Join ClubPF
Nick Bennett 56 min

The Death of Marketing Attribution


Join Rand Fishkin in a revealing discussion on the evolving challenges of marketing attribution, exploring why traditional models no longer hold and what the future holds for marketers.



0:00

So, Nick, I'm really excited about this.

0:05

This is just on a personal note very quickly.

0:08

I started my career as a content marketing strategist,

0:13

was my title, is basically an account manager

0:15

for a portfolio of about 30 brands

0:19

at a content marketing agency in Boston called Brafton.

0:24

And every Friday we would get together

0:26

and watch Rand's Whiteboard Friday videos

0:29

because none of us, we were all more or less fresh

0:31

out of college, none of us knew what we were doing

0:33

and we were supposed to be advising directors

0:35

of marketing on a SEO and content strategy.

0:38

So, Rand, I feel like I owe you my career.

0:41

I'm really excited to have you here

0:44

and you've done a lot since then too.

0:46

Those were, that was a lot, a long time ago.

0:48

So, you are now co-founder and CEO of Spark Toro,

0:52

which is a fantastic product.

0:55

I hope we talk more about it.

0:57

If we don't, I'm gonna give it a plug later

0:58

as I use it all the time.

1:00

And also, co-founder and CEO of Snapbar Studio.

1:03

So, and I mentioned Whiteboard Fridays,

1:07

but that's, folks don't know

1:09

when you were founder and CEO at Moss.

1:12

- Right, yeah.

1:14

- So welcome.

1:14

I'm so excited to be here with you and talking to you

1:17

and learning from you.

1:18

I attended your conference Spark together

1:21

at the last year. - Oh yeah, wonderful.

1:24

- And heard you talk about this topic

1:28

a little bit there and immediately was like,

1:30

I wanna hear more about this.

1:32

And so I'm really excited to bring it to you,

1:33

Club PF today.

1:35

Thanks for being here with us.

1:37

- Yeah, gosh, my pleasure, Colin.

1:39

Thank you for having me.

1:41

Yeah, it's funny.

1:45

I think a lot of folks know me from Whiteboard Friday

1:47

oddly enough, those videos have a long, long half life.

1:51

I still find folks discovering them to this day,

1:54

which is kind of exciting.

1:55

Also makes me feel guilty

1:56

for not producing more video content.

1:58

Maybe as Sparktorro matures,

2:03

I'll need to get back in front of the Whiteboard.

2:06

So-- - Because my therapist would say,

2:07

you gotta turn that guilty for into a thankful for somehow.

2:10

- Hmm, yeah, yeah, I like that.

2:13

That's a turn about is fair play.

2:16

- Well.

2:17

- So yeah, we do, we, you know,

2:18

we've got an exciting presentation today.

2:20

I think Colin, I know that one of the questions

2:24

I always get whenever I present is,

2:26

will the slides be available afterward?

2:28

Are you gonna make those available to folks?

2:30

- Be sure, and the recording.

2:32

- Okay, excellent, terrific.

2:36

So, where am I?

2:38

This up and we're gonna share screen.

2:43

Okay.

2:48

Is that look good to everybody?

2:52

- Looks great, yeah, thanks.

2:55

All right, terrific.

2:57

So, my contention here for a lot of marketers

3:01

is that the world, the universe of attribution,

3:06

which is different than measurement,

3:08

we're gonna talk about those is over.

3:10

And there are a variety of reasons

3:13

that this has come to an end.

3:15

So I'm gonna set some context

3:17

and then I'm gonna explain where I think

3:20

the smartest marketers are putting their energy and effort

3:24

in order to overcome this big challenge.

3:27

Oh, one of the problems for sure is that,

3:31

and I am as guilty of this as anyone,

3:34

especially as somebody who did lots of education

3:36

in the marketing field for the last couple of decades,

3:38

is that we spent all this time convincing executives

3:42

that they should invest in digital channels

3:44

because precisely because they were not only measurable,

3:48

but attributable.

3:50

And our big weapon in that fight

3:53

was that, you know, unlike putting dollars against radio

3:56

or television or billboard advertising

3:59

or local campaigns or mailers,

4:02

is that we could precisely tell you

4:05

what how many of your dollars were going to any channel

4:08

and how much return on investment you were getting

4:10

from each of those channels

4:12

by building these sophisticated, moderately sophisticated

4:16

attribution models.

4:19

And that came from basically this idea

4:22

that everything in digital was measurable,

4:25

that, you know, when you looked at your marketing dashboard,

4:30

you could see exactly,

4:32

not just how much you spent on an ad channel,

4:34

but you could see the number of people

4:37

who came from that channel,

4:39

what they visited on your site

4:41

before converting to a sale

4:43

or signing up for an email list.

4:45

And that buyer journey was completely trackable.

4:52

And then then it all went sideways.

4:54

And then that era ended.

4:57

And I'm gonna go through what sort of the core causes of this,

5:02

but really it's the combination of all of these things

5:07

that have ended it together.

5:08

And they've come to a head in just the last couple years.

5:12

And especially this year,

5:13

we'll talk about the death of third-party cookies.

5:15

But so third-party cookies ending in 2024.

5:20

This is coming quite soon.

5:21

You probably read about the cookie apocalypse for a while now.

5:24

So I won't belabor the point,

5:26

but Chrome, which has the dominant market share worldwide

5:29

and certainly in the West,

5:31

is the primary one that's gonna be phasing them out

5:35

after I think it's September of this year,

5:37

somewhere around there.

5:39

And once that happens, the tech is dead all over.

5:43

First-party cookies, which are still alive,

5:46

still around, you can still drop a cookie.

5:48

You can, you know, remember someone's log in,

5:50

you can see what activity they had on your website

5:54

or in your app on previous visits and tie those together.

5:58

Yes, you can, but it is not anywhere close

6:02

to what it was five or six years ago.

6:04

The ability to drop, track and measure cookies,

6:09

cross-browser, cross-device and cross-session

6:13

are all declining.

6:14

And this is because many more consumers

6:17

are embracing multiple devices,

6:19

many more consumers are embracing tech that blocks

6:23

cookie tracking, including ad blockers.

6:27

And when this happens, right, when you lose 20, 30%

6:32

of your ability to, you know, build a sophisticated,

6:37

sort of, or accurate model of visitor traffic representation,

6:43

most of these attribution models collapse.

6:45

They can't survive the loss of about more than about 30

6:48

or 40% of accurate, you know, tracking systems.

6:53

And that makes sense, right?

6:54

'Cause the model is built on, okay, we got, you know,

6:57

600 visits and, or 600 sales, 100 of those sales

7:01

came from this particular channel in this particular week

7:04

and they had these previous activities.

7:07

But if that becomes 20 or 10, it just doesn't,

7:12

it just doesn't work out.

7:14

So, ad analytics blockers, those multi-device journeys,

7:18

right, 20%, 20% of all browsers in the world

7:22

blocking Google Analytics tracking way, way higher

7:26

in tech savvy sectors.

7:27

Most of you listening to this probably have some form

7:31

of blocking system.

7:33

And ad blockers aren't just blocking ads,

7:36

they also block tracking techniques.

7:38

Like we've talked about, I am essentially saying,

7:41

I think you can forget about tracking an individual

7:44

pre to post login, which means

7:47

the sophistication level that you could have in the past

7:50

around saying, oh, this content was consumed by people

7:54

who eventually became these customers is mostly gone.

7:58

You can only really do that realistically post login.

8:02

Meaning you have to capture someone's, you know,

8:05

whatever email and password, you have to get them

8:07

to log into your system before you can do it,

8:08

which means you really don't know what top of funnel looks

8:11

like, you barely know what middle of funnel looks like.

8:13

You only could realistically, the only part

8:16

of the funnel that's attributable is the bottom of funnel.

8:21

Turns out digital ads quite good in that sector

8:25

and all the awareness, SEO, content, PR, email,

8:29

all that stuff often sits above that.

8:31

If that weren't enough, we have a bunch

8:35

of new privacy laws.

8:37

Many of us are probably in the United States right now

8:41

and what's possible and legal here,

8:43

which we often think of as like where the web is at,

8:46

is not possible or legal in nearly any other jurisdiction.

8:50

You go to Germany, you go to France, you go to Italy,

8:53

you go to Vietnam or Singapore or Indonesia

8:55

or Australia or New Zealand or South Korea or China

8:58

and the types of tracking that's still possible

9:01

in the United States, not possible in these places.

9:03

So this cuts off a huge additional number

9:06

of your potential trackable measurable audience.

9:10

And technically, Google Analytics is sort of still legal

9:14

in the EU, but realistically, it's so complex to set it up

9:18

that almost no one is even able to use that technology,

9:22

which means the only thing they're really measuring

9:23

is what we got from old log files, right?

9:25

Like we had this many visits to this page,

9:28

this many requests from this many browsers,

9:30

were they real browsers, we don't know.

9:32

Let's come to the culprits here,

9:39

the causes of many of this pain.

9:41

I bet a lot of marketers, while those of you

9:43

who watch "Whiteboard Friday" probably didn't forget.

9:45

But a lot of marketers have probably forgotten

9:47

how easy it was to attribute organic content back in the day.

9:52

A ton of this was through SEO.

9:53

You used to be able to see every single visitor

9:57

what they searched for in Google

9:59

before they clicked on your link and came to your website.

10:03

And with that data, you were able to reverse engineer,

10:05

hey, these keywords are really valuable for me,

10:07

these are not, this is how many people,

10:09

this is where the demand is coming from,

10:11

this is what from my site is ranking,

10:13

for which other keywords.

10:14

It was a beautiful age, right?

10:17

We had this data from 1997 to 2015, 1415, gone, gone.

10:22

Right?

10:28

That ability was incredible.

10:32

In 2023, Google sort of learned, especially post-pandemic,

10:37

that their monopoly power was so untouchable

10:40

that they could shrink the share of searches

10:45

that result in a click and still make enough money,

10:48

still make more money through advertising dollars.

10:51

They essentially answer so many user queries

10:54

right in the search results now.

10:56

And of course, search generative experience,

10:58

which supposedly will roll out at some point,

11:01

is gonna do this even more.

11:02

So the reality is, yes, advertising clicks

11:09

are still technically on the rise slightly, right?

11:12

The compound annual growth rate is not super high anymore.

11:15

Google's essentially plateauing as a business,

11:17

which is why I think they're doing more extraction

11:21

of value on their existing tech

11:24

versus trying to innovate and invent new things.

11:26

This is a big part of why all these big tech companies

11:28

are doing layoffs.

11:30

Percent of clicks on organic is declining,

11:33

keyword data is hidden, good luck tracking any visitors,

11:36

unless they're logged in to your site.

11:38

And the percent of searchers who are satisfied

11:40

by whatever Google surfaces, it keeps rising.

11:45

So there've been a lot of arguments that like,

11:47

oh, Google's getting worse, but user satisfaction

11:50

as measured by Google sort of saying,

11:52

oh, this person searched for this keyword,

11:54

then they didn't search anymore

11:56

because they were satisfied by an answer.

11:58

That's as high as it's been.

12:00

Google also, on the advertising side,

12:03

is doing this really sketchy stuff.

12:05

I'm sure you've read about or heard about performance max

12:07

or P max campaigns and how manipulative and problematic

12:12

those are.

12:12

That is only getting worse.

12:16

And the net net here is that,

12:20

this is not just Google, this is gonna be Apple, Amazon,

12:22

Facebook or Meta, all the advertising platforms

12:25

are essentially switching from a more granular,

12:29

you can see all the data, you sort of know,

12:31

which sites or which YouTube channels

12:33

or which Facebook pages or the demographics

12:36

of people who saw your ads to a world

12:38

that is much less granular, to a world

12:41

that cuts attribution models.

12:43

You can see this in Google Analytics, right?

12:45

They now only have the two that make Google ads

12:48

look the best.

12:49

And this willingness to literally make search results

12:54

worse in order to sell more ads,

12:56

which again, value extraction, this was surfaced

13:00

by the Department of Justice's,

13:02

the United States Department of Justice,

13:04

suing Google for abuse of monopoly power.

13:06

And this was disclosed, this is an email

13:09

from the head of ads to the rest of the team,

13:11

including the C suite at Google saying,

13:15

"Hey, people, do you wanna get your short-term bonuses?

13:19

If so, you need to roll back this improvement

13:22

to search results 'cause the ad clicks are dropping."

13:26

And what happened?

13:27

Of course they rolled it back, right?

13:28

Everybody wants to get their bonus.

13:30

What's so good.

13:32

All right, so beyond search though,

13:35

this is happening in social too.

13:37

I'm sure you've noticed this.

13:39

If you log into analytics, you can see

13:41

that 76% of all visitors or whatever,

13:43

some huge number come from direct or type in.

13:47

How's that possible?

13:48

When I look at the URLs, did somebody really type in

13:51

zero-click, hyphen content, hyphen the,

13:53

hyphen counterintuitive?

13:55

No way, impossible.

13:57

That's not being bookmarked.

13:59

What's really going on here?

14:01

Well, we ran an experiment and turns out

14:06

a ton of these social networks and content networks

14:11

hide all their referral data by default.

14:14

And some hide a percent, like LinkedIn was like hiding 15%,

14:19

14% of referrals.

14:21

There's not, I don't think this is wise by the way.

14:27

I think that TikTok would do better

14:29

if they told websites, hey, this traffic came from TikTok.

14:34

But instead, if you look in your referrals,

14:36

you're not gonna see any TikTok at all,

14:38

because TikTok intentionally hides

14:41

and obfuscates the referral.

14:44

This is same story with WhatsApp and Discord

14:47

and Mastodon and Slack, all those B2B communities in Slack

14:51

that are sending your SaaS and software customers traffic,

14:55

you're never gonna be able to see it or measure it.

14:57

You can't build attribution around it

14:59

because it's completely missing.

15:02

So there's a blog post on our website about dark traffic

15:07

as it comes from different social networks

15:09

and the percent with attribution,

15:11

the percent that are attributed to direct

15:13

that you can check out, but huge problem, right?

15:17

Massive issue.

15:18

And last but not least, the rise of zero-click everything.

15:23

So every one of these platforms,

15:26

this is not just social networks, right?

15:29

I'm talking about YouTube and Google

15:31

and places like that as well.

15:34

But all of these places that have links out

15:39

that the algorithm prioritizes native content,

15:44

meaning Google is gonna put Google's answer

15:47

to your search query above everything else,

15:50

the one that has the sort of rich snippet

15:53

that requires no click.

15:55

YouTube is gonna hide by default the description

16:00

underneath a video link.

16:02

So when you go to a YouTube page, you see a video,

16:05

often, very often in the description,

16:07

there's gonna be a link to a website, right?

16:10

A third party website, but it'll be hidden.

16:12

YouTube will make sure that you can only see

16:17

the text before that link.

16:19

Instagram, the only place you can do it

16:21

is the link in bio, we all know that post

16:23

that include link in bio, perform worse than those

16:26

that don't, Twitter, when Elon bought it.

16:29

I mean, this was happening before,

16:30

but he made sure that external links are devalued even more.

16:35

And when you open source the algorithm,

16:36

he like, you know, showed explicitly

16:38

that if there's a link in the post,

16:40

it's gonna intentionally perform worse.

16:42

This is same as true on Reddit,

16:44

same as true on LinkedIn, on TikTok.

16:46

The only place you can put it is link in bio

16:48

and they're testing no bios.

16:50

So basically, what we're seeing,

16:53

what I'm seeing is stuff like this.

16:55

This is a post of mine that has a link,

16:58

impressions, you know, 444 impressions, great.

17:01

This one has no link about 10x, right?

17:04

10x the reach, 10x the engagement,

17:08

because the platforms don't want you to link out.

17:11

They wanna keep all the traffic on their site, right?

17:14

Every platform is all about not letting people leave.

17:20

TikTok hiding the user's link in the bio as a test.

17:23

Just getting even worse.

17:25

Okay.

17:27

So I ran through a whole bunch of things

17:29

that are nightmarish for anyone

17:31

who wants to do measurement and marketing.

17:33

But this is my core contention, which is in 2013,

17:38

it was challenging, but possible,

17:41

to build a sophisticated attribution model

17:44

and make your digital marketing investments

17:47

intelligently based on that.

17:48

And in 2023, you can still measure a lot of this stuff, right?

17:52

You can measure like, oh, I have this many new followers

17:55

on TikTok, right?

17:56

I had this much engagement on this LinkedIn post.

17:58

Or we think that this many people searched

18:01

for this Google keyword and I rank number four for it.

18:04

Like you can measure, but you cannot build

18:08

a realistic attribution model.

18:11

Anyone who tells you that they've got a, you know,

18:14

great attribution model beyond people

18:17

who just are trying to attribute their big tech ad spend

18:21

and are spending, you know,

18:22

100 million plus dollars a year

18:24

so that they've got a sizable enough cohort to measure.

18:28

I don't believe them at all.

18:29

I don't think it's realistic.

18:31

I have never seen, I've given this presentation a few times,

18:35

people who come up to me afterwards have been like,

18:37

"Let me show you what we're doing."

18:38

And I don't buy it.

18:40

It doesn't pass muster.

18:42

So let me be clear before you are like,

18:46

"Oh my God, why are we even measuring anything?"

18:47

Like we can't, measurement and attribution

18:49

are two different things.

18:50

So attribution is, I see this ad while I'm scrolling Reddit.

18:55

I visit the website, I reach a call to action.

18:58

I complete that action and the analytics of,

19:03

what's this company, Clay, right?

19:05

Of Clay.com tells Clay.com Reddit was responsible

19:09

for this purchase, right?

19:11

Not just the visit, but the purchase.

19:12

Or they're like, "Hey, we've seen this visitor four times

19:15

in the past, so we're gonna give 25% of the credit

19:17

to the Reddit ad."

19:18

Or we've got a weighted model.

19:20

We're gonna give 40% of the credit to the last click.

19:22

Or the weighted model starts at the beginning of the journey.

19:25

So we're giving 10%, whatever the model is.

19:28

Measurement is entirely different.

19:30

Measurement is HBO runs an ad for Last of Us on,

19:35

oh, it's now called Max, isn't it?

19:37

Well, I had to get rid of the brand name everyone knows

19:40

for the thing that we all associate with.

19:43

Do you have this association to call on?

19:45

Whenever I think Max, I always think Skinamax

19:47

from like the '90s.

19:49

Yeah, and also it used to be called HBO Go.

19:53

So this is like the second or third time they've changed

19:55

and relaunched this app and it's exactly the same user

19:58

experience, it doesn't make any sense.

20:00

I mean, the only thing that they changed

20:01

is you don't automatically log in again.

20:03

You have to like reset your credentials

20:05

on your television every time, it's real annoying.

20:08

So HBO, I'm gonna call them that,

20:12

run this ad for awareness.

20:14

They get stats about how many people saw it.

20:16

This is how many views the ad got,

20:18

this is how many hovers it got, this is how many clicks,

20:20

this is how many people watch the trailer.

20:22

And then they can compare that to the ads

20:24

they've run for their other shows.

20:25

So Last of Us versus Distant Lands or Harley Quinn

20:28

or Flight Attendant or whatever shows they've got.

20:30

Historically, we called these vanity metrics

20:35

because they did not tie directly to dollars, right?

20:39

You could not say, oh, well, it's that many clicks,

20:44

that's 11% higher than what we saw with Harley Quinn

20:47

or whatever, but what is the money?

20:52

Like how many people bought because they saw this ad?

20:55

You can't prove that.

20:57

We don't, with the answer is we don't know

20:59

how many people bought because of this ad.

21:01

We only know the people who saw it.

21:03

Measurement is not attribution and it doesn't try to be.

21:07

I think the problem with attribution

21:08

is that it's faking science, right?

21:12

It's saying, hey, this thing is true

21:15

based on a theory of how we think human models

21:20

or human beings work and, well, gosh,

21:24

if I hadn't seen that Reddit ad,

21:25

I would have not converted, but nine out of 10 people

21:28

who saw it would have converted.

21:29

I think that's pseudoscience.

21:33

I think that's mythical thinking.

21:38

So in 1960, little story time,

21:41

but 1960, right?

21:42

Coca-Cola runs this billboard.

21:45

They spent millions, but look at that neon.

21:47

That thing's giant, conveniently sitting above

21:50

a four roses bourbon ad, right?

21:53

'Cause the idea is bourbon and Coke together,

21:55

but of course Coca-Cola doesn't wanna advertise them together

21:58

because then it's a liquor ad and blah, blah, blah.

22:01

But this billboard in Chicago, Michigan Avenue, right?

22:07

How did they decide that they were gonna run that?

22:09

Obviously they didn't have sophisticated

22:11

digital attribution, right?

22:13

But it must have been something else.

22:15

Well, back in the day, right?

22:18

In the sort of Don Draper Mad Men era,

22:21

they were looking at the audience,

22:22

like how many people drive through

22:23

and walk down Michigan Avenue in Chicago?

22:26

What's the potential reach, right?

22:28

That we can expect there?

22:29

What's the context of this?

22:30

Okay, does Michigan Avenue sort of have a,

22:33

you know, it's sort of like the,

22:34

like New York's Fifth Avenue or a lot of smaller towns

22:38

as main streets, right?

22:39

It's got this cache, cultural cache, market comps, right?

22:43

They looked at, well, when we ran this billboard

22:45

in Detroit, how to do it?

22:46

And we ran this billboard in Cincinnati, how to do it?

22:48

When we ran this billboard in Los Angeles, how to do it?

22:51

And then we do lift-based measurement, right?

22:54

So essentially, we look at stores in Chicago

22:59

from the same quarter or same month last year

23:02

after the billboard went up, before the billboard went up

23:05

and after the billboard went up

23:06

and then we decide, huh, it looks like we saw a 5% lift.

23:10

We didn't run any other campaigns in that local area.

23:14

And so we're attributing that 5% lift

23:18

to probably the billboard.

23:19

Do we know for sure it was a billboard?

23:21

I can know.

23:22

But that's what marketers used, right?

23:28

The whole advertising profession was built

23:31

on vanity metrics, the things we've been dismissing

23:34

for the last quarter century.

23:35

Here's a potential customer, right?

23:40

Consuming content about your product.

23:42

This is for the video game disco Lee Zim,

23:44

which I personally think is a masterpiece,

23:47

but it's an acquired taste for sure.

23:49

And, right, so they're watching on YouTube,

23:52

they're seeing a TikTok video about it.

23:54

They might go to Google and get so much information

23:57

from the page that there's no need to click, right?

23:59

No reason to visit the site at all.

24:01

Sorry, I'm getting allergies

24:06

'cause it's a little warm already.

24:08

And those people might read some third party reviews,

24:12

but you know what they're never doing?

24:15

They're never visiting your website, right?

24:18

That the folks who run disco Lee Zim

24:20

have only two kinds of attribution.

24:23

Number one is view through ads.

24:24

They know the ads that they bought

24:25

and how many people viewed them.

24:28

And they know potentially last click

24:32

if people came and sign up the email list

24:34

on their actual website,

24:36

or there are some metrics that you're allowed to get

24:38

from the Steam page, the page that sells the PC game.

24:42

But if you optimize for exclusively these two forms

24:48

of marketing, you're gonna destroy,

24:52

absolutely destroy an underinvest in all brand building.

24:56

All organic, all content marketing, all SEO,

25:00

because of those things, all social media,

25:02

because those things can't be built

25:04

around attribution models anymore, right?

25:07

They're just not present, that data does not exist for you.

25:10

Those vanity metrics can't be tied to attribution.

25:13

So look, I mean, you want an attribution model.

25:18

The only way that works is if the analytic software

25:21

that you're using and the model that you've built

25:23

can record all the devices that made all the visits

25:26

and tie them all together.

25:28

Otherwise, you are just BSing yourself

25:31

and your team and your company and your boss,

25:33

and you are absolutely under serving your market

25:37

with your advertising and marketing.

25:39

None of this is gonna happen in a cookie list,

25:42

you know, 2023 browser privacy law regulated,

25:46

ad blocker filled multi-device world, not happening.

25:51

There are lots of vendors who will try and convince you

25:56

that everything I've said to you is wrong,

25:58

because they would like to sell you very expensive software

26:00

based on machine learning,

26:01

or now they're gonna call it artificial intelligence,

26:03

'cause it's way hotter, that build samples

26:07

based on some, you know, MMM marketing mix model.

26:11

And you can read all about these.

26:13

The math behind them is pretty impressive.

26:16

I am not,

26:23

I'm not trying to pour shade, throw shade on

26:26

what these vendors are doing, right?

26:27

They, I think they have their market.

26:30

I think there's a reason that they can be invested

26:35

in reasonably, and that is if you're spending tens of millions

26:38

or hundreds of millions of dollars

26:39

on big tech advertising channels,

26:41

and what you really wanna know is

26:43

which big tech advertising channel is contributing

26:46

more or less to my, you know, eventual sales.

26:51

This is reasonable.

26:52

This is a reasonable thing to invest in.

26:55

But the MMM stuff only works if every single brand touch

26:58

can actually be included.

27:00

And if you think that's possible,

27:02

there's a bridge on Mars that they should sell you instead.

27:06

Here's a brief abbreviated list of things

27:08

that marketing mix modeling will never, cannot measure.

27:13

Watch your talk at a conference, this webinar.

27:15

No, impossible.

27:17

There's no way, there's no way a marketing mix model software

27:20

can say, oh, RAM gave a presentation, right?

27:22

The PF community, the presentation was well received.

27:25

Lots of people went over to SparkToro

27:27

and signed up for a free account.

27:28

And then this percent of them became customers.

27:32

Nope, zero attribution, absolutely impossible.

27:35

Private Facebook book group, no way.

27:37

Friend emailed you about it.

27:39

Nope, not measured.

27:40

Word of mouth mentioned in a YouTube video, Reddit comment.

27:42

Boss forwarded your newsletter,

27:44

learned about you from some branded merch.

27:46

We don't have branded merch, but right.

27:49

Just not fundamentally impossible.

27:51

And word of mouth, in my opinion,

27:53

is responsible for the overwhelming majority of sales.

27:57

So look, that's not even, I'm not even mentioning

28:02

all the things that yes, technically,

28:04

it could be that you could put something

28:06

into your attribution model,

28:08

but there's no way anyone is gonna put

28:10

most organic content, most one to one outreach,

28:12

most comment marketing, most SMS, most streamer reviews,

28:15

virtually anything unpaid to do, just not, just toss it out.

28:20

So, this is the part of the presentation where you go,

28:25

right, Rand, are you just telling me

28:28

that because ads are the only thing

28:30

I can realistically put attribution behind,

28:33

that I should just spend a ton more money on ads?

28:36

Yeah, of course, you know me.

28:37

I love digital marketing going entirely

28:40

to Google and Facebook's bottom line.

28:42

That's exactly what I'm gonna tell you to do.

28:43

Oh, all right.

28:46

In case you don't believe me.

28:48

Look at all these companies that put hundreds of thousands

28:51

and millions of dollars into these channels,

28:54

and then pulled back on them, right?

28:55

eBay, tens of millions of dollars, Uber,

28:58

a hundred million dollars Chase,

28:59

running ads across hundreds of thousands of sites

29:02

and then 5,000, Procter and Gamble seeing Lyft

29:04

when they pulled back on their big tech ad spend.

29:07

These are not folks who don't have the sophistication

29:10

or the marketing mix modeling software

29:12

or the teams of analysts and data scientists,

29:15

right, putting stuff into this.

29:17

What could possibly be going on?

29:19

How could it be possible that by pulling back

29:21

on big ad tech spend, they are seeing improved results

29:25

over and over again?

29:26

This is probably the most famous one a couple of years ago,

29:28

right, Brian Chesky talking about how they pulled

29:30

all their spend and saw the same traffic

29:32

and there's the New York Times from last year

29:34

reporting that yes, indeed, that is what happened.

29:38

You can see it in their numbers.

29:39

I'm gonna explain this through a little parable

29:43

and then I'll give you some tips and tactics

29:48

that I recommend and we can do some Q&A.

29:50

So, woman in, this is in Milan,

29:55

she owns the pizzeria, Sibia.

29:58

And she's like, gosh, you know,

30:00

I'd love to sell some more pizzas, right?

30:03

Everybody loves a good pizza.

30:06

But my business is not as robust as it could be.

30:09

So she hires three little neighborhood scams, right?

30:12

These teenagers to go run around and pass out flyers

30:16

with discount codes, right?

30:17

It's got like, you know, 25, 30% off a pizza.

30:20

She's got a green flyer, she gives to one kid,

30:22

a red flyer and a white flyer.

30:25

And a month later, she looks and she's like, wow,

30:27

you know, like 40% of my receipts are coming from

30:31

that kid with the green flyer.

30:32

Like the green flyer is just boosting business so much.

30:34

This is incredible.

30:36

You know what, kid, you're hired,

30:38

the other two scamps run along.

30:41

And then six months later, she talks to her accountant

30:43

and her accountant's like, well, you know,

30:45

business is up a little bit, yeah, about 8%, 9%.

30:48

But because of the discount code, you know,

30:50

so many of your existing customers using this

30:53

account code that, you know, your total receipts

30:56

are actually down.

30:56

The one's like, how could that be?

30:58

I don't, doesn't make sense, you know,

31:00

the restaurant feels busier.

31:03

Maybe it doesn't feel as busy as I thought, but gosh.

31:07

So what does she do?

31:09

She puts on a disguise, she follows the kid

31:13

out of the pizzeria and she sees what he does, right?

31:17

And what does the kid do?

31:18

The kid goes out of the pizzeria,

31:19

he's got his big stack of green flyers.

31:21

He ducks into an alley around the corner and then he waits.

31:25

And as she's watching, she sees this kid just waiting

31:28

and waiting and then as soon as somebody's walking,

31:30

like they're gonna walk into the pizzeria,

31:32

he jumps out and hands them the flyer with the discount code.

31:35

Right?

31:37

Because why work hard to make more sales

31:40

when you could just take credit for sales

31:42

that were already going to happen?

31:45

AKA Facebook and Google's motto.

31:49

You know, the hardest wall of pills here, my friends,

31:54

is that the best marketing opportunities

31:57

are the least measurable.

31:58

That's just how it is.

32:02

And if you can't convince your boss, your team,

32:05

your client, your execs, that this is true,

32:09

shown this presentation, right?

32:12

Because nobody disagrees with this in theory,

32:14

they only disagree with it in their own metrics.

32:17

Everyone knows that this is true, right?

32:20

It has to be true.

32:22

This is frustrating, but it is the bitter truth.

32:27

So what's the solution here, right?

32:30

We talked about problems, we talked about context.

32:33

What am I actually recommending?

32:35

Well, look, in 2024 and beyond 2024 and beyond,

32:40

when just there's four, okay, and beyond.

32:47

Oh, great, yeah.

32:51

Oh, I think my screen sharing ended, hang on a second.

32:57

There we go.

32:57

We're gonna reshare that screen.

33:03

This presentation, perfect.

33:07

Let's go back to where we worked.

33:09

Number one, you can throw money at the big tech ad providers

33:16

who are gonna be very happy, very happy to take your money

33:21

and provide you with attribution, right?

33:23

They will say, yes, oh, yeah, you spent $20 million with us

33:28

and you made $35 million, we're taking credit.

33:32

You know what they're great at?

33:33

They're great at being that kid with the green flyer

33:35

and jumping in front of your customers.

33:37

What does Google know better than anyone else?

33:39

Where people go on the web before they buy?

33:43

What does meta know better than anyone else?

33:45

Where people go on the web before they buy?

33:47

And what are they great at?

33:48

Showing your ad to people who are going to buy anyway.

33:53

You can try this with any company in the world.

33:55

I guarantee if you cut all advertising spend for 90 days,

34:00

you are gonna see that a tremendous number of your sales

34:05

that are attributed to advertising come anyway.

34:10

Is that number 50%, is it 75%, is it 95% like Airbnb?

34:15

You will never know unless you cut it.

34:17

The second thing you can do, your second option

34:21

is build a hard to measure dashboard, right?

34:23

That infers lift from channels for that madmen, Don Draper,

34:26

you know, Chicago, Michigan,

34:27

having you think that we talked about where you are testing

34:31

different campaigns and you're watching the lift

34:33

that you see from that and you're investing in them one by one,

34:36

geographically or time series based 20th century

34:40

style measurement, right, old school.

34:43

And number three, conceivably,

34:45

you can just measure the big stuff.

34:47

Traffic, conversions, revenue, ignoring almost everything else

34:52

and investing in creative, hard to measure,

34:55

but vanity metrics, lifting channels and tactics.

35:00

This works, this is, it's actually what we do.

35:05

So, you know, you run these big tick ads,

35:08

some of them are probably gonna work fine for big brands,

35:11

but if you're trying to grow a brand,

35:12

I don't believe in these.

35:13

I don't think small and medium businesses

35:15

challenge your brands benefit.

35:17

If you're gonna test it at least test shutting it off,

35:21

that is my one big pitch, huge pitch.

35:24

Second, you can do this hard to measure dashboard thing, right?

35:28

So, across all these channels,

35:29

there are ways to measure each one's potential,

35:32

not return on investment, but vanity metrics sort of model

35:37

and then build a model of your own, right?

35:40

You can build something like this that says,

35:42

"Oh, okay, on brand interest stuff,

35:44

this is what it was like in July."

35:47

That turned into this in August,

35:49

that turned into this in September.

35:51

What kind of lifted we see,

35:53

how many more customers did we get, all of that?

35:56

And a few software companies are building products like this.

35:59

This one is from Brand Ops,

36:00

which I think is actually quite good

36:02

if you need a kind of high quality measurement dashboard

36:06

as opposed to attribution dashboard.

36:08

I think they've got a nice solution for you.

36:10

And option number three,

36:11

you measure the big stuff and ignore the rest.

36:15

Almost every marketer hates this part,

36:18

but at SparkToro, we don't just reject attribution,

36:22

we sort of ignore most measurement entirely.

36:24

We built this dashboard, I ran it for a few months,

36:27

and I was like, this is not worth it.

36:29

It's so much energy and effort to measure the stuff.

36:32

Properly testing and measuring, properly building models

36:35

is so costly, so time intensive

36:39

that it is more worthwhile to waste a big portion

36:43

of my marketing effort.

36:44

I would rather waste money on the channels, waste money,

36:49

sorry, not digital advertising,

36:50

but waste organic effort, right?

36:53

In my time, Amanda's time,

36:54

doing things that, well, we thought this was gonna work,

36:57

but it turned out it didn't work.

36:58

I am happy to throw an hour at a podcast recording

37:03

that ended up only reaching 50 people

37:06

and it wasn't our core audience.

37:07

That's fine.

37:08

I'd much rather do that than waste 100 hours

37:12

building this dashboard or hiring a full-time data science

37:14

person and a full-time measurement person

37:16

and waste a ton of KCMICTO's time

37:19

building all these measurement tactics

37:21

into our platform, adding much of stuff.

37:25

What we do is we find sources that reach our audience,

37:28

we try to provide unique value,

37:30

we hopefully wow that audience

37:32

and then try and bring them to our site, right?

37:35

So like, when I do that podcast,

37:37

I'm asking, hey, go check out SparkToro

37:40

if you wanna try the Free Audience Research tool,

37:43

and then hopefully turn their audience into our audience.

37:46

This model over and over again is what has driven

37:49

big results for us and that the core part of it

37:52

is not attribution or even measurement.

37:56

It's good intuition on those sources

37:58

and good content and good intuition

38:03

and audience understanding and empathy

38:05

for what's gonna impress people.

38:07

We don't use paid media, we do virtually no SEO,

38:10

we get no SEO traffic outside of our brand name,

38:12

but all these spikes are email newsletters

38:16

and interviews and industry podcasts, right?

38:19

We're essentially saying, we want a source of influence

38:23

for our audience to come amplify our brand.

38:26

What could we do with them that would earn that coverage?

38:29

And then the list of things is infinite.

38:31

This is challenging, it's challenging,

38:35

but I am gonna argue it is a far better use

38:39

of virtually all small and medium business marketers,

38:42

time and energy and the other stuff, right?

38:47

The way I measure is are my customers engaged

38:50

with this platform and the topics that are relevant?

38:53

Do I like participating here?

38:55

Does it fit with the things that I'm good at

38:57

and I get energy from?

38:58

And are my vanity metrics going up, right?

39:01

More followers, more clicks, more total visits,

39:05

more people try on the tool.

39:07

That's what we do, might work for YouTube.

39:09

- Oh, I think I'm supposed to mention Colin,

39:14

that if folks like, you can try SparkToro,

39:18

which sort of helps with a lot of the,

39:20

for free at SparkToro.com.

39:22

- It's truly very, very cool.

39:27

And because you're being so humble about it,

39:31

I'll tell people, you can put in, for example,

39:33

it does many, many things, but put in,

39:36

hey, I wanna know more about the people

39:39

who follow Rand on Twitter.

39:41

Are you on Twitter?

39:43

- Yeah, I mean, who's on Twitter anymore?

39:45

But it'll go across social channels and websites and stuff.

39:49

- Yeah, so I wanna know more about the people

39:51

who follow Rand, look that up and I'll tell you,

39:55

they also follow these other people.

39:57

Maybe you can get on their podcast

39:59

and also they go to this website a lot

40:02

and they search for this kind of thing frequently.

40:04

And so like it gives you these kind of indicators

40:08

of where you can meet your audience,

40:10

which like you said, Rand,

40:12

I think one of the fundamentals of marketing.

40:14

So you're not getting lost in these little like

40:16

maybes and data and like maybes and attribution.

40:19

You go like, yeah, this is what my audience likes and wants.

40:22

Just like it's the first principle, go there, do that.

40:25

- Yeah, I mean, look, Colin,

40:27

I was a huge fan of surveys and interviews,

40:31

but I wanna, you know, I have empathy for folks

40:34

and recognize that a quality survey

40:37

in an interview process is like six months

40:39

and at minimum, tens of thousands of dollars, right?

40:43

If you're doing it sort of at professional scale.

40:46

And Casey and I were like, that sucks.

40:49

That process sucks.

40:50

Could we just do, you know, for the whole internet

40:55

that core process for marketers, like tell people

40:59

what YouTube channels does your audience follow?

41:01

What podcasts they subscribe to?

41:02

What email newsletters are they subscribing to?

41:04

What websites do they visit?

41:05

What keywords they search for?

41:07

Like as opposed to having to sort of one by one tease that out

41:11

of from surveys and interviews,

41:13

which have accuracy problems too,

41:15

can we just use public data on the web

41:18

and clickstream data from private panels to get that?

41:22

And the answer is, yeah, it works, right?

41:25

Like it's really cool.

41:27

It really does help solve this, that frustrating problem.

41:31

And you know, when like I present a lot of problems

41:34

like these, right, the zero click attribution,

41:37

the dark social thing,

41:38

and I don't wanna present those without some,

41:42

well, what's the next step, right?

41:44

And what I don't like doing is just pitching SparkToro.

41:48

So like there are alternatives too.

41:50

If you have an enterprise scale budget,

41:52

I really do recommend brand watch.

41:53

I think they have an outstanding product.

41:55

It's expensive, but you know, you talk to someone

41:58

on their sales team, you'll see a demo.

42:00

It's quite good, right?

42:01

It's kind of like SparkToro, but at enterprise scale.

42:05

And there's another company in the,

42:08

I think they're in the EU called Helixa.ai,

42:12

like and another one called Audience,

42:15

like Audience, but with an S instead of a C.

42:18

So those are all good alternatives too.

42:21

- Thanks.

42:22

I've got a couple of questions.

42:24

Alex, thanks, Alex.

42:25

Alex says, "Randy, how do you think this timeline we're on

42:29

will affect the direction of,

42:30

or the director of demand gen job in marketing?

42:33

Will that role become harder to be successful in

42:36

and to justify?"

42:37

- Probably the answer is yes, but it really depends.

42:46

I think that there,

42:48

unfortunately for the next decade,

42:51

what I see is Google and Facebook and Apple and Amazon

42:56

using their kind of monopoly power

42:58

in the digital advertising world to continue to convince

43:02

mostly CFOs and CEOs that their data is good enough

43:07

and more accurate than what, you know,

43:11

some CMO or some director of marketing

43:14

or someone who heard Rand on a podcast would have you believe.

43:18

And so you should keep investing with them.

43:20

And because so few boards of directors

43:24

and executives are willing to risk a down quarter,

43:30

they would rather sort of pay the kid with the green flyer

43:34

to take credit for sales that were already gonna happen

43:36

than risk losing the 8% lift in sales.

43:39

So, does that suck?

43:44

Yes, do I think it's dumb?

43:45

Yeah, does it give you a fantastic way to compete

43:49

against big companies by outsmarting them?

43:52

Yes.

43:53

- To some extent, it's, I think about this all the time

43:59

and I think maybe because marketers like to gripe and moan,

44:01

but like, you know, nobody, my CEO is not telling my VP

44:06

of finance how to do the accounting, right?

44:12

'Cause there's rules and laws

44:13

and they have to do the right things.

44:16

And so the CEO goes, I'll just trust you.

44:20

- Yeah.

44:21

- Until you mess up.

44:22

That never happens with a marketer.

44:24

They never go, okay, I'll just trust that you know

44:26

what you're doing.

44:26

These companies have made it just simple enough

44:31

that anybody who can look at a dashboard

44:33

can kind of grok what they want them to grok,

44:37

which is like, hey, we made you money.

44:39

So there's kind of like a cultural problem

44:43

in the way that marketing is trusted or not trusted

44:46

from the C-suite.

44:47

And I think to some extent we've done it to ourselves

44:49

through our obsession with attribution.

44:52

- Yeah, yeah.

44:53

Well, and I mean, we sort of had the rug pulled out

44:55

from under us, right?

44:56

So Google was sort of like, hey, here's all this amazing data.

44:59

Do you remember Pre Cambridge Analytica calling?

45:02

Facebook used to show you demographics, psychographics,

45:05

other pages that your audience visited, you know,

45:08

all this stuff.

45:08

So the bigger of an audience you got on your Facebook page

45:12

or Instagram or whatever, the more data

45:15

these players would provide to you.

45:17

It was really quite extraordinary.

45:18

In fact, the Facebook, the old 2016,

45:23

pre-2016 Facebook audience page looked a lot

45:27

like what Spark Toro looks like today, right?

45:29

It was like, it was quite similar.

45:31

Is there, you know, you got the pie chart with your gender

45:33

and age breakdown and then job titles

45:35

and descriptions and roles and, you know,

45:38

people who visit your page also visit these pages

45:40

and websites and follow these other people,

45:43

that information was gold and they took it away

45:46

under the guise of privacy, but realistically,

45:49

they took it away because now you have to spend money

45:51

with them to reach your audience

45:52

instead of knowing where you can go organically,

45:55

which I think is BS, right?

45:57

That data is your data.

45:59

Like that, you built that audience, it belongs to you.

46:03

I think it is shadiest hell that Google took away keyword data.

46:08

I think it's shadiest hell that this dark social problem exists.

46:12

I think it's shadiest hell that advertising options

46:14

like performance max take away all of your data

46:17

and then promise to tune things for you,

46:19

but then you'll never be able to fix it

46:20

or know what's wrong in the future.

46:22

It's just, I think Corey Doctorow has a great blog post

46:25

about this that folks can look up.

46:27

It's called the inshitification of platforms.

46:31

- It's so good.

46:31

- Which, yeah, which sort of describes the trend

46:36

of how all of these platforms operate, you know,

46:39

Amazon being a great example of one where it's like,

46:41

hey, here's this, you know, very cheap and inexpensive place

46:46

to buy goods and services delivered over the internet

46:49

incredibly fast.

46:50

Really, really good from a customer service

46:52

and consumer standpoint.

46:54

But then, right, oh, suddenly Amazon is often much more expensive

46:59

than a lot of other places on the web

47:01

and almost everything in the paid ad on Amazon, right?

47:06

It's so hard to find something organic.

47:08

The ads are so tiny, you're rarely clicking

47:12

on them even intentionally.

47:13

And so now it's become, you know, just not a great choice

47:18

but a default choice because you opted into Prime

47:20

because, well, the TV thing and the instant delivery

47:24

and the free shipping and the free returns.

47:26

And when I go to Whole Foods, you know,

47:29

I don't want to pay full price, all of that.

47:31

So Prime. - Yeah.

47:32

- Yeah.

47:33

I'll try to find it, Alec.

47:37

- Oh yeah, let me, it's EN,

47:41

shitification of the internet.

47:46

(mumbles)

47:49

I'll put it in the chat for you there.

47:52

- Oh, I lost my train of thought.

47:56

What were you just saying?

48:01

You were just talking about,

48:02

oh, here's what I was gonna say.

48:06

There's this trend towards, especially, I guess,

48:11

in B2B, and there's some of it in B2C2,

48:14

marketers talking about, you have to make a media company

48:18

or, you know, owned media, first party data,

48:23

there's community building, which is, you know,

48:26

what this is, club PF, people first,

48:29

if anybody here is not a member, check it out, club PF.co.

48:32

It's for people who believe that, you know,

48:37

marketing is about human beings.

48:39

And so like, if Rand's talk resonated with you,

48:42

I think the rest of what's going on in that,

48:44

in our community will.

48:46

So anyway, what do you think about that,

48:47

like that movement to build a media brand

48:51

or like a community?

48:53

Do you think that's part of a solve?

48:55

- For some people, yes.

48:59

Like, one of the things I really shy away from Colin

49:02

is suggesting to any one marketer or any company

49:07

that, hey, this thing worked for me,

49:11

and so you should just copy that and do it this way.

49:13

Or, you know, community building is a great channel,

49:16

everyone should do community building.

49:17

That's absolutely not true.

49:19

I think the only universal advice I can really give

49:22

when it comes to choosing channels and tactics

49:24

is choose channels and tactics that reach

49:28

the audience you need to reach, right?

49:31

That includes your customers,

49:33

but probably also includes the people

49:34

who influence your customers,

49:36

which are often a different group, you know,

49:38

whatever media, journalists, podcasters,

49:40

bloggers, social media influencers, all that kind of people.

49:44

And then choose channels and tactics

49:49

that you are uniquely good at, right?

49:53

Like you have some strength there,

49:55

you can provide unique value

49:57

that other people who are playing in that space

49:59

cannot provide.

50:00

So maybe that's, you know,

50:01

you're really great at data journalism,

50:03

and so blog posts and PR is terrific.

50:06

You're really good at community,

50:07

and so building a community is, you know,

50:09

a good opportunity, as long as the people you're trying to reach

50:12

are people who would actually enter a community.

50:15

And then the third one is do things

50:17

that you personally enjoy doing and get energy from.

50:22

Because I have never seen it happen.

50:24

I've never seen a marketer or a founder who's like,

50:26

"Ugh, I hate Instagram so much,

50:29

but gosh, I'm really good at it."

50:31

Nope. Nope.

50:33

That's how it works, right?

50:36

So if you want to find the channel

50:38

and tactics spread that, or structure that I recommend,

50:42

it's sort of hitting all three of those notes.

50:44

That's very different from recommending

50:48

like everyone should XYZ.

50:50

Yeah, there's this thing that I was just talking about

50:54

in somebody on LinkedIn yesterday

50:55

where you get the CEO who like constantly forwards you stuff,

51:00

your biggest competitor is doing and says,

51:02

"Why aren't we doing this?"

51:04

And like, "What do you say to that?"

51:06

And it's because they're doing it.

51:09

(laughs)

51:11

We're better at this other thing.

51:13

There has to be like a good answer for that.

51:16

And a lot of marketers aren't prepared with a good answer.

51:18

They go, "Oh, I'm sorry."

51:19

And then they start doing it.

51:20

And that's not the right thing to do.

51:22

I absolutely agree.

51:24

I mean, I think that the power of focusing

51:28

on your unique value and what you're good at,

51:32

and then you can, right?

51:33

Like there's things you can learn

51:35

from competitive intelligence and looking at competitors

51:37

and your CEO forwards you to that thing.

51:38

And you're like, "Oh, well, we shouldn't do that exact thing

51:41

'cause we hate that channel.

51:42

"That channel doesn't work for us

51:43

"or it's not part of our mix, hasn't performed well for us,

51:46

"whatever."

51:47

But we can take the intuition of using this sort of,

51:51

whatever, ideation or emotionally driven ad or whatever,

51:57

and do that in the channels and tactics

51:58

where we're good at, right?

52:00

And so then you sort of are like,

52:02

"Oh, great idea, boss.

52:03

"I'm gonna apply that to the thing that we do well."

52:06

So boss, as opposed to being like,

52:08

"Oh, yeah, I'm just gonna go and jump on this hype train."

52:13

Yeah.

52:14

Yeah.

52:15

It's the hedgehog concept that I--

52:19

Like exactly.

52:19

I learned it from Jim Collins, yeah.

52:22

I think there was, there was a great question that--

52:27

Did I skip something?

52:28

How do you justify the spend this is from--

52:32

I think this is from Bethany.

52:35

How do you justify the spend to a board with the attribution,

52:39

without the attribution or measurement?

52:41

I.E. when they say,

52:43

"If you can't tie these marketing activities directly

52:45

"to a sale, why wouldn't we just invest

52:47

"in more sales folks?"

52:48

So, I think this is the core,

52:52

like the reason this presentation exists, right?

52:55

That all of this, hey, let me set the context for you.

52:58

Here's all the reasons why attribution is sort of BS anymore,

53:01

all of that.

53:02

Unfortunately, it is marketers job

53:06

to educate their boards, their CEOs,

53:09

about all of this stuff, right?

53:10

Like I'm presenting it to you,

53:12

but you can kind of be like,

53:14

"Ah, goddamn it, Rand.

53:15

"Like now I have to go convince my board

53:18

"that all these things that you're saying are true."

53:21

Yes, I'm sorry, right?

53:23

Like that sucks.

53:24

That's also part of the job of being a marketer, right?

53:27

Part of the job of being a marketer,

53:28

or a CTO or a CFO, right?

53:30

Like the CFO has to tell the CEO, like,

53:33

"Oh, why are we changing accounting practices to this way?

53:37

"Well, it's because of these laws and regulations

53:39

"or because it's gonna help us with forecasting

53:41

"or whatever it is, and let me educate you on that."

53:44

Same thing, marketers have to educate their teams of,

53:46

"Hey, you got used to this thing for 20 years?

53:48

"It does not work that way anymore.

53:50

"Here's why, here's what changed in the last five years.

53:53

"Here's what's about to change with third party cookies.

53:55

"Here's what happened when all these companies

53:57

"cut back on their spend, right?

53:58

"So let's think about how we invest in marketing

54:02

"in a different way.

54:04

"I really do apologize that that sits

54:06

"on the shoulders of ordinary marketers.

54:08

"I wish I could personally get in front of all your boards

54:11

"and be like, "Hey, it's like this."

54:14

But, you know, maybe the recording will be helpful to folks.

54:19

- I think it's especially hard in B2B,

54:21

because in B2C you can do the test relatively easily.

54:24

Shut off your ads for even a week,

54:27

and you'll see whether they were giving you a lift or not.

54:30

Maybe not.

54:32

- Yeah, I mean, I usually recommend at least a month or a quarter.

54:37

And that is because one of, I think one of the realistic

54:45

counterpoints to, you know, shut off your digital ad spend

54:48

and, you know, see how many sales continue to come in,

54:50

is that two things are true.

54:53

One advertising can speed up or slightly accelerate the journey.

54:58

So for example, let's say someone is, you know,

55:01

was going to buy your software

55:04

or was going to buy, you know, your new book.

55:08

But you shut off the ads and so instead of them buying it

55:13

in seven days after they first learned about it,

55:16

it took 14 or 21 or three months.

55:20

That can happen, right?

55:23

And so you're essentially watching,

55:26

you're trying to see a longer term trend with the shut off

55:29

in advertising rather than just the one-to-one measurement,

55:34

because we do know that advertising can, you know,

55:37

accelerate a sale that was going to happen.

55:40

And there's also the brand effect, right?

55:42

So some advertising is not just transactional,

55:46

but brand advertising, I would argue that most digital marketing,

55:49

digital ad spend, you know, a Google ad, a Facebook ad,

55:53

are not brand.

55:55

They are action driven, not awareness driven.

55:59

So your mileage may vary.

56:03

I think it's reasonable to have counterpoints around that.

56:06

- Makes sense.

56:09

That's time.

56:10

Thank you so, so much, you've been so generous to come

56:13

and share your thoughts and arguments with us

56:17

that it's now our burden to carry back to our C-suite

56:19

and board.

56:20

- You're welcome, slash I'm sorry.

56:23

(both laughing)

56:26

- People follow, follow around, go check out Spark.

56:30

You will not regret it.

56:32

And we'll see you around in Club PF.

X

This is a test comment.

X

This is a longer test comment to see how this looks if the person decides to ramble a bit. So they're rambling and rambling and then they even lorem ipsum.


author