with that mindset and
8:57
trying to make it
8:59
harder for Google to succeed.
9:03
And so brand is like this
9:04
stereotypical example.
9:05
And we see it with brands.
9:08
Of all sizes, even from, you know,
9:12
seven figure brands to 10 figure
9:13
brands, you go into their
9:14
Google ads account and it's very
9:17
rare that brand and non brand is
9:19
truly fully segmented
9:22
apart.
9:23
And it has to be because if you say
9:26
I want a five X right.
9:28
And you put brand in there, right,
9:33
that's going to generate a 20 X.
9:35
And then Google is
9:35
like, cool, we got a 20.
9:37
We banked the 20 X here.
9:39
That means we can waste money on
9:41
stuff that doesn't perform.
9:42
It'll average out to a five.
9:44
And the advertiser is
9:46
happy because it says a five.
9:48
But they haven't dug that layer
9:50
below and seen all those conversions
9:53
are happening to brand.
9:55
80% of the budget is 10% of the
9:58
conversion value and 10% of the
9:59
budget is 90% of the
10:01
conversion value.
10:02
And blended, it looks great.
10:04
They run a holdout test on it.
10:05
Like, hey, this is
10:06
not incremental at all.
10:07
Like we chopped our Google and
10:09
nothing happened.
10:10
It's like, well, yeah, of course.
10:11
It's because your brand and non
10:13
brand are blended together.
10:14
And so if there is one thing that
10:18
people take away, get them split
10:21
apart completely, right?
10:23
And split in search.
10:25
So negative it out in
10:27
all your non brands.
10:28
If you want to run it, put in its
10:30
own manual CPC search campaign.
10:33
But you have to do it
10:33
on shopping too, right?
10:35
Like shopping inventory serves
10:36
against brand queries.
10:38
And so if you're not segmenting out
10:41
branded queries and shopping,
10:43
if you don't have a
10:44
branded shopping campaign,
10:45
all that's going to roll into your
10:47
Pmax is going to roll into your
10:48
regular shopping.
10:49
And it's going to be the same
10:50
problem that you would have with
10:51
search as Justin Shopping.
10:53
That is, I think, the biggest
10:55
problem and also the
10:58
easiest problem to fix.
11:00
Is that a one-time thing?
11:01
Or do you have to
11:01
check it regularly?
11:02
Because Google is trying to jam
11:03
brand back into things
11:04
or brand adjacent things.
11:06
So if you do a good job with your
11:08
negative keyword list,
11:10
you're going to get 99% of it out.
11:14
It's not going to
11:15
cover all misspellings.
11:16
And so it does make sense so much in
11:17
a while go through and negative out
11:19
the misspellings.
11:21
When you launch new campaigns,
11:22
you do need to apply the negative
11:23
keyword list to the campaign.
11:26
But it's mostly a one-time, 30
11:31
minutes max sort of activity.
11:35
So there's a lot of brands.
11:38
Brad was saying there's a lot of
11:38
small brands or smaller brands that
11:40
listen to this podcast.
11:41
And I think, you know, have always
11:44
wanted to test Google.
11:45
And I think people
11:45
get excited about brand.
11:49
They get excited about launching
11:50
Google and they get excited about
11:51
launching YouTube ads as well.
11:54
So when you're looking at those
11:57
brands that you've
11:58
worked with in the last,
12:00
let's just say 12 months, because I
12:02
know it changes over time.
12:04
What are they doing?
12:05
What's the mix of brand, non-brand?
12:08
Are they doing shopping?
12:10
Are they doing YouTube?
12:12
What does that mix look like?
12:13
And what are the things that you've
12:14
seen be truly incremental to their
12:17
results and their growth?
12:19
I'm going to say something, and
12:20
maybe this is my first hot take,
12:21
even though it's not a hot take
12:22
because it's driven by data.
12:23
Google and YouTube done well are the
12:27
most incremental advertising that
12:29
any brand will ever have,
12:30
period, full stop, done.
12:33
We've done this with nine-figure
12:34
brands where 87% of the traffic
12:38
coming from YouTube is
12:39
net new to the brands,
12:41
whereas Meta is
12:41
sitting around like 60, 62%.
12:43
Google and YouTube are so
12:48
incremental and expose so many
12:52
people to traffic
12:55
that is nowhere else.
12:58
There are 34 million Americans that
13:01
are not on Meta properties,
13:04
are not on TikTok,
13:06
are not in any other
13:07
digital properties,
13:08
but YouTube.
13:09
Literally, the only place you can
13:11
reach these 34 million
13:12
people are on YouTube.
13:14
Totally net new,
13:15
do not exist anywhere else.
13:17
They're only on YouTube.
13:19
And you can reach them there, right?
13:20
And so,
13:22
and that's just the people that
13:23
aren't anywhere else, right?
13:24
It's not even counting the people
13:25
that you don't reach
13:26
because they're not in a
13:26
conversion-focused audience
13:27
in Meta or anything like that.
13:30
So,
13:32
massively, massively, massively,
13:34
massively incremental.
13:35
It's just set up
13:36
poorly most of the time.
13:38
Now,
13:39
to kind of like loop
13:40
back and it's like,
13:40
okay, well, how do you
13:41
make Google incremental?
13:43
Well,
13:44
one is,
13:44
it's get brand and non-brand
13:46
separated, right?
13:48
And then what you're doing is
13:49
different based on the scale
13:51
of the brand and where
13:52
your budget allocation is.
13:53
The first thing I'd say is like,
13:56
don't run brand at all.
13:57
Like negative it out and
13:58
then don't have it running,
14:00
period.
14:01
And I think Zach will probably agree
14:03
with me on that, right?
14:04
It's generally a waste of money.
14:06
And it's one of the very first
14:08
holdout tests that we run
14:09
with any client that comes on,
14:12
whether it's Meta in
14:13
Google or just Google only,
14:14
is like,
14:15
let's test brand,
14:17
like see if it's incremental
14:17
because it's usually not.
14:21
And it's something
14:21
you can just turn off,
14:22
which is great, right?
14:23
If you can allocate
14:24
that budget to things,
14:25
put it as a YouTube shorts instead.
14:27
Like what a better way to do it.
14:30
So if you can get away
14:32
with not running brand,
14:33
don't run brand.
14:34
And then the structure is like,
14:39
it's going to be pretty consistent,
14:42
but the budget
14:42
allocations will be different.
14:43
So non-brand shopping or like a
14:48
non-brand shopping Pmax,
14:51
right?
14:52
Non-brand search.
14:53
And then if you are running brand,
14:55
brand search,
14:56
brand shopping,
14:56
and then some degree
14:58
of demand gen shorts
15:01
as kind of like the
15:03
third component of that.
15:05
So,
15:05
you know,
15:06
we're looking at
15:07
three to five campaigns
15:08
as kind of like the
15:10
default structure.
15:12
When do you,
15:13
so then,
15:14
okay,
15:14
so that's for everybody.
15:16
So what is scaling it out look like?
15:18
Is it horizontally
15:19
adding more campaign
15:20
by product,
15:21
product category?
15:22
Is it just scaling it
15:23
up by cranking budgets
15:24
and sending it to the moon?
15:26
Obviously,
15:27
YouTube is probably
15:27
handled fairly differently
15:29
because there's a
15:29
creative component to it,
15:31
whether that's shorts
15:32
or just,
15:32
you know,
15:33
general YouTube placements.
15:35
So you can pick either side,
15:37
maybe Google is the place to start
15:38
and then work over to YouTube.
15:40
But like,
15:40
what is actually,
15:41
what does it look
15:42
like to start scaling up
15:44
the Google side?
15:45
Yeah.
15:45
So I think
15:46
the easiest scale to
15:49
unlock is shopping,
15:50
right?
15:50
It's the shopping inventory.
15:53
And it's easy,
15:56
but it's not simple
15:58
because what
15:59
dictates how shopping works
16:00
is really the structured data
16:02
that you're giving Google,
16:05
right?
16:05
So it's the quality
16:06
of your product feed.
16:08
It's the descriptions.
16:09
It's the images.
16:11
It's the secondary images.
16:14
The color data,
16:15
the size data.
16:16
It's how optimized your titles are,
16:19
right?
16:19
It's making sure that
16:20
there aren't price mismatches
16:21
or any other issues
16:22
that are impacting
16:23
the quality of the data
16:26
that is being fed to Google
16:27
because that's how Google decides,
16:29
oh,
16:31
this is,
16:31
this is,
16:33
you know,
16:33
a blue sneaker,
16:35
right?
16:35
It's,
16:36
well,
16:36
you have to give them all that data.
16:37
And so if someone
16:37
searches for blue sneaker,
16:39
right,
16:39
you want to show up for that.
16:40
And so the only way
16:42
you're going to show up
16:42
is if you have good,
16:43
clean data.
16:44
And reviews are another huge
16:46
component of that.
16:47
People are going to go,
16:49
they're going to
16:50
search for blue sneakers.
16:52
What's going to get surfaced
16:53
and what's going to get clicked on.
16:54
Stuff that has good data
16:55
so that Google knows
16:56
that's what you're selling.
16:57
And then things that
16:58
have a lot of reviews
16:59
because that's a trust signal
17:01
and people are going to click on it.
17:02
And so feeding Google good data
17:06
is going to impact
17:08
how much shopping volume
17:12
you're going to be able to acquire.
17:13
And then there are other parts
17:17
that obviously brands can't control.
17:18
So like if you,
17:19
the larger your SKU library,
17:21
the greater the
17:22
surface area you can cover.
17:24
And so generally
17:24
the higher the scale
17:26
that you can conceivably get.
17:30
Similarly to that,
17:32
the better your
17:32
business economics are,
17:34
right?
17:34
If you think about
17:35
search impression share,
17:36
like a pie
17:37
or a hundred percent of search
17:39
impression share,
17:40
like no business is going to be able
17:43
to economically monetize
17:44
a hundred percent of search
17:45
impression share.
17:46
But if your
17:47
economics are not so good,
17:48
right?
17:48
If your revenue per session
17:50
is generally pretty low
17:51
or your margins are pretty low,
17:53
like the percentage of
17:55
the search impressions
17:57
where you can profitably target it,
18:00
it's going to shrink.
18:01
And so
18:02
your business
18:05
metrics also like dictate
18:06
kind of the scale
18:07
that you can achieve
18:08
on shopping as well.
18:10
But
18:12
there's a ton of scale
18:13
in shopping.
18:15
Friends,
18:16
quick break.
18:16
This episode of Scalability School
18:17
is brought to you by
18:18
Northbeam Incrementality,
18:21
which honestly is so awesome.
18:23
Let me just tell you
18:23
a little bit about it.
18:24
If you're a marketer,
18:25
you probably agree
18:26
incrementality testing is broken.
18:28
It's insane.
18:29
You got to do all this crazy stuff.
18:31
But with Northbeam,
18:31
you do it right in the platform.
18:33
It builds tests
18:34
using your own MTA data,
18:36
launches them directly
18:37
in the ad platform
18:38
and automatically monitors them
18:40
throughout the test.
18:42
And if something threatens validity,
18:44
you're alerted to it
18:45
within Northbeam
18:46
before it ruins the results.
18:49
The really nice thing
18:50
is it solves all these problems
18:51
you hear about
18:52
incrementality testing
18:53
all the time,
18:54
like labor,
18:55
spreadsheets,
18:55
statistical reference,
18:56
designing tests
18:57
is normally so slow,
18:58
right,
18:59
and error prone.
19:00
So with Northbeam Incrementality,
19:01
it's automated and accurate
19:03
and your team
19:04
is able to spend
19:05
with confidence.
19:07
Give it a shot
19:07
with Northbeam Incrementality.
19:10
Really,
19:10
the future of incrementality
19:11
is here
19:12
and Northbeam
19:12
has figured out a way
19:13
to make it easy.
19:15
So, I mean,
19:15
it's interesting,
19:16
you know,
19:16
I've never heard somebody
19:17
give the case
19:18
for shopping that heavily,
19:19
so that's interesting.
19:20
I think you've had this thing
19:21
about called
19:21
hidden search terms
19:23
and you have a framework
19:24
of Google inventory,
19:27
which is premium,
19:29
mid-tier,
19:29
and garbage.
19:30
And that's what,
19:31
this is,
19:32
you know,
19:32
taken from your word.
19:33
So talk about
19:34
hidden search terms
19:35
and the research
19:36
around this
19:37
and what things
19:39
folks can do
19:39
to bring some
19:40
of these search terms
19:41
out that maybe
19:42
they weren't,
19:43
you know,
19:44
aware,
19:44
ever familiar with
19:45
in terms of targeting.
19:47
And this is,
19:47
again,
19:48
this gets to the whole,
19:49
there's additional layers
19:50
underneath Google
19:52
that when you just
19:53
look at it,
19:55
it's painting
19:56
a very small picture
19:58
of what's actually
19:59
going on.
20:00
so keywords
20:02
and search terms
20:03
are not even
20:05
remotely the same thing,
20:07
right?
20:07
It's like,
20:07
oh,
20:08
well,
20:08
you know,
20:09
I'm targeting blue shoes.
20:11
Like,
20:11
that's the keyword
20:12
that I'm targeting.
20:12
That's what I'm bidding.
20:13
It's just like,
20:14
no,
20:14
that can match
20:15
to hundreds
20:16
of different search queries
20:17
that are somewhat related
20:20
to completely unrelated
20:22
depending on your match type,
20:24
right?
20:24
And so those are
20:25
the search terms
20:25
that surface.
20:27
And so those are the things,
20:28
search terms are what people
20:29
are actually typing
20:30
into Google to find
20:31
and then your ad
20:32
is surfing against,
20:33
right?
20:34
And Google,
20:35
up until,
20:36
I think it was probably
20:36
2020,
20:37
like 2021,
20:38
I can't remember
20:39
the exact year.
20:39
At this point,
20:40
it was a while ago.
20:41
They used to show
20:42
95%
20:45
of all of the terms
20:47
that people typed in
20:48
to,
20:49
that surfaced your ad.
20:50
So you'd have
20:51
95% visibility,
20:53
90 to 95% visibility
20:54
in everything
20:55
that people searched for
20:55
that your ad surfed against.
20:57
And for privacy reasons,
21:01
they decided
21:03
to hide
21:04
a huge percentage
21:06
of the search terms
21:08
that people searched for
21:09
that your ad surfed against.
21:10
And they lumped them together
21:11
in what is known
21:12
as other search terms,
21:14
which you can't see
21:16
what they searched for.
21:17
You can only see
21:18
a single line
21:19
of the aggregated performance
21:20
of what is hidden.
21:22
And that's a problem,
21:23
right?
21:23
Like you're paying for it.
21:25
You should know
21:25
what people are searching.
21:26
And the performance
21:27
of the stuff
21:28
that is hidden
21:28
is generally
21:29
worse than
21:31
what is visible.
21:33
And that's because
21:34
in my estimation,
21:37
right,
21:37
it behooves Google
21:38
to hide the stuff
21:39
that people would
21:40
try to exclude
21:41
if it's not related.
21:44
Right?
21:44
If my keyword is,
21:46
you know,
21:46
blue shoes
21:47
and,
21:49
you know,
21:50
the matching is
21:51
blue tablecloth,
21:53
like,
21:54
that's
21:55
not related.
21:56
Right?
21:57
But you can get
21:58
matches that are
21:59
that far away.
22:01
It's actually
22:02
pretty shocking
22:03
how not close
22:05
search terms
22:06
can be to the keyword
22:07
that triggered them.
22:08
And now that they
22:09
can hide those,
22:11
it's easier for you
22:12
to keep bidding on them
22:13
as opposed to
22:14
historically,
22:15
you know,
22:16
turn them into
22:16
negatives and remove them.
22:18
Yeah.
22:18
You put together
22:19
like a pretty
22:20
gnarly report,
22:21
like actually
22:21
digging into this.
22:22
Do you have any
22:23
of the numbers
22:23
like handy
22:24
out of the top
22:24
of your head
22:24
that you can
22:25
just like rip off?
22:26
Or is that
22:27
a little bit
22:27
too specific?
22:28
I mean,
22:29
do you want,
22:29
I have the report
22:30
somewhere probably.
22:31
Pull it up.
22:31
Yeah.
22:32
If it's not,
22:32
if it's not a huge
22:33
pain to pull up,
22:34
I'd be super curious
22:34
to dig into that
22:35
because I remember
22:36
reading it
22:36
I'm pretty sure
22:37
that,
22:37
is that what
22:38
Ginny responded to?
22:40
And I do want to note
22:42
this is only for
22:44
search campaigns
22:46
and shopping campaigns
22:47
where you have
22:48
full search
22:49
career report
22:49
data available.
22:50
But we looked at
22:51
around $20 million
22:52
in ad spent
22:54
in search
22:55
and shopping spent
22:56
and found that
22:57
for every
22:59
dollar of ad
23:00
spent
23:01
that you are
23:03
running through Google,
23:04
they're skimming
23:05
almost a full dollar.
23:06
off of your efficiency
23:09
with hidden search terms.
23:10
So,
23:11
if you were to spend
23:14
a million dollars
23:16
on ads
23:17
and everything
23:18
was visible,
23:19
you would generate
23:19
about $5.6 million
23:20
in revenue.
23:21
In reality,
23:22
you're only generating
23:23
about $4.75
23:24
million in revenue.
23:26
And that $850,000
23:27
shortfall
23:28
is due to the
23:29
inefficiency
23:30
of
23:31
hidden search terms.
23:33
Yeah.
23:34
So you're looking at
23:34
like
23:34
a 5.6x
23:36
for visible
23:37
versus
23:39
a
23:39
4.75
23:40
for
23:41
full blend
23:42
including
23:43
hidden.
23:44
Like worst case scenario,
23:45
historically,
23:46
before this change,
23:47
you could just
23:47
exclude them
23:48
and,
23:49
okay,
23:49
I've excluded them,
23:50
I'm not wasting money
23:50
on them.
23:51
And now,
23:52
it's just being
23:53
lumped in there.
23:54
And so,
23:55
what you're saying
23:56
is best case scenario,
23:57
you can actually
23:57
exclude them
23:58
and then the additional
23:58
spend,
23:59
you'd be able to continue.
24:00
You're spending that money
24:01
anyway,
24:01
so it's going to be
24:02
spent at that more
24:03
efficient return number.
24:06
Am I understanding
24:06
that correctly?
24:08
Well,
24:08
I mean,
24:08
you can't get rid
24:09
of the stuff that's hidden,
24:10
right?
24:10
For sure.
24:11
Stuff is always
24:11
going to be hidden
24:12
and it is a game
24:14
of whack-a-mole
24:15
to consistently
24:16
exclude things
24:17
to try and pull
24:17
up efficiency numbers.
24:18
You can aggregate
24:20
data through
24:21
what's known
24:22
as Ngram analysis
24:23
to pull stuff
24:25
out of hidden,
24:27
like,
24:27
and exclude stuff
24:28
that would otherwise
24:29
be lumped into
24:30
hidden search terms.
24:32
Generally,
24:33
hidden search terms
24:33
are stuff
24:35
that Google
24:36
categorizes
24:38
as, like,
24:38
low volume,
24:39
right?
24:40
So it's not getting
24:41
a lot of volume
24:42
against those queries
24:43
and so those
24:45
are generally
24:45
going to be
24:45
longer tail
24:46
search queries
24:47
and by longer tail,
24:48
I mean,
24:48
more words
24:49
in the query.
24:51
And so
24:51
you can aggregate
24:53
what is visible
24:54
and find terms
24:55
that are
24:56
low performing
24:57
and when you
24:58
exclude those,
25:00
it will also
25:01
generally pull out
25:02
stuff that was
25:04
hidden
25:05
that also included
25:06
that term
25:06
that is
25:08
not effective.
25:10
Not usually
25:10
going to get
25:11
like a lower
25:13
percentage
25:14
of hidden
25:15
but you can
25:16
reduce the
25:18
inefficiency
25:18
overall
25:19
by going
25:21
through and
25:21
pulling stuff
25:22
out.
25:23
Cool.
25:23
Personally,
25:24
I'm very
25:24
interested in
25:25
like the
25:25
actual
25:26
tactical levers
25:27
of,
25:28
okay,
25:28
we can dig
25:28
into shopping.
25:29
Andrew,
25:30
if you're cool
25:30
with going
25:31
down that route,
25:32
I'm just going
25:32
to start ripping
25:32
some questions.
25:34
Absolutely.
25:34
Feel free to
25:35
take it wherever
25:36
direction you want
25:37
to.
25:37
So you mentioned
25:38
you've kind of
25:38
got like four
25:40
to five campaigns
25:40
that are standard.
25:41
I think we'll
25:43
ignore the brand
25:43
side of this
25:44
completely for now
25:44
because I think
25:45
your basic
25:45
TLDR is like
25:47
turn it off
25:47
or highly
25:48
limit the spend
25:49
that's going
25:50
into it at all
25:50
if you're going
25:51
to spend into it
25:52
and just be
25:53
thoughtful around
25:53
that.
25:54
If you're in a
25:55
vertical where you
25:55
need to spend
25:56
on brand,
25:57
you'll know
25:57
because it's
25:58
the exception.
26:00
Fair.
26:01
Cool.
26:01
Maybe worth
26:02
coming back to
26:03
that because
26:03
that's interesting.
26:04
But then shopping
26:06
is where you're
26:07
seeing the majority
26:07
of the volume.
26:08
You said every
26:09
brand could spend
26:09
six figures.
26:10
It's kind of
26:11
like a take.
26:12
So as people
26:13
are thinking about
26:14
setting up shopping,
26:15
you said the
26:16
biggest lever
26:16
is kind of
26:17
just like the
26:18
inputs to Google
26:19
which is
26:20
functionally the
26:20
feed and maybe
26:21
some offshoots
26:22
of that.
26:22
So what are
26:23
brands doing wrong
26:24
when they set up
26:25
their shopping?
26:25
You mentioned
26:26
the title and
26:27
the actual data
26:28
is not set up
26:29
properly.
26:30
So let's say
26:31
that they are
26:31
starting to chip
26:32
away at updating
26:33
that and maybe
26:34
their 80%
26:35
of what you
26:35
call like this
26:36
is a good
26:37
input.
26:38
The feed is
26:39
solid.
26:39
What's like
26:40
what's still
26:40
missing?
26:41
And then if
26:42
they wanted to
26:43
increase spend
26:43
volume through
26:44
shopping,
26:44
like what are
26:45
they actually
26:45
doing?
26:45
It is actually
26:46
I should say
26:47
is one of the
26:47
biggest challenges
26:48
with Google
26:49
is because
26:50
everything is
26:53
different depending
26:54
on small
26:56
changes in a
26:57
brand, right?
26:57
Like if you
26:58
have 10,000
26:59
SKUs, like
27:00
your approach
27:01
is going to be
27:02
completely different
27:03
than if you
27:04
have 50 or
27:05
100 SKUs
27:06
or 10 SKUs,
27:07
right?
27:08
And so the
27:10
feed is table
27:11
stakes, right?
27:12
And having that
27:13
fully optimized
27:14
and I will say
27:14
like most people
27:16
just connect their
27:16
Shopify feed to
27:17
Google Merchant
27:18
Center and
27:18
they're like,
27:19
I'm done,
27:19
right?
27:20
Like, no.
27:22
And if that is
27:22
you, like, go get
27:24
data feed watch.
27:25
It's not that
27:25
expensive.
27:26
It's going to
27:27
allow you to
27:28
really optimize your
27:28
Google fields,
27:29
which is super
27:31
important.
27:32
And we're
27:32
going to
27:32
assume you
27:32
have reviews
27:33
piping in
27:33
because that's
27:34
also very
27:34
important.
27:35
Shopping from
27:36
a structural
27:36
level and I
27:37
think Google
27:38
in general,
27:40
and this is
27:41
something that
27:42
we've seen
27:42
time and time
27:43
again on
27:44
search, on
27:45
shopping,
27:45
absolutely on
27:46
YouTube.
27:47
Google is not
27:48
nearly as
27:49
efficient a
27:50
budget allocator
27:51
as Meta.
27:53
And so, so
27:54
many people come
27:55
from the Meta
27:56
space where it's
27:56
like, trust
27:57
Meta, they
27:58
know what they're
27:59
doing, go
28:00
broad, like drop
28:01
it all
28:01
together, right?
28:04
You cannot do
28:05
that with Google.
28:06
You're going to
28:06
have a bad
28:07
time if you're
28:08
doing that with
28:08
Google.
28:09
And that
28:10
extends to
28:12
shopping as
28:13
well.
28:13
And then you
28:15
need segmentation
28:16
and you need
28:18
thoughtful segmentation
28:19
in order to
28:20
unlock volume.
28:22
What this
28:23
segmentation
28:23
looks like is
28:25
really dependent
28:26
on number of
28:28
SKUs and
28:30
business
28:31
priorities.
28:31
If you're
28:32
a retailer
28:34
and you
28:34
carry a
28:35
hundred brands,
28:36
your shopping
28:37
structure is
28:38
going to look
28:38
materially different
28:39
than if you're
28:39
a supplement
28:40
company and
28:41
you've got
28:41
20 SKUs.
28:42
Just try the
28:42
example,
28:43
though, Colin,
28:43
of let's say
28:44
there's this
28:45
lower amount
28:45
of SKUs,
28:46
20 SKUs.
28:47
How are they
28:48
structured?
28:48
Are they
28:48
structured based
28:49
on which
28:50
one's driving
28:51
the best
28:53
revenue for
28:54
you with the
28:54
best gross
28:55
margin?
28:55
How do you
28:57
start to
28:57
structure that?
28:58
I think the
28:59
initial structuring
29:00
that you're
29:00
going to want
29:00
to do,
29:01
and again,
29:01
this is a
29:01
phase one,
29:02
is you have
29:04
one campaign
29:05
and you have
29:06
ad groups
29:07
in the
29:08
shopping campaign
29:09
that are
29:10
segmented either
29:11
by a
29:12
category or
29:13
if you have
29:14
really very
29:15
few products,
29:16
you can even
29:17
segment them
29:18
individually.
29:19
So you're a
29:20
supplement company,
29:21
you have
29:22
supplements for
29:24
digestion and
29:25
supplements for
29:26
sleep, right?
29:27
You can break
29:27
them up into
29:28
those individual
29:29
categories and
29:30
then make sure
29:31
that you have
29:32
negative keywords
29:33
at the ad
29:33
group level to
29:35
make sure that
29:35
you're routing
29:36
queries to the
29:36
correct ad
29:38
groups.
29:38
So that's kind
29:39
of like a
29:39
we're just
29:40
getting started
29:42
structure.
29:43
And then
29:44
once you have
29:46
that running,
29:46
you're going to
29:47
start to see,
29:47
okay, this
29:49
product is
29:49
potentially working
29:50
really well and
29:51
performing really
29:52
well.
29:53
Maybe that's
29:54
something where
29:55
you actually
29:55
can, and
29:56
we've done
29:57
this, like
29:57
we'll have a
29:59
shopping campaign
30:00
with just a
30:00
single product in
30:01
it so that
30:02
it's not
30:03
competing for
30:05
budget with
30:06
the other
30:07
products in
30:08
that campaign.
30:10
It's not super
30:10
common where
30:11
we'll fully pull
30:12
out an
30:12
individual product,
30:13
but it'll
30:14
happen.
30:15
Over time,
30:16
breaking the
30:17
campaigns apart
30:18
based on by
30:20
performance and
30:21
like bucket
30:22
them together
30:23
really into
30:24
like four
30:25
main structures
30:27
which are like
30:27
high spend,
30:28
high performing,
30:30
high performing,
30:31
low spend,
30:33
just low
30:35
performing,
30:36
and then like
30:36
no spend,
30:38
right?
30:39
And so,
30:40
you know,
30:41
you don't want
30:42
like your top
30:43
10% of SKUs
30:44
to be competing
30:46
for budget
30:46
with the 80%
30:49
of SKUs
30:50
that get 2%
30:51
of the total
30:51
clicks,
30:52
right?
30:52
It's historically
30:54
called, you
30:56
know, heroes,
30:56
sidekicks,
30:57
villains, and
30:58
zombies, right?
30:58
It's kind of
30:59
the four
30:59
monikers for
31:01
the different
31:01
structures.
31:03
And you
31:03
break them
31:04
all apart.
31:05
And SKUs
31:05
can shift in
31:06
between the
31:07
different campaigns,
31:08
but the
31:09
reasoning behind
31:11
that is
31:12
Google's budget
31:13
allocation is
31:14
not incredibly
31:15
efficient.
31:16
And if
31:18
you're lumping
31:18
them together,
31:19
you're just
31:19
leaving money
31:20
on the table.
31:22
But the
31:23
first segmentation
31:24
starts in
31:25
an individual
31:25
campaign at
31:26
kind of the
31:27
accurate level
31:27
by grouping
31:28
products separately.
31:29
And grouping,
31:30
does grouping
31:30
the products
31:30
on its own
31:31
give to your
31:33
point?
31:33
Because it
31:34
helps Google
31:34
sort out
31:36
delivery a bit
31:36
better because
31:37
you're inserting
31:38
kind of an
31:38
opinion into
31:39
how that should
31:40
be structured.
31:41
Does that
31:41
alone add
31:42
more volume?
31:43
Or is it,
31:45
hey, we need
31:45
to split these
31:45
things up and
31:46
then we go
31:46
edit product
31:48
titles to have
31:49
more, you know,
31:49
like, are we
31:50
trying to increase
31:50
the search volume
31:51
by getting more
31:52
keywords?
31:52
Like, I'm like,
31:53
very tactical as
31:54
somebody who's not
31:56
running Google
31:56
accounts.
31:57
Like, so you're
31:59
segmenting those
31:59
things, but like,
32:00
what else is the
32:01
team doing on a
32:01
weekly basis to
32:02
try and drive
32:03
more volume?
32:05
Yeah, so I
32:05
mean, you're also
32:06
removing inefficient
32:07
search terms,
32:09
right?
32:10
that's just an
32:10
ongoing negative
32:12
keyword application
32:15
is an ongoing
32:16
piece of work
32:18
in Google,
32:19
on shopping,
32:20
on search,
32:20
everywhere.
32:21
Product optimization,
32:22
title testing,
32:23
that is also a
32:24
key component of
32:25
it because,
32:27
again, like,
32:28
you operate
32:29
under the
32:30
understanding that
32:32
there is a
32:32
finite amount
32:33
of search volume
32:35
for whatever it
32:36
is that you're
32:37
doing, right?
32:38
And when you
32:39
find things that
32:39
work, you want
32:41
to be getting a
32:42
high, as high a
32:43
click share as you
32:43
possibly can.
32:45
So optimizing to
32:47
try and maximize
32:47
your click share,
32:49
generally, like,
32:50
it's through titles,
32:50
it's through more
32:52
reviews, price,
32:55
obviously, also a
32:56
component to it,
32:57
using promotional
32:58
things, promos to
33:00
get the sort of
33:02
slash through and,
33:04
like, sale tags and
33:05
things like that.
33:05
And then adjusting
33:06
targeting, right?
33:07
And that's one of
33:07
the added benefits
33:08
there, too, is,
33:09
like, if you have
33:10
things broken up,
33:11
you can have a
33:12
specific target for,
33:14
you know, mascara,
33:16
whereas blush has a
33:18
completely different
33:19
target and you can
33:20
control those
33:21
efficiency targets
33:22
individually.
33:23
Yeah.
33:24
What role do
33:25
images play in
33:26
shopping?
33:26
Like, are you guys
33:27
testing images
33:28
consistently or is it
33:29
not a big lever?
33:31
Generally, like,
33:31
pretty strict
33:32
limitations around
33:34
images in
33:35
Google.
33:37
rules around, like,
33:39
not having texts
33:40
on the images,
33:41
like, not putting
33:42
promotional information
33:43
on images in the
33:45
feed.
33:46
So most things do
33:48
default to
33:49
white box.
33:50
You can include
33:50
lifestyle images
33:51
as well.
33:53
There are fields for
33:54
lifestyle images that
33:56
always worth doing.
33:57
Basically, and this is
33:59
another general rule,
34:00
if there's an
34:00
optional field that
34:01
you can use, it's
34:03
not optional.
34:03
Just use it, right?
34:05
Like, the more
34:06
optional things you
34:07
include, the better
34:08
your feed score is
34:10
going to be in
34:11
Google's estimation
34:12
and the more
34:14
opportunity you have
34:15
to surface.
34:16
So there's not a
34:18
ton that you can do
34:19
from an image
34:20
standpoint in terms
34:21
of testing.
34:22
You can do a
34:24
little bit, but
34:26
it's tough.
34:28
We've had clients
34:29
where their product
34:30
is in a box and the
34:32
box has text on it
34:33
and they'll reject
34:34
the image because
34:37
of the amount of
34:37
text that's on the
34:38
box.
34:39
I hate to host this
34:39
phrase, but it's kind
34:40
of phrased this way
34:41
intentionally, but like,
34:41
is there anything
34:42
hacky in shopping that
34:43
you can do that
34:44
consistently works?
34:45
Like, you know,
34:46
there's this theme on
34:48
Meta for a while
34:48
where everybody did
34:49
the we're sorry kind
34:51
of ad in that
34:51
positioning.
34:52
Like, there's a
34:52
format that pops up
34:53
that you can just
34:54
run that playbook
34:55
across brands and it
34:56
works very consistently.
34:57
Is there any
34:58
comparable on the
34:59
Google side, whether
35:00
it's like hacky
35:01
feeling?
35:01
Maybe it's specific to
35:02
shopping, maybe not.
35:03
Not really.
35:05
And I would say
35:06
that in general,
35:07
like, it's not
35:09
how we like to
35:11
approach advertising.
35:13
Whatever arbitrage
35:14
exists, I mean,
35:15
you're opening
35:15
yourself up to
35:17
potential issues
35:18
down the road.
35:20
You know, like,
35:23
there was a time
35:23
where you could
35:24
like, get promo
35:26
text onto images
35:28
by, you know,
35:31
making it less
35:31
obvious and
35:32
tricking the
35:33
sort of machine
35:34
to think it
35:34
wasn't there,
35:35
even though it
35:35
clearly was.
35:37
And you're just
35:38
going to get your
35:38
feet suspended
35:39
and you're going
35:39
to have to redo
35:40
all the work.
35:40
So, you know,
35:41
I generally
35:42
operate on the
35:43
do things the
35:45
right way
35:45
principle and
35:47
you're going to
35:48
be in good
35:49
shape.
35:50
Brad's always
35:50
trying to hack
35:51
an idea, you
35:52
know.
35:52
Look, I'm just,
35:53
you know, you
35:54
never know what
35:55
exists to me.
35:55
I'm the anti-hack
35:56
guy.
35:56
I'm the let's
35:57
do it by the
35:58
book and, you
36:00
know, always be
36:00
on the right
36:01
side.
36:02
I also lean that
36:02
direction, but,
36:03
you know, sometimes
36:04
people like, you
36:05
know, a little
36:05
sauce every once
36:06
in a while.
36:07
Do not invite the
36:08
eye of Sauron,
36:09
you know.
36:10
That's fair.
36:12
So let's just
36:13
talk about YouTube
36:13
briefly in going
36:16
through that, right?
36:16
I think this is a
36:18
place where people
36:18
are doing a ton
36:19
with taking stuff
36:21
from meta, putting
36:22
it on shorts.
36:24
What do you see
36:25
that works, that
36:26
drives incremental
36:27
results for folks
36:29
and how do you
36:30
think they should
36:31
approach YouTube?
36:32
How much time
36:32
do we have?
36:34
YouTube is the
36:35
most complex
36:36
place to
36:38
advertise and
36:39
the hardest place
36:39
to advertise in
36:40
digital.
36:41
And YouTube
36:42
is like six
36:45
or seven
36:46
different things,
36:49
right?
36:49
So like when we
36:50
talk about YouTube,
36:51
I mean, we're
36:51
talking about so
36:53
many different
36:53
things.
36:54
And so, you
36:55
know, I'll start
36:56
with kind of the
36:58
the low, lowest
36:59
of low-hanging
37:00
fruit, which is
37:01
that YouTube
37:01
shorts thing.
37:02
It's the demand
37:03
gen conversion
37:04
focused YouTube
37:05
shorts component
37:06
where you can
37:07
take your
37:08
high-performing
37:09
9x16 video
37:10
assets from
37:12
meta, put them
37:14
in a demand
37:15
gen campaign,
37:16
and run it
37:18
purely on
37:18
YouTube shorts
37:19
placement.
37:21
You know, get
37:22
kind of click
37:23
based, pretty
37:24
good click
37:24
based attribution
37:25
out of it.
37:26
you do have
37:28
to do
37:28
audience
37:28
targeting.
37:29
That's one
37:29
of the things
37:30
to remember.
37:31
You know, I
37:31
always say like
37:32
YouTube today
37:34
is kind of
37:35
like meta
37:36
in 2015
37:37
where, you
37:38
know, you
37:38
do the 1%
37:39
lookalike
37:40
audience, right?
37:41
Like we
37:42
were on
37:44
this call
37:44
all old
37:45
enough to
37:45
remember the
37:46
old lookalike
37:47
audience stacks
37:48
and things
37:49
like that.
37:50
So you do
37:51
have to do
37:51
audience
37:52
targeting there.
37:53
But that's
37:54
kind of like
37:54
the lowest
37:56
form of
37:59
YouTube
37:59
advertising.
38:00
Do you have
38:00
any issues
38:02
with the
38:02
measurement on
38:03
shorts?
38:03
Like, is it
38:04
cool?
38:05
Like, do you
38:05
trust what
38:06
YouTube is
38:06
reporting on
38:07
shorts enough?
38:08
Or, like, I'm
38:09
thinking of the
38:09
brand that's
38:10
not paying
38:11
house.
38:12
I'm sure
38:12
house is
38:12
fantastic.
38:13
I know they're
38:14
fantastic.
38:14
But like,
38:15
just a quick
38:16
sidebar on
38:17
measurement of
38:18
shorts.
38:19
It sounds like
38:19
it's low-hanging
38:20
fruit, but
38:20
what about the
38:21
measurement side?
38:23
So, shorts
38:24
measurement is
38:26
okay, but
38:27
not amazing.
38:29
And Google
38:32
is, it's
38:35
always going to
38:35
be under
38:36
counting the
38:38
efficacy of
38:38
shorts.
38:39
And I will
38:39
say this very
38:41
declaratively.
38:44
If you don't
38:46
have the
38:46
ability to
38:48
do holdout
38:48
testing, do
38:49
not spend a
38:50
dollar on
38:50
YouTube outside
38:51
of shorts.
38:52
Shorts is the
38:52
only place you
38:53
can spend
38:53
money.
38:54
Do not
38:55
do in
38:55
feed, do
38:56
not do
38:56
in
38:56
stream, do
38:57
not do
38:57
CTV, do
38:58
not do
38:58
any of
38:59
it.
38:59
Like, if
39:00
you are
39:00
not doing
39:01
holdout
39:01
testing,
39:03
don't spend
39:04
money on
39:05
sort of
39:05
traditional
39:06
YouTube.
39:07
Just only
39:07
spend money
39:08
on shorts
39:08
because that's
39:09
the place
39:09
where you're
39:09
going to
39:10
get some
39:11
click-based
39:12
attribution
39:13
and the
39:15
rest of the
39:15
YouTube
39:16
inventory.
39:17
And this
39:17
is why
39:17
YouTube
39:17
inventory is
39:18
so mispriced,
39:19
right?
39:19
It's like,
39:20
it is hard
39:20
to measure.
39:21
You do
39:22
really need
39:22
to be doing
39:22
those holdout
39:23
tests.
39:25
So, stay
39:26
in shorts
39:27
and don't
39:28
venture out
39:29
of the
39:29
shallow
39:30
waters into
39:31
the deep
39:31
end of
39:32
YouTube.
39:33
I mean, I
39:33
think it's a
39:34
matter, I
39:34
think that's
39:35
good about the
39:35
measurement and
39:36
making sure that
39:37
you have a plan
39:37
for measurement
39:38
and makes a ton
39:39
of sense.
39:40
To me, it's one
39:40
of these things
39:41
that you hear
39:41
people talking
39:42
about.
39:42
They don't
39:43
really know how
39:43
to set it up,
39:44
right?
39:44
I think a lot
39:45
of people spend
39:45
a lot of time
39:46
optimizing their
39:47
meta ads
39:47
accounts.
39:48
And I
39:48
think YouTube
39:49
is a place
39:50
that takes a
39:50
lot of energy
39:51
and attention
39:51
and people
39:52
just want to
39:53
repurpose stuff,
39:54
which you can
39:55
do to a degree,
39:56
but I guess I'm
39:57
just curious of
39:57
any other tips
39:58
and tricks
39:59
and tactics
40:00
that you
40:00
believe are
40:02
really valuable.
40:02
The other big
40:04
thing is audience,
40:05
right?
40:06
Audience still
40:06
matters on
40:07
YouTube.
40:09
And Google's
40:10
targeting on
40:11
YouTube is the
40:13
best and most
40:14
granular targeting
40:15
that exists today
40:16
in digital.
40:18
Like, you really
40:19
can target your
40:20
ICP.
40:21
And Google's
40:22
bidding algorithm
40:23
is not at the
40:24
level where you
40:27
can just go
40:28
broad and it
40:29
will find your
40:29
ICP, right?
40:30
So, you know,
40:32
if you sell
40:33
beauty products,
40:36
like, an in-market
40:38
audience for
40:39
beauty products is
40:40
a good audience
40:41
to test on
40:43
YouTube, right?
40:44
Like, go
40:45
narrow.
40:46
Go top 10%
40:48
household income
40:49
women between
40:50
the ages of
40:51
35 and 54
40:53
who have
40:54
X, Y, and Z
40:55
interests, right?
40:56
Like, the kind
40:57
of stuff that we
40:58
used to do on
40:59
Meta, like,
41:00
still totally
41:01
applicable on
41:02
Google.
41:03
You should
41:03
definitely be
41:04
doing that sort
41:04
of targeting
41:05
with YouTube,
41:07
even on
41:08
shorts, right?
41:11
And, like,
41:13
shorts is
41:13
huge.
41:14
Like, it's
41:15
bigger than
41:16
reels.
41:17
It's bigger
41:17
than TikTok.
41:19
Two billion
41:20
people use
41:21
shorts a month.
41:23
Like, shorts
41:24
is colossal.
41:26
So the
41:26
inventory is
41:27
there for
41:28
essentially any
41:29
advertiser to
41:30
spend thousands
41:31
of dollars a
41:32
day on just
41:33
shorts as a
41:34
placement.
41:35
But, again,
41:36
you gotta
41:37
take what's
41:38
old is new
41:39
again.
41:40
You gotta
41:40
take the
41:40
old rules of
41:41
what you did
41:42
on Meta,
41:43
you know,
41:43
back in the
41:44
day and get
41:45
granular with
41:46
your audiences.
41:47
And take
41:47
your winners,
41:48
right?
41:48
Like, you
41:49
know, I
41:49
think short
41:49
form video
41:50
that performs
41:52
well is
41:54
short form
41:54
video that
41:55
performs well.
41:56
You know,
41:56
obviously, like,
41:57
TikTok's kind of
41:58
like the gold
41:58
standard.
41:59
Like, if it
42:00
performs on
42:00
TikTok, it'll
42:01
perform everywhere
42:02
else.
42:02
Like, it
42:03
performs on
42:04
Meta, it
42:04
might not
42:04
necessarily perform
42:05
on TikTok.
42:06
But take
42:07
your winners,
42:09
drop them
42:09
into a demand
42:10
gen shorts
42:11
campaign, have
42:13
some granular
42:14
audience targeting
42:14
with different ad
42:16
groups and doing
42:16
granular audience
42:17
targeting.
42:19
And, like,
42:20
that's a super,
42:21
super easy
42:22
place to start.
42:24
It will have,
42:25
like, pretty good
42:25
click attribution
42:26
to it.
42:28
And, you know,
42:29
you're not going
42:29
to spend $20,000
42:30
a day on it,
42:31
but $2,000, $3,000,
42:33
$4,000, $5,000 a
42:34
day, I mean,
42:36
it's not nothing.
42:36
I'm very curious
42:37
about the role of
42:39
landing pages
42:39
across Google.
42:41
I know from a
42:42
shopping side,
42:43
it would have to
42:44
be handled
42:44
differently because
42:45
it needs to go
42:45
to functionally
42:46
a PDP.
42:48
So, I guess you
42:48
could split it
42:49
into two parts.
42:50
Like, what are
42:50
the role of
42:51
landing pages from
42:51
a search perspective?
42:52
I'm assuming you
42:53
have a lot more
42:53
flexibility in what
42:54
you can and can't
42:55
do from a search
42:55
side.
42:56
But, to your
42:57
point earlier,
42:57
maybe there's a
42:58
little bit less
42:58
volume overall in
43:00
search.
43:01
But then I'm just
43:01
curious, like,
43:02
have you seen,
43:03
like, CRO-specific
43:04
tests to PDPs that
43:06
have been impactful
43:06
for Google traffic
43:07
specifically?
43:08
Like, obviously,
43:09
price and offer is,
43:10
you know, maybe not
43:11
where I'm going
43:11
with this.
43:13
But I'm just curious
43:14
if you've come
43:14
across any, like,
43:15
CRO wins that
43:16
have unlocked
43:16
crazy scale,
43:17
whether it is a
43:18
shopping or search
43:19
side.
43:20
Absolutely.
43:21
Yeah.
43:21
Yeah.
43:21
And landing page
43:23
is huge.
43:24
You know,
43:25
landing pages are
43:27
being important
43:28
on Google,
43:30
like, predates
43:30
them being
43:31
important on
43:32
Meta.
43:32
I mean, it's
43:33
historically been
43:34
very important
43:35
because landing page
43:36
quality score
43:36
is one of the
43:39
components of
43:40
quality score.
43:41
And quality score
43:42
impacts your
43:43
CPCs.
43:43
And so a
43:44
better, higher
43:45
converting landing
43:46
page will
43:48
increase your
43:48
quality score,
43:49
which means it
43:50
helps your ad
43:50
rank, which
43:51
means your
43:51
CPCs are
43:52
going to be
43:52
lower.
43:53
So it's
43:54
always been a
43:56
key component
43:56
to it.
43:57
And you can
43:57
run shopping
44:00
to landing
44:02
pages.
44:03
And especially
44:04
if you are
44:04
in, you know,
44:06
maybe something
44:06
that does not
44:06
have a lot of
44:07
SKUs, if you're
44:09
in kind of like
44:09
a, you're a
44:10
supplement with
44:11
like a massive
44:12
hero product,
44:13
you probably don't
44:14
want to be sending
44:15
that to a PDP.
44:16
You do want to be
44:16
sending that to a
44:17
landing page that
44:18
is specifically
44:19
structured for
44:20
that traffic.
44:21
Because again,
44:22
we get back to
44:23
the how can we
44:24
maximize the
44:25
inventory that we
44:26
can target
44:26
profitably, right?
44:28
Like that's going
44:28
to impact your
44:28
revenue per session,
44:30
right?
44:30
And that means
44:31
you can adjust
44:32
your targets and
44:34
you're going to be
44:34
able to get a
44:35
greater click share,
44:36
right?
44:36
Like I would
44:38
love to have a
44:38
90% click share
44:40
for search terms
44:42
that are good
44:43
for me, right?
44:45
Because that's
44:46
what's going to
44:46
increase my volume.
44:47
So landing pages
44:48
are massively
44:49
critical to
44:50
search, but
44:51
also to shopping.
44:52
And I, you
44:53
know, I'm not
44:53
saying create a
44:54
custom landing page
44:55
for every single
44:55
product that you
44:56
have, but if it
44:58
gets a lot of
44:58
volume and you
45:00
look and you
45:00
see, okay, my
45:01
click share is
45:02
relatively low.
45:03
Like, okay, well,
45:04
how am I
45:04
increasing that?
45:05
How am I
45:05
increasing the
45:06
click share for
45:06
this specific
45:07
product?
45:08
Like, let's
45:08
make a landing
45:09
page for it
45:10
because now our
45:11
revenue per session
45:12
goes up, which
45:13
means we can, you
45:14
know, adjust our
45:16
bids, which
45:17
means the
45:18
percentage of
45:19
the inventory
45:20
that we can get
45:20
and get profitably
45:21
is going to go
45:22
up.
45:23
So, yeah,
45:24
landing page is
45:25
massively critical.
45:26
It's the key to
45:28
unlocking additional
45:29
non-brand search
45:30
as well because
45:31
people are looking
45:32
for specific
45:33
things and
45:35
your products
45:36
might have
45:38
20 different
45:40
angles for what
45:41
it does, right?
45:42
Like a
45:42
magnesium supplement,
45:43
like it's great
45:44
for sleep, but
45:45
it's also like
45:46
good for focus
45:47
and all these
45:49
other things.
45:49
And if you're
45:50
sending all of
45:51
those queries to
45:52
your PDP,
45:54
that's terrible,
45:55
right?
45:56
It's like best
45:57
supplement for
45:57
sleep.
45:58
All right, well,
45:59
there you go.
45:59
Now we're hitting
46:00
our sleep queries,
46:01
right?
46:01
Supplement for
46:02
focus.
46:04
Now we're hitting
46:05
something completely
46:05
different.
46:06
And so, you
46:07
know, and this
46:08
is, I think,
46:09
the biggest
46:11
challenge with
46:12
Google is that
46:13
outside of
46:14
YouTube, there's
46:15
not one big
46:18
unlock.
46:19
You're not going
46:19
to find a killer
46:20
creative and
46:22
scale your
46:23
budgets 10x.
46:24
Like that's not
46:25
how it works.
46:26
It is brick
46:27
by brick
46:28
incremental
46:29
layering of a
46:32
few percentage
46:33
points of
46:33
efficiency here.
46:35
Unlocking a new
46:36
class of keywords
46:37
that's going to
46:38
add another,
46:39
you know,
46:40
$3,000, $5,000,
46:41
$10,000 of ad
46:42
spend a month.
46:43
That's running
46:44
efficiently, right?
46:45
You're layering
46:47
on these
46:49
bricks and
46:50
building it
46:51
one piece at a
46:52
time.
46:53
There's no,
46:55
hey, we found
46:55
an absolute
46:56
banger, like,
46:57
let's rip it to
46:57
the moon.
46:58
Like, that's just
46:58
not how it
46:59
works.
47:01
And, you know,
47:02
I think that
47:04
lends itself to a
47:05
specific type of
47:06
person who is
47:08
very in the weeds,
47:09
who is going to
47:10
do all of this
47:10
work.
47:11
And also why
47:12
I think
47:13
most people
47:16
who deal
47:17
with Google
47:18
or most
47:19
brands are
47:20
like, oh,
47:20
Google doesn't
47:20
work or,
47:21
you know,
47:23
it's because
47:23
they're not
47:24
willing to pay
47:25
people who can
47:26
actually do this
47:27
to build it
47:28
and build it
47:29
over time
47:30
because it
47:31
does layer
47:32
over time.
47:33
It is not,
47:34
we took it
47:35
from $10,000 a
47:35
month to
47:36
$400,000 a
47:38
month in,
47:38
you know,
47:39
two months
47:40
with one
47:40
banger ad.
47:41
It's layered
47:42
over time.
47:43
I think we have
47:44
a new podcast
47:45
called Skill Issue
47:46
that can come out
47:47
of this to bring
47:47
this full circle
47:48
where we just
47:49
like, we'll just
47:50
pull up ad
47:50
accounts and
47:51
Colin, you can
47:52
just publicly
47:53
shit on
47:54
everybody's
47:54
Google account.
47:56
It's so hard.
47:57
There are not
47:57
that many Google
47:58
people in the
47:59
world who are
47:59
good at it.
48:02
It is,
48:04
and I've seen
48:04
this with
48:05
clients who
48:08
have been
48:09
wanting to
48:10
fire us
48:11
for a very
48:11
long time,
48:12
not because
48:12
we're not
48:12
doing a great
48:13
job, but
48:13
just because
48:14
we're expensive
48:14
and they're
48:14
like, you're
48:15
too expensive.
48:16
We don't
48:16
want to pay
48:16
you.
48:17
We're going
48:17
to in-house
48:19
this.
48:19
I'm like,
48:20
okay, good
48:20
luck.
48:22
There are
48:23
just not
48:23
that many
48:24
people who
48:24
are legitimately
48:25
good at
48:26
Google because
48:27
the number
48:27
of years it
48:28
takes to be
48:29
really good at
48:29
Google is a
48:31
lot and the
48:32
attention to
48:33
detail is
48:34
massive and
48:36
the number
48:36
of
48:38
individual
48:39
settings that
48:40
if done
48:41
incorrectly
48:42
will completely
48:44
invalidate
48:45
everything else
48:45
that you
48:46
did is
48:47
incredibly
48:48
high.
48:49
You can do
48:50
all of the
48:50
things that
48:50
I've said
48:51
so far
48:52
through here
48:52
completely
48:53
correctly,
48:54
but if you
48:54
have your
48:55
conversion set
48:56
up incorrectly,
48:56
which is very
48:57
common,
48:58
let's say you
48:59
have two
48:59
different primary
49:00
conversions,
49:01
it's going to
49:01
be completely
49:02
non-incremental.
49:02
It's going to
49:03
generate zero
49:03
revenue and it's
49:04
going to look
49:04
great, but
49:06
it's so easy
49:07
to have
49:08
page view
49:09
plus purchase
49:10
as primary
49:12
conversions and
49:13
they're the
49:13
account defaults
49:15
and so they're
49:16
there in that
49:16
campaign and
49:17
you're telling
49:18
Google, I
49:19
value page
49:19
views and I
49:20
value purchases
49:21
and so obviously
49:22
one is much
49:22
easier to get
49:23
than the other
49:24
and so Google
49:25
is going to
49:25
just be like
49:26
generating traffic
49:27
that's firing
49:28
page views and
49:29
it's going to be
49:30
racking up like
49:31
efficiency in there,
49:32
but you're not
49:32
going to be seeing
49:33
it on your
49:33
bottom line and
49:35
it's because that
49:36
one single
49:37
setting was done
49:38
incorrectly and
49:39
now it's all
49:40
worthless, right?
49:41
And there's
49:42
dozens of
49:43
those throughout
49:45
the account and
49:47
you need someone
49:48
who can be
49:49
completely on top
49:50
of that, knows
49:51
where they are
49:52
and isn't going
49:54
to make a mistake,
49:55
right?
49:56
And there's just
49:56
it's hard work
49:58
and there's not
49:59
a lot of people
50:00
who have that
50:02
depth of
50:03
knowledge to
50:04
do it in a
50:05
platform and
50:06
then you add on
50:07
YouTube, which is
50:07
a totally different
50:08
platform within the
50:09
platform and it's
50:10
like, okay.
50:11
It's definitely a
50:12
lot.
50:12
Well, you've given
50:13
us a lot to think
50:14
about, Colin.
50:15
Lots of absolute
50:17
sauce for folks.
50:18
I appreciate your
50:19
time.
50:20
Thank you for
50:20
coming on and
50:21
appreciate you
50:22
being here.
50:23
Yeah, this was a
50:24
lot of fun and
50:25
happy to come on
50:26
and jam anytime.
50:28
Thank you everybody
50:28
for listening.
50:34
This episode is
50:35
brought to you by
50:35
Brad's company,
50:37
Work Marketing.
50:38
If you need a
50:39
D2C marketing
50:40
agency, let me
50:41
tell you,
50:41
Homestead is
50:42
great, but
50:43
Work Marketing
50:44
is also fantastic.
50:46
And let me tell
50:46
you, you
50:47
aren't going to
50:48
find friendlier
50:49
people out there
50:50
in the e-commerce
50:52
space.
50:53
So we decided to
50:54
do these little
50:55
ads for each
50:55
other's companies.
50:56
So hopefully you
50:57
find it interesting.
50:57
But seriously,
50:58
great team at
50:58
Work Marketing,
51:00
very smart.
51:00
Brad and Jordan
51:01
are incredibly
51:02
dialed in.
51:03
I just gave them
51:04
a lead, already
51:05
made this brand
51:06
that I gave them
51:07
like, I don't
51:07
even know,
51:08
double the revenue
51:10
that they had the
51:10
previous month or
51:11
something.
51:12
So, you know,
51:13
it's very exciting
51:13
to be connected
51:15
with Brad.
51:15
And if you need
51:16
a great agency,
51:17
there's really no
51:17
one better.
51:18
Zach, anything to
51:19
comment on Work
51:20
Marketing?
51:21
Yeah, I mean,
51:21
if you want an
51:22
agency that cares
51:23
about your business
51:24
much more than
51:25
they care about
51:25
their own website,
51:26
I just tried to
51:27
load WorkMarketing.com
51:28
and it was broken.
51:29
So they're definitely
51:30
going to give more
51:31
of a shit about
51:31
your business
51:32
than their own.
51:33
So I highly
51:33
recommend Brad
51:34
and the team
51:35
over at Work.
51:35
They've been
51:35
incredible.
51:36
We've referred a lot
51:37
of business over
51:38
to them as well.
51:38
Really, really good
51:39
as far as like
51:40
cracking funnels
51:41
and figuring out
51:41
like rapid growth
51:42
for brands.
51:43
So I highly
51:44
recommend these guys.
51:45
The only way that we
51:55
grow this podcast is
51:56
by you sharing it
51:58
with your friends.
51:59
Honestly, like
51:59
reviews kind of
52:00
don't really mean
52:00
anything too much
52:01
anymore.
52:02
They're really
52:02
meaningful, but they
52:03
don't do a lot for
52:04
the growth of the
52:04
podcast.
52:05
And so sharing
52:06
YouTube links,
52:07
sharing Spotify
52:08
links, sharing
52:09
Apple, whatever we
52:11
call it under the
52:12
podcast app now,
52:13
anything you can
52:14
share, the better
52:15
we're going to be.
52:16
Guys, anything else
52:17
you want to say on
52:18
this?
52:18
Yeah, please go check
52:19
us out on YouTube.
52:20
Rack up those views
52:21
for us.
52:22
We'd love to see it.
52:22
And then subscribe.
52:23
Make sure to subscribe
52:24
on YouTube as well.
52:25
And I relentlessly
52:26
refresh the YouTube
52:26
comments because it
52:27
dictates my mental
52:28
health for the day.
52:29
So please say
52:29
something nice about
52:30
all of us.
52:32
Thank you, everyone.
52:33
Thanks for listening.
52:34
Honestly.