Latest / Genuin Connections / Genuin Connections: Episode 13 -- Debra Aho Williamson, Founder at Sonata Insights

Transcript

Genuin Connections: Episode 13 -- Debra Aho Williamson, Founder at Sonata Insights · Genuin Connections

0:05
This is genuine connections featuring
0:07
collaborative conversations about
0:09
community, content, creativity,
0:12
commerce, and the convergence of media
0:14
and marketing.
0:19
Hi, I'm Matt Worse, CMO at Genuine, and
0:21
welcome to this episode of Genuine
0:23
Connections, where we'll be joined by
0:26
one of the sharpest, most strategic
0:28
minds in this emerging AI economy. She's
0:32
someone who's been predicting and
0:34
explaining the biggest shifts in media
0:36
and marketing for nearly two decades. I
0:39
first met Deborah when we were building
0:42
early social media into a marketing
0:45
experience and ecosystem as well and now
0:49
she's become a leading expert in just
0:51
yet the latest digital trend really
0:54
revolution before founding Sonata
0:56
Insights Deborah A Williamson spent 20
0:58
years as a lead analyst at e-arketer
1:02
where she wrote more than 120 reports on
1:05
digital platforms social media usage
1:09
trends, advertising behavior, and really
1:12
the evolution of online consumer
1:15
interaction. And during that time, she
1:17
helped marketers, agencies, platforms,
1:20
startups, investors understand not just
1:24
what was changing, but why it matters.
1:27
Deborah was early to spot the power of
1:30
social, early to articulate how digital
1:32
discovery was going to reshape search
1:34
and advertising. And now today she's
1:37
leading the industry conversation about
1:39
how AI is redefining the rules of media
1:43
and engagement. I read her stuff
1:45
religiously. Her work, including her
1:48
newsletter, the AI ad economy, which you
1:50
should all go follow on LinkedIn, helps
1:52
brands understand how people are using
1:55
AI to discover and evaluate and decide
1:58
what that means for the future of
2:00
marketing and advertising. And in her
2:01
newest research, which we'll talk to her
2:03
about, she challenges this old notion
2:06
that organic and paid media live on
2:08
separate planes and separate clocks.
2:12
She's showing how AI environments are
2:14
collapsing discovery and decision into
2:18
one place and how brands have to rethink
2:20
credibility and authority and
2:22
performance in an age where answers, not
2:24
clicks, will drive outcomes. Something
2:26
that we talk a lot about at Genuine.
2:28
Now, she advises companies on how to
2:30
blend organic presence with AI
2:32
advertising, how trust, another topic
2:35
that we love, how that becomes
2:38
performance currency and what marketers
2:41
need to do now if they want to lead
2:43
rather than follow in this next era. So,
2:46
without further ado, let's welcome to
2:48
the show Deborah Aaho Williamson.
2:52
>> Hi Debbie, welcome to Genuine
2:54
Connections. So grateful to have you
2:55
here with us. Thank you. Well, thanks
2:58
for having me and it's really great to
2:59
catch up with you because it's it's been
3:01
a while, right?
3:02
>> It's been it's been a minute. Certainly
3:04
in the digital marketing media life
3:08
cycle here. We've probably been through
3:09
three or we missed web three. We didn't
3:12
get to catch up during that phase. The
3:14
metaverse I don't think we caught up
3:15
during, but now really substantial
3:18
movement that's happening. So, I'm glad
3:21
you're here because I you are at the
3:23
forefront as both an analyst and a
3:26
thought leader. So, I'm excited to jump
3:27
in with you. Are you ready?
3:29
>> I am ready. Let's go.
3:30
>> Okay. You've been at the center of
3:33
multiple digital revolutions from really
3:37
ad supported web to social, which is how
3:40
you and I first met, to now really the
3:42
AI transformation. Is there anything
3:45
about this moment, this shift as we kick
3:48
off that's different than anything
3:51
you've felt, seen, or experienced so
3:54
far?
3:54
>> Yes. uh the pace
3:57
>> the pace of change is so much faster um
4:02
because AI AI moves fast.
4:07
>> Um and
4:09
>> if you you look back at what we were
4:11
capable of doing with AI um and what
4:14
even the experience as a consumer of
4:16
using a platform like chat GPT was even
4:20
a year ago uh to what it looks like now.
4:23
um things are moving so quickly. So I
4:26
would say that is probably the biggest
4:28
differentiation from the digital
4:30
transformations I've seen before is that
4:33
um everything is moving very very
4:35
quickly and the marketing community um
4:38
you know we're always racing to catch up
4:41
uh with the latest and greatest and I
4:43
think the challenges are even bigger
4:45
with AI than they were in any of the
4:48
previous eras that I've been involved
4:50
in. And maybe this episode needs an
4:52
expiration date because it's all going
4:54
to be outdated in like six weeks anyway.
4:56
But when you think about that consumer
4:58
behavior, which is something I know that
5:00
you've been consistently tracking
5:02
throughout,
5:04
what are the clearest signals that AI as
5:07
an umbrella term is already shaping how
5:11
people discover and evaluate and even
5:16
buy things?
5:17
>> Yeah, I mean so much to unpack there. So
5:19
I think the best thing to think is to do
5:21
is to start with how I think about AI
5:23
and you're exactly right. Um the
5:25
conversation I mean AI isn't new. It's
5:28
you know decades old. We've seen AI
5:30
being integrated into everything from,
5:33
you know, Google Maps to Netflix
5:35
algorithm, you know, recommendations um
5:38
on the back end. Things that maybe
5:40
consumers weren't necessarily
5:42
engaged with or realizing that they were
5:44
using AI, but clearly AI has been and
5:47
and machine learning along with it has
5:49
been part of the foundation of a lot of
5:51
the great technological advances that
5:53
we've seen over the past years.
5:56
um where I really draw my attention and
5:59
thank you for pointing out the consumer
6:01
angle because um basically everything
6:03
that I do when I research and when I
6:05
think about a new technology or a new
6:08
digital transformation moment is I start
6:10
with a perspective of a consumer and
6:12
what is their experience and how are
6:14
they interacting and are wanting to
6:16
engage with it and how are they wanting
6:18
to use this as part of their life um
6:21
making a choice to do it not like in
6:23
Netflix where you know you're kind of
6:24
looking through Netflix and the
6:26
algorithm is showing you certain things
6:28
and it's AI that's driving that but you
6:30
as a consumer don't really know that you
6:32
are in this case actively saying I'm
6:35
going to go to chat GPT because I'm
6:37
planning a trip to as I am Yuma Arizona
6:40
to go catch up with some friends in a
6:42
couple of weeks. I've never been to Yuma
6:44
Arizona. I could go to Google but hey
6:46
I'd rather go to chat GPT and say hey
6:48
ChatGpt tell me about Yuma Arizona. Um,
6:51
that's what I'm talking about when I
6:53
talk about AI and the digital
6:56
transformation associated with AI. It's
6:58
really about how are people using
7:00
consumers using AI and how do marketers
7:03
fit into that conversation. Well, and
7:06
there are tools that obviously are
7:08
happening behind the scenes from both a
7:11
development and curation
7:15
perspective in addition to the creation
7:17
that we've talked about that and I think
7:20
this is somewhat new from an AI
7:22
perspective because you're right, it's
7:23
been around in a lot of different
7:24
iterations but sometimes it was over
7:27
here in some group that was doing
7:29
something interesting or maybe not so
7:31
interesting but important. Now it has
7:34
become pervasive in all aspects of
7:36
business. But from a marketer
7:38
perspective,
7:39
I I go back and forth on this, but do we
7:42
think that marketers are ahead of
7:44
consumers on AI adoption or mostly
7:46
reacting to what consumers are already
7:48
doing?
7:50
>> Yes.
7:53
>> Okay. So, so, so yes and yes. when
7:56
chatbt was launched in 2022 into 2023,
8:01
uh you started to see just a little bit
8:03
of consumer usage of it, but a whole lot
8:05
of kind of marketing industry um
8:08
awareness of AI and what it could do. So
8:11
I look at the years like 2023 and 2024
8:14
as times where um marketers spent a lot
8:18
of effort you good valuable effort
8:21
thinking about how to use AI inside
8:23
their organization um to make their
8:26
marketing processes better faster more
8:28
efficient uh in you know more integrated
8:32
into what they're doing in their daily
8:34
lives as business people and those are
8:37
valuable things to think about. Um but
8:40
what I worried about already back in
8:43
those day earliest days was okay well
8:47
don't forget to think about what's going
8:49
on outside of your company and how your
8:52
external marketing also needs to change
8:54
because consumers are going to be using
8:56
these tools more and more. I mean I knew
8:59
this a couple years ago just from seeing
9:01
what was out there. And so I think
9:04
marketers were ahead of consumers in
9:06
that they were already focused
9:08
internally for the past several years.
9:11
Very smart again, very wise to figure
9:14
out how to use AI to orchestrate your
9:16
marketing to improve and make make it
9:19
more efficient. We're seeing lots and
9:20
lots of great use cases for that. Uh but
9:23
what I didn't want companies to forget
9:25
is that the consumer behavior was going
9:27
to start ramping up. And it was early,
9:30
right? A couple years ago, we saw, you
9:32
know, a few people like early adopters
9:34
going into chat GPT and kind of poking
9:36
around at it and seeing, oh, well, this
9:38
looks kind of interesting. I'm going to
9:39
see what I'm going to do here. Uh but as
9:41
the months have gone by and now we're
9:44
into 2026 and ChatGpt
9:47
uh is a couple years old, it's becoming
9:50
a bigger and bigger foundation of
9:52
consumer behavior. And you're seeing,
9:53
you know, upwards of like say 50 60% of
9:56
people in surveys saying yes, I use AI
10:00
in some way. And when you see that
10:03
happening in such a quick time frame,
10:05
you know that this is something that
10:06
then now marketers need to start
10:08
catching up to from an external
10:10
marketing perspective.
10:12
Well, what it I think has resulted in um
10:16
at least from my perspective is the
10:18
blurring of lines between product and a
10:22
consumer experience and marketing as a
10:25
consumer experience. And it depends
10:26
obviously on what the intended or
10:30
business model is in each individual
10:32
case. But every marketer now needs to
10:35
understand product in a way that they
10:37
haven't before. And that is certainly a
10:41
new mindset that hopefully people are
10:44
open to in this industry and at a
10:47
leadership level but also depends on the
10:50
environment. So, how much of the AI
10:55
opportunity is about ads
10:59
themselves versus context and experience
11:04
where and how those ads actually
11:07
engage or live within?
11:10
>> Um, I think it's both honestly. I just
11:14
wrote so I have a Substack called the AI
11:16
ad economy and I just um recently wrote
11:19
a a whole article about the idea that um
11:23
being present in these environments as
11:25
marketers organically um in other words
11:29
AEO answer engine optimization GEO u you
11:33
know use the term what whatever term an
11:35
you know AI search whatever it is being
11:38
present in those environments
11:40
organically is incredibly important for
11:42
brands um because people are forming
11:45
opinions. They're learning about
11:47
products uh from the first time all the
11:50
way to being able to purchase them in
11:52
these environments based on the things
11:54
that the chat that uh chat GBT or AI
11:57
overviews you name it co-pilot knows
11:59
about them. So that's super important.
12:02
Um but what's very interesting to me is
12:04
the layer on top of it which is the paid
12:07
advertising.
12:08
ChatGpt sells advertising. You know,
12:10
Copilot has advertising obviously, you
12:13
know, Google has advertising within AI
12:15
overviews and and an AI mode. And so
12:18
marketers need to make sure that they
12:20
integrate whatever this new growing paid
12:24
advertising opportunity alongside the
12:27
very important task of being present
12:29
organically when someone is searching
12:31
and looking for something that is
12:33
related to what they're selling.
12:36
Well, in the last iteration of how this
12:38
all came to be with search, right, there
12:40
were behaviors and search engines and
12:43
experiences that happened in
12:46
Google and Bing. But ultimately, when
12:50
you were on site on a brand's website,
12:52
on a commerce site, a retail site, even
12:54
content publishers, whatever it may be,
12:57
they had to adapt to the consumer
12:59
behavior by creating on-site search. And
13:03
I think the same thing is going to
13:04
happen where the behaviors of AI,
13:08
you know, prompting and searching and
13:12
result engineering and optimization will
13:15
need to also happen in environments
13:17
where people are still going and in some
13:19
cases where trust is higher. I'm more
13:22
likely to trust something that happens
13:24
in an environment or in a branded
13:26
ecosystem that I already know and have
13:28
shopped at as opposed to maybe something
13:29
new. But now there's a layer of
13:32
connectivity that may not have existed
13:34
before. That's we can get into the
13:37
technical terms, right? The protocols
13:38
that connect these things. But even the
13:40
behavior around conversational AI can
13:43
compress discovery and consideration and
13:45
all that the the the funnel right into
13:49
one interaction in one place. So what
13:51
does that do for traditional funnel
13:53
marketers and the behaviors that they've
13:56
learned and grown up building?
13:59
Well, I mean, I think we've spent a few
14:01
years now talking about how the
14:02
traditional funnel is already broken.
14:05
So, I would hope that smart marketers
14:07
have already realized that,
14:09
>> but now it's broken again is what I'm
14:11
right. Like, it's it's even more
14:13
compressed.
14:14
>> But you're right. What AI does is it it
14:16
it does compress that possibility of
14:18
getting from awareness to purchase into
14:21
one complete conversation. And it could
14:24
be a really short conversation. You
14:26
know, one example that I um like to use
14:29
is, and this actually happened to me a
14:31
couple years ago, but it's still really
14:32
relevant, is I was searching for a new
14:35
brand of facial moisturizer because the
14:36
one that I normally purchase at
14:38
Nordstrom was out of stock. And so, I
14:40
went to Chat GPT and I thought, well,
14:42
I'll just use this as an opportunity to
14:44
not only find something similar to the
14:46
product that I was looking for, but
14:47
maybe cheaper because the one I was
14:48
using was kind of exper expensive. So,
14:51
um, within about a 15 to 20 minute back
14:54
and forth with Chat GPT, I became
14:55
introduced to not one, not two, not
14:58
three, but four brands of moisturizer
15:00
that I'd never heard be of before. Um, I
15:03
learned about them all. I chose one. I
15:06
decided that I was going to purchase it.
15:07
I went to Amazon. I bought it. I'm now a
15:10
loyal customer of that brand of
15:11
moisturizer. And I have not gone back to
15:14
the purchase that I made at Nordstrom
15:16
for years and years. That's how quick it
15:19
happens. And I think that's the story
15:20
that I think marketers need to wrap
15:22
their heads around.
15:24
>> Well, and they're all still figuring out
15:26
where and how to leverage human
15:28
influencers and creators. And now
15:30
there's yet another influential
15:33
engine that can drive this that again
15:38
the speed of this is is somewhat
15:39
staggering. and the speed. I go back to
15:42
how SEO, right, organic search and paid
15:46
search took some time to really
15:50
become more of a unified thought process
15:53
and strategy for a lot of marketers. I
15:56
wonder if AEO and paid AI media, I don't
16:00
know if there's an acronym for that yet.
16:02
U pay, not pain, pain may I don't know.
16:06
>> I call it PAM.
16:07
>> Yeah. Um
16:08
>> it could be pain for marketers but
16:10
>> we need pain relief right the is that
16:14
that convergence has to be happening
16:17
much faster than paid search and paid
16:20
social did right
16:22
>> absolutely and and yes by the way so the
16:24
the term that I'm I like to use is AI
16:26
media so think about social media think
16:28
of retail media
16:29
>> let's talk about it as AI media now that
16:31
when I and when I when I talk about AI
16:33
media just to put a definition around it
16:36
I I define AI media as an adup supported
16:40
experience inside an AI assistant, an AI
16:43
enhanced search or AI native interface,
16:46
but eventually actually AI agents as
16:48
well. But these are places where people
16:50
ask questions, research and make
16:52
decisions. And so when I think about AI
16:55
media and I talk about that as a term,
16:57
this is what I'm referring to. Uh but
16:59
you're right, the transition from
17:01
organic to paid is happening much faster
17:04
in AI than it happened in previous
17:06
digital transformations. So if you think
17:09
about going back to search and even the
17:10
days of the earliest websites, the
17:12
earliest websites started in 1994. Um
17:16
and then in uh four years later in 1998,
17:19
Google came around and people went and
17:21
realized, oh well um you know, we need
17:23
to actually be present in these search
17:25
engines like Google. Um so search engine
17:27
optimization became a thing. Um it
17:29
wasn't until two years later in 2000
17:31
when Google introduced advert paid
17:33
advertising. So companies had basically
17:35
a span of six years to get used to hey I
17:38
have a website now we need to advertise
17:40
it. And then if you look at social a
17:43
similar thing happened. So, Facebook
17:45
launched in 2004. Uh, you know, a few
17:48
years later, uh, we started to see
17:50
organic, you know, brand pages. Um, they
17:52
introduced brand pages and I think it
17:54
was around 2007. You might remember
17:56
that. You might have built a few brand
17:58
pages back in the day.
17:59
>> Rings a bell. Yeah.
18:01
>> Yes. Yes. So, it was all about being
18:02
organic and realizing, oh gosh, I could
18:04
use a platform like Facebook to talk to
18:06
my customers. Um, and then lo and
18:09
behold, five years after that, in 2012,
18:12
uh, Facebook kind of said, "Yeah, let's
18:14
let's not let you do that so much
18:16
anymore. We're going to turn down the
18:17
reach of your organic material and start
18:20
making you buy advertising." And, um, it
18:22
started introducing feed ads,
18:24
coincidentally, the same year in 2012.
18:27
So, brands did have a long period of
18:29
time to go from, okay, here's what
18:31
Facebook is. Let's build brand presence.
18:34
Let's buy advertising.
18:36
Let's look at chat GPT. Chat GPT
18:38
launched to 2022. Um couple years later,
18:41
we've all started talking about answer
18:43
engine optimization. 2026, less than,
18:46
you know, four years after chat GDP's
18:48
launched, it's starting to test
18:49
advertising. There's the compression
18:51
right there. This is how fast, getting
18:53
back to the start of our conversation,
18:55
this is how fast things are moving in
18:57
this business.
18:58
>> Yeah. I think the flip side to that
19:00
speed, right? There's a lot of
19:01
opportunity and positive value that
19:04
comes of it. There's also risk for
19:07
brands, right? Because it's still
19:09
a work in progress in many cases in
19:12
terms of accuracy. It's getting better
19:14
and that speed is increasing and
19:16
accelerating and the velocity of
19:18
adoption is increasing. But the risk of
19:21
brands being diluted or conquested or
19:26
rendered obsolete because they're not
19:30
optimizing the right way is also much
19:34
faster or um accelerating more quickly.
19:37
So how do you think about brands
19:41
protecting
19:42
from AI and is there a way that they can
19:45
protect from AI with AI?
19:49
That's a big question. So I think what
19:52
I'll start by saying is that um what I
19:55
think everybody has learned in the past
19:58
couple of years of doing or starting to
20:01
think about and activate on answer
20:03
engine optimization is it's really
20:06
complicated.
20:08
>> It's not like search. It's like search
20:11
on steroids because you have multiple
20:14
platforms that each interpret
20:17
information a different way. whether
20:19
it's Gemini or Copilot or Perplexity or
20:22
or or ChatGpt for example, each of them
20:25
has a different way of looking and
20:27
ingesting information. It goes to
20:30
different sources. You have to deal with
20:31
that. Um and then you have to consider
20:34
that the queries are going to be
20:37
different. You know, if I say to a chat
20:40
GPT, show me the best running shoes for
20:43
trail running, it might give me one
20:45
response. But if I just tweak it a
20:48
little bit and I say, I'm interested in
20:50
starting to learn trail running. Help me
20:53
understand what shoes might be
20:55
appropriate for me for that. It
20:57
interprets that context differently. And
20:59
it's not necessarily going to show you
21:01
the same results. So it becomes this
21:04
huge task ongoing task for businesses to
21:07
make sure that they stay that they stay
21:10
relevant and I don't think anybody's
21:12
really figured out the the the right
21:15
formula yet. Well, it even
21:16
differentiated, right? Because if
21:19
everything becomes
21:21
well risks becoming sort of a reversion
21:24
to the mean and maybe I'm
21:25
oversimplifying it in a way, but does
21:28
brand matter anymore or is everything
21:30
just like Amazon basics, right?
21:33
Batteries. It's all just batteries,
21:35
right? in terms of what the value is. So
21:37
that's the the threat and the
21:39
opportunity for brands to differentiate
21:41
some themselves whether it's organically
21:43
or through AI media boosting to get
21:47
above a generic result.
21:50
>> Right.
21:51
>> Absolutely. And I do think I think that
21:52
is one of the potential
21:55
powers of AI media. I mean we're early,
21:58
right? I
21:58
>> Yeah. I don't know that this is exactly
22:01
going to happen like I think it is, but
22:03
I think there is a potential for brands
22:05
to uh either use paid advertising on
22:10
platforms like chat JPT to boost what
22:12
the platform already says about them uh
22:15
kind of like the platform is an
22:17
influencer like the chat GBPT is an
22:18
influencer and you're boosting the
22:20
influencers's content. Mhm.
22:21
>> Uh or to um actually maybe even subtly
22:26
change the narrative or provide a
22:28
different viewpoint if for some reason
22:31
the what the chat GBT does says about
22:33
you isn't exactly how you want to be
22:35
perceived. Um, another thing and then
22:38
thinking outside of these platforms, I
22:40
think that there's going to be a way to
22:42
use paid advertising
22:44
on other platforms outside of plat chat
22:47
GPT and that's going to be ingested in
22:49
some way into how these platforms start
22:52
to think about you. So, um there's going
22:55
to be a lot of ways I think that paid
22:56
advertising becomes an important tool to
23:00
help companies not only be relevant but
23:03
like to your point to differentiate um
23:06
as consumers are starting these journeys
23:08
and figuring out and gravitating towards
23:11
uh AI platforms to solve like everyday
23:15
issues.
23:16
>> Some AI platforms are obviously
23:17
embracing ads, some are resisting them.
23:20
There will be a reckoning as the economy
23:23
plays out as is always the case. But the
23:27
accessibility and acceleration of this
23:30
technology means that other players may
23:33
evolve quickly too. And even you know
23:36
Walmart
23:38
is an AI platform like sorry but it is
23:41
and people may not think of it that way
23:42
but it is. So as everyone kind of
23:47
becomes more AI savvy and then AI led
23:50
not just AI powered which is a term I
23:52
hate but actually AI le what is the sort
23:56
of divergence between those who are
23:58
embracing
24:00
monetization media models platform
24:03
economics versus those that aren't like
24:05
what does it tell you about where the
24:06
ecosystem may land
24:09
>> marketers have a lot of learning to do
24:11
about AI
24:12
>> okay
24:13
>> going back to this idea that they're
24:15
still figuring out how to incorporate AI
24:17
into their organizations,
24:19
you know, internally in their day-to-day
24:20
processes. We're still at the earliest
24:22
stages of that. Um, and you know,
24:24
there's surveys that keep coming out
24:26
that say, you know, companies aren't
24:27
embracing AI as quickly as as they
24:30
should. And um, there's there's thing
24:34
things are happening that that, you
24:35
know, that are getting ahead of
24:36
companies. And then there's a few a few
24:39
companies that are saying, "Oh, you
24:40
know, we've embraced AI so completely
24:42
that we're going to, you know, fire half
24:44
of our company because we don't need
24:46
them anymore." Um, there's a lot of kind
24:50
of internal dynamics that are still
24:53
being developed within businesses and we
24:56
know, you know, you and I have been in
24:57
the marketing industry for a long time.
24:59
um new media formats come along and you
25:02
know there's the a few companies that go
25:04
out and decide I'm gonna you know I'm
25:06
gonna be I'm gonna test this. I know I'm
25:08
ready. I think I'm going to go out and
25:10
test this. Um but many many other
25:12
companies take a lot longer to embrace
25:15
it. U because they kind of got to get
25:17
their internal um organizational stuff
25:21
in order before they can start to think
25:23
about okay how do I actually layer this
25:25
media on um in
25:29
at scale. Uh or or maybe they're waiting
25:32
for uh you know the early adopters to
25:35
play around for a while and make the
25:37
mistakes and
25:38
>> past followers can win. Right. Yeah.
25:40
>> Right. Yeah. So, so I I mean I think
25:42
I've kind of gotten a little bit aside
25:44
from your, you know, soft track from
25:45
your question, but uh the companies that
25:49
I think are going to have the advantage
25:51
in this are are the ones that are going
25:53
to be uh simultaneously working on
25:57
improving their internal use of AI in
25:59
their marketing organizations, but also
26:01
keeping close tabs on how their
26:03
consumers are changing. Um, you know,
26:06
are they are they still going to Google?
26:08
Are they not? Um, are they using social
26:10
media less? Um, you know, to are they
26:13
listening to influencers less? I'm
26:14
making these scenarios up, but things
26:16
that people should make pay attention
26:17
to. Are these influencers become less
26:20
important? Because Chhat TBT is becoming
26:22
more of an influencer in their lives.
26:24
Those are the kinds of things that I
26:26
think will help marketers to if if they
26:28
pay attention to them will help them to
26:30
become, you know, better and more
26:32
integrated, you know, across both
26:34
internal and external. when it comes to
26:36
AI.
26:38
>> Yeah.
26:38
>> I remember in the early days visiting
26:40
the Facebook new headquarters when it
26:42
opened up in Menllo Park, they had signs
26:44
on the wall of probably like Mark
26:46
Zuckerbergms and one was fail fast,
26:49
right? So the faster you fail, the
26:51
faster you learn. And again, excitement
26:54
and risk both two sides of the same
26:56
coin. There are mistakes that early AI
26:58
media adopters will make and there's a
27:00
cost of waiting too long to experiment
27:02
in AIdriven media environments. I think
27:05
finding that balance is going to be
27:07
different for everyone. And it may come
27:09
down to
27:11
what your business objectives and model
27:15
is on one hand and what your audience
27:18
and consumer
27:20
needs are on the other, which is still
27:23
marketing 101. In in our world, in my
27:26
world, I think about publishers and
27:28
retailers and consumer platforms who are
27:32
all wondering how AI affects their
27:34
traffic. Is this is no one coming to our
27:36
site anymore or engagement like are
27:38
these real or is this genuine? But also
27:42
monetization. So, do you see from one of
27:47
their vantage points um a a real
27:50
opportunity for someone to kind of step
27:52
in like are retailers going to be the
27:54
ones that drive this forward? Is it
27:56
content and publishers who lead the way
27:59
that others follow from a business model
28:01
perspective?
28:03
>> Well, I mean, I think from a paid
28:05
advertising standpoint, I think the
28:07
first types of companies that should be
28:08
advertising in AI media would be
28:11
retailers. And I think we've seen that
28:13
borne out in some of the early
28:14
advertisers on chat GBT. Um you know
28:17
travel being another category. Um I
28:19
think you know electronics being a third
28:22
category based on really heavy use
28:25
consumer usage of these platforms to
28:29
research you know an electronics
28:30
purchase or a new phone or something
28:32
like that. So these are the kinds of
28:34
companies that I think should move
28:35
first. Uh and you know so so but but I
28:40
think when you think about it onsite too
28:42
um you know you mentioned Walmart
28:43
earlier uh Amazon obviously you know
28:46
Amazon has Rufus Walmart has Sparky you
28:49
know they are integrating AI into the
28:52
experience of using those platforms
28:54
which I think is really smart. So I
28:56
think on-site and onplatform experiences
28:58
of AI are actually going to be really
29:00
important uh not only because uh they
29:04
these companies can very easily
29:06
integrate them but also these are these
29:08
are places where people are comfortable
29:09
with right like we all shop at Amazon we
29:12
kind of know what it's like to shop at
29:13
Amazon we're not making a new behavior
29:15
necessarily to maybe say ask ask Rufus a
29:19
question in the same way that it's a new
29:21
behavior to go like externally in the
29:23
chat GPT. So I feel like and I've seen
29:26
you know we've seen data about this um I
29:29
can't remember who unfortunately but a
29:31
lot of um you know there's more and more
29:34
purchases on Amazon um are incorporating
29:37
conversations with Rufus in some way. Um
29:39
and I think Walmart is seeing some
29:40
similar traction with with Sparky and
29:42
its impact on on purchasing. So um
29:45
having these on-site experiences in
29:47
retail destinations I think is really
29:49
powerful. Um and that probably extends
29:51
into publishers as well.
29:53
>> Yeah. Yeah, I mean I think I mean
29:55
obviously Walmart and Amazon are the
29:58
biggest and have the most resource and
30:00
capital to commit to um these types of
30:04
not just experiments but actual like
30:08
investments and bets and infrastructure
30:11
change. I think the opportunity to your
30:13
point of building onsite may not be as
30:16
quick but this is you as as a partner
30:19
that can provide those services. I think
30:21
you're right and I I know you're right
30:24
because we've got the data to prove it.
30:26
But the data is also a good question
30:28
that I think others are going to start
30:29
asking. I have an answer, but I'm
30:31
curious what yours is because if clicks
30:34
or site visits
30:36
are important or engagement is important
30:38
or maybe not or is being cannibalized,
30:41
what becomes the new performance signal
30:43
that marketers should care about?
30:46
>> Metrics is a big question. Uh yes,
30:49
you're right. I mean, if you're engaging
30:50
with an AI experience on site, you
30:53
automatically have the benefit of that,
30:56
you know, of of of the metrics that come
30:58
from from that directly from that
31:00
>> zero party data we call it, right?
31:02
>> Yes. Exactly. So, and and I think that
31:05
gives these platforms a really strong
31:08
advantage in terms of who owns the
31:10
customer. Um, where it gets tricky is
31:13
the referrals from a chat GBT, the
31:15
referrals from an AI overview, the
31:17
referrals from from, you know, a
31:19
co-pilot or whatever. Um, and and how do
31:22
you when and how do you attribute that?
31:25
Um, these are the things that are all
31:27
still very much in development.
31:30
>> Trust is a topic that comes up a lot in
31:32
AI conversations.
31:34
It's paramount to what we've built and
31:37
why we've built what we've built. And it
31:39
keeps coming up. I yeah, I read your
31:41
your newsletter. I love it. It keeps
31:42
coming up in your research. How are how
31:45
are consumers balancing curiosity about
31:49
AI with the skepticism, especially
31:52
around advertising and branded content?
31:55
>> Yeah, you're you're right. I mean
31:56
there's a lot of layers to the trust
31:58
question in AI and um some of it has to
32:01
stems from do I trust that these AI
32:04
companies have my human best interests
32:07
in mind um or is AI going to just
32:10
suddenly take over and you know dominate
32:12
the world. I mean these are kind of big
32:14
existential questions that people are
32:17
confronting when they're using uh an AI
32:19
platform or thinking about the power of
32:21
AI. So um there's trust on that level
32:24
which I think is something that you know
32:26
everybody needs to get comfortable with
32:28
right you know do you think that these
32:31
are going to get so smart that they are
32:33
going to um you know take away the
32:36
humanity or they're going to start
32:38
interacting and engaging with themselves
32:39
and humans are going to be you know kind
32:41
of left to the side kind of getting big
32:44
and philosophical with you for a moment
32:46
here but
32:48
the thing that I think
32:51
coming down to the brand trust level and
32:53
coming down to the trust in in terms of
32:57
things that people typically care about
32:59
uh you know on a more granular level
33:01
which is like privacy um data uh is is
33:06
it going to hallucinate um these are
33:09
sort of the more practical types of
33:11
trust
33:13
>> I I liken this back to some research I
33:16
did on social media um when I was an
33:18
analyst at e-arketer And we literally
33:20
called it a digital trust report. Uh we
33:23
um surveyed people about their trust in
33:26
different social media platforms and
33:28
their level of trust and you know we
33:30
created a ranking and um lo and behold
33:34
very few people trusted social media
33:36
platforms um to protect their privacy to
33:39
uh you know
33:41
make them feel safe to show them you
33:44
know information that was accurate. Uh
33:47
so every year we published this report
33:49
saying oh my gosh trust is you know
33:50
brands be aware of this. Um but what
33:54
ended up happening was that didn't stop
33:56
people from using social platforms.
33:58
Every year the usage kept going up and
34:00
maybe a few people stopped using because
34:02
of this trust concern but overall
34:04
looking you know writ large at the type
34:07
at the consumer behavior people are
34:09
still using social media. So bringing it
34:12
back to AI, the reason I tell that story
34:13
is because I think this a similar thing
34:15
is going to happen with AI. People are
34:17
going to be concerned about their
34:18
privacy. They are going to be concerned
34:20
about hallucinations. They are concerned
34:22
about the idea that um you know that
34:25
that AI might uh give them
34:27
misinformation. Um they might be
34:29
concerned about the their their data and
34:32
what happens to it.
34:34
uh I don't think we're going to see
34:36
people pulling back, you know, in grand
34:39
scale from from this. I only think more
34:42
and more people are going to realize the
34:44
benefits of using AI and it's going to
34:46
increase.
34:48
>> Yeah. I think to to the conversation and
34:51
points we made earlier, even if it's not
34:53
in the environments where they're
34:55
currently engaging as an offsite
34:58
dedicated AI experience, if the
35:01
technology gets better but becomes more
35:04
ingrained into and integrated into
35:08
existing trusted brand experiences,
35:10
right? I shop with my daughter at
35:12
Sephora and Sephora has an incredible
35:16
back engine of AI curated content and
35:19
commerce enabled experiences.
35:23
I will continue to trust Sephora if the
35:25
technology delivers
35:27
in that in that example. So I think it
35:30
comes back to the environment again as
35:32
we were talking about earlier. So
35:34
>> absolutely.
35:35
>> Okay. So, one last question before I let
35:37
you go and thank you again for just this
35:39
incredible wisdom and uh and knowledge
35:43
here. As we think ahead, I don't want to
35:45
say five years because God, who knows,
35:48
but maybe
35:50
18 months to two years, what about AI
35:54
marketing will feel obvious then that
35:57
still feels uncertain today?
36:00
>> It's just even 18 months is a hard thing
36:02
to talk Okay, you you pick the time
36:05
frame up.
36:06
>> I'm glad you didn't ask me about five
36:07
years because it it you can't p you know
36:10
you could try to predict five years, but
36:12
yeah.
36:13
>> Um I I you know I followed social for a
36:15
long time. Social changed every single
36:17
year uh and something new happens. So I
36:19
knew you could you could try to predict
36:21
what was going to happen five years from
36:22
now, but you really were best to predict
36:24
maybe what's happen one year from now.
36:26
Um with AI, I could even predict six
36:28
months, but let me try let me just try
36:30
the 18-month thing just just for fun. I
36:33
think we're first of all, I think we're
36:34
all going to be in realization that um
36:38
showing up organically and in AI is a
36:41
lot harder than it looks and um
36:44
hopefully we're it's it it's going to be
36:47
something that I companies have
36:50
developed a few companies now stand out
36:52
18 months from now. Hopefully a few
36:54
companies now stand out as the ones that
36:56
actually have figured it out. I know
36:58
right now it's like a thousand companies
37:01
are trying to sell you your AEO
37:03
services. So we're probably going to
37:04
have a few companies that are going to
37:05
be the ones who are like, "Yeah, okay.
37:07
We know what to do here." Um it's
37:09
difficult, but we've kind of cracked
37:10
that nut. So that's one thing. Uh I
37:13
think that uh we'll see paid advertising
37:16
on AI media become something where, you
37:20
know, in the US maybe a couple billion
37:23
dollars is is being spent. I mean, I've
37:24
seen forecast of like 30 billion by
37:27
2030.
37:28
>> Um, it may not ramp up that as quickly
37:30
as that because of a lot of the
37:32
underlying factors that we've talked
37:34
about. Um, but it's definitely growing
37:36
and it has a long ramp ahead of it. If
37:38
you look at the growth of social media,
37:40
the first several years of social media
37:42
was basically very very small until
37:44
companies like Facebook really figured
37:46
out how to create an ad product that
37:49
made advertisers go, "Oh, okay. Now I
37:52
really need that." um and they took away
37:54
organic reach. But that's that's the
37:57
kind of thing that I think is going to
37:58
be happening. Um finally, I think the
38:00
thing I'll say is uh you know, we don't
38:02
know necessarily, we spent a lot of time
38:04
thinking about and talking about chat
38:05
GBT. We don't know if Chat GBT is going
38:07
to be the winner in this space. Um I
38:09
don't think we can say right now 18
38:11
months from now that chat GBT is going
38:13
to be the leader. Uh social media, we
38:15
all talked about MySpace back in the day
38:17
and then Facebook came along and who who
38:19
the heck was MySpace? So I I don't
38:22
necessarily think that that that um we
38:24
can rule out like Google. We've hardly
38:26
mentioned the word the name Google in
38:28
this entire conversation. Uh which we
38:30
should have, but we can't rule out
38:32
Google. Google has the the
38:33
infrastructure, the the the knowledge,
38:36
the advertiser relationships, um the
38:38
technology and all of that to I think
38:41
stand out. So if I were to guess, it's
38:44
going to be Google and Chat GPT kind of
38:46
vying for the top spot when it comes to
38:49
AI media. Interesting. Okay. Well, let's
38:53
then set a date. We can always move it
38:56
up or back where we revisit this, but
38:58
I'm sure in the meantime, in between now
39:00
and then, you will continue to provide
39:02
us with near real time, if not real time
39:04
updates as to the leaders, fast
39:07
followers, um, and all the changes that
39:10
keep taking place at a staggering rate.
39:13
So, Deborah, thank you so much for being
39:15
here. So, I want people to uh follow or
39:18
or subscribe to the AI ad economy,
39:20
right? That's the newsletter. I love it.
39:22
Was about weekly, once a week or every
39:24
other week. Weekly on Substack on also
39:27
available on LinkedIn.
39:28
>> Yeah. Okay. Substack and LinkedIn. We
39:30
know we know them well. Um so, thank you
39:32
for for this and for helping us make
39:34
another genuine connection.