1
00:00:32,055 --> 00:00:37,592
Hello, hello, hello, welcome to another
episode of the Always Be Testing Podcast.

2
00:00:37,592 --> 00:00:41,847
I'm your host, Ty DeGrange, and I'm here
in beautiful Santa Barbara at CJU.

3
00:00:41,847 --> 00:00:44,901
We're talking to Catherine Tomargo right
now,

4
00:00:44,901 --> 00:00:47,325
the AI Zar. How you doing?

5
00:00:47,846 --> 00:00:48,846
That's

6
00:00:48,146 --> 00:00:50,552
a new one. I love it. Doing great.

7
00:00:50,552 --> 00:00:51,552
Can't complain really.

8
00:00:52,109 --> 00:00:56,924
Just got out of a fun agency CJ
conversation,

9
00:00:56,924 --> 00:01:01,333
diving deep into all the things and CJU is
near and dear to my heart because

10
00:01:01,333 --> 00:01:05,913
I was a CJU employee and it's great to be
back in Santa Barbara.

11
00:01:05,913 --> 00:01:07,927
We've got gauchos on our on our team.

12
00:01:07,927 --> 00:01:10,387
I'm a former gaucho, so it's good to be
back.

13
00:01:10,387 --> 00:01:11,387
And thanks

14
00:01:11,128 --> 00:01:12,128
for joining us.

15
00:01:11,925 --> 00:01:13,405
Yeah, thanks for having me. Yeah.

16
00:01:14,009 --> 00:01:15,009
know a little

17
00:01:14,419 --> 00:01:18,162
bit about what's happening all this AI
stuff so hopefully you can help

18
00:01:18,162 --> 00:01:22,494
us and you also know a lot about CJ
because you've kind of been with CJ a
while.

19
00:01:22,494 --> 00:01:23,655
How did you

20
00:01:22,657 --> 00:01:23,657
Yeah.

21
00:01:23,675 --> 00:01:25,737
how did you how'd you do that?

22
00:01:26,298 --> 00:01:29,440
How did I do that? It's a great question.

23
00:01:29,440 --> 00:01:36,007
I mean for many I'd say about

24
00:01:36,007 --> 00:01:40,090
a decade now, we've been talking about how
affiliate is the channel of channels,

25
00:01:40,090 --> 00:01:46,296
how it we can pretty much do almost
everything on a performance basis.

26
00:01:46,296 --> 00:01:50,650
And I think I just got really lucky early
on with clients that were,

27
00:01:50,650 --> 00:01:51,650
you know

28
00:01:52,185 --> 00:01:58,553
~ so I was able to to test into

29
00:01:58,553 --> 00:02:01,978
a lot of new emerging models and that
eventually just sort of,

30
00:02:01,978 --> 00:02:07,165
you know, set me up for for being at the
forefront of what's happening now within

31
00:02:08,887 --> 00:02:10,548
Yeah, it's come up so much on the pod,

32
00:02:10,548 --> 00:02:13,871
it's obviously the namesake, but you you
lived it and made sure that

33
00:02:13,871 --> 00:02:15,363
it was happening for your clients.

34
00:02:15,363 --> 00:02:18,901
It's something that Tom and other
colleagues and I have talked about,

35
00:02:18,901 --> 00:02:22,418
which is beyond important. Coming up
through CJ,

36
00:02:22,418 --> 00:02:25,991
was there you know was there something
that sort of stood out for you

37
00:02:26,072 --> 00:02:32,037
or or taught you with those experiments
that kind of like how did you get to be
like

38
00:02:34,773 --> 00:02:37,395
Obviously it's become massively important,

39
00:02:37,395 --> 00:02:43,379
but I'm excited to hear how you went from
CJ Client Services to leading

40
00:02:43,479 --> 00:02:45,400
the AI charge procedure, which is really

41
00:02:45,400 --> 00:02:46,400
really amazing.

42
00:02:46,289 --> 00:02:50,653
Yeah, so I mean again it it comes back to
this test and learn framework.

43
00:02:50,653 --> 00:02:55,556
That's something I really encouraged a lot
of my clients to ensure they were always

44
00:02:55,556 --> 00:02:59,349
doing and with testing and learning
they're you know,

45
00:02:59,349 --> 00:03:04,984
you gotta look at the data. ~ so just
having that

46
00:03:07,097 --> 00:03:12,301
Data focus or data focus just persists
throughout everything that that I do.

47
00:03:12,301 --> 00:03:15,750
Earlier, I would say like late twenties
four,

48
00:03:15,750 --> 00:03:16,750
early twenty.

49
00:03:20,505 --> 00:03:27,322
that I helped develop from CJ around
understanding where referral traffic

50
00:03:27,322 --> 00:03:30,975
was coming. ~ where for both, you know,

51
00:03:30,975 --> 00:03:33,827
publishers and where advertisers were
referring traffic,

52
00:03:33,827 --> 00:03:37,641
et cetera, et cetera, ~ was able to pick
up on the fact that

53
00:03:38,603 --> 00:03:40,855
early days, maybe not so much right now,

54
00:03:40,855 --> 00:03:45,509
but early days, ~ some of these emerging
platforms were actually driving lots

55
00:03:45,509 --> 00:03:46,900
of traffic to publisher sites.

56
00:03:46,900 --> 00:03:52,984
And so quickly understood that as both an
opportunity ~ to

57
00:03:54,462 --> 00:03:56,263
improve that for publishers, right?

58
00:03:56,263 --> 00:03:59,095
So if if you're not yet receiving kind of
traffic,

59
00:03:59,095 --> 00:04:01,987
how can you do that? And because there
wasn't a ton out there,

60
00:04:01,987 --> 00:04:06,439
it's really required just for the digging
in and we are.

61
00:04:06,439 --> 00:04:11,012
Yeah, absolutely. And when clients are
coming to you and coming to CJ to say,

62
00:04:11,012 --> 00:04:13,774
I need help figuring out AI visibility,

63
00:04:13,774 --> 00:04:17,236
what are some of the things that they're
getting and how what does that

64
00:04:17,236 --> 00:04:18,236
look like?

65
00:04:18,173 --> 00:04:21,917
Sure. So we back to data,

66
00:04:21,917 --> 00:04:28,573
right? ~ we've spent spent a lot of last
year and beginning of this year just

67
00:04:28,573 --> 00:04:32,456
trying to lay the foundation,

68
00:04:32,456 --> 00:04:35,169
right? Or or understand and unpack what it
is that we

69
00:04:41,421 --> 00:04:47,793
the tech stack should be ~ since the
existing tech stack doesn't quite tackle

70
00:04:47,793 --> 00:04:53,866
a lot of the pieces and so ~ with that
research we've developed this sort

71
00:04:53,866 --> 00:05:00,198
of we'll call it our own tech stack or or
suite of tools ~ that allow brands

72
00:05:00,198 --> 00:05:04,400
to understand that starting with
visibility right so like what visibility

73
00:05:07,011 --> 00:05:10,553
far as we're concerned and I think most of
the industry is whether or

74
00:05:10,553 --> 00:05:15,145
not a brand appears across these
environments and and how they appear.

75
00:05:15,145 --> 00:05:17,667
So there are tools in place that allow
brands to track that.

76
00:05:17,667 --> 00:05:21,229
That's sort of where we start ~ is
establishing those baselines

77
00:05:21,229 --> 00:05:24,710
or conducting an audit in order to
establish those baselines.

78
00:05:24,710 --> 00:05:30,964
And then from there looking at the sources
that are driving that visibility

79
00:05:30,964 --> 00:05:34,566
on behalf of that brand to that then you
know we can then devise a plan.

80
00:05:34,566 --> 00:05:36,617
But it's all grounded in data.

81
00:05:36,601 --> 00:05:37,601
I

82
00:05:36,681 --> 00:05:40,864
love that. And what you shared earlier
today was really cool and like citation,

83
00:05:40,864 --> 00:05:46,879
sentiment, mention, like you talked a
little bit about like what that maybe
looks

84
00:05:46,879 --> 00:05:48,700
like without giving up the whole story.

85
00:05:48,700 --> 00:05:52,313
It's a really I loved what some of that
visibility you shared.

86
00:05:52,405 --> 00:05:58,528
Yeah, sure. So with this this new touch
point we'll call it,

87
00:05:58,528 --> 00:06:04,851
~ because it hasn't yet received the full
~ channel crown just yet or designation.

88
00:06:04,851 --> 00:06:07,052
but was just information

89
00:06:06,145 --> 00:06:07,145
Okay.

90
00:06:07,052 --> 00:06:10,853
here on the pod. Not quite. I just think
the industry hasn't fully aligned.

91
00:06:10,853 --> 00:06:15,155
Like you, GEO, are a channel now.

92
00:06:15,155 --> 00:06:18,717
but the

93
00:06:19,425 --> 00:06:21,870
It is new and it is different.

94
00:06:21,870 --> 00:06:24,355
~ it is similar to

95
00:06:26,177 --> 00:06:27,597
Other things that we've done in the past,

96
00:06:27,597 --> 00:06:29,248
but it is nonetheless different.

97
00:06:29,248 --> 00:06:33,549
And where it sits is really like in this
upper funnel space.

98
00:06:33,549 --> 00:06:36,920
Yeah. And as such the the KPIs are
different.

99
00:06:36,920 --> 00:06:38,550
Right. So we talked about visibility.

100
00:06:38,550 --> 00:06:40,381
Yeah. That's one of the main KPIs.

101
00:06:40,381 --> 00:06:42,911
How is your brand appearing? That's what
that is.

102
00:06:42,911 --> 00:06:44,082
On a per kind

103
00:06:43,657 --> 00:06:44,657
Problem.

104
00:06:44,082 --> 00:06:46,492
of basis. Yeah. A prompt, a platform,

105
00:06:46,492 --> 00:06:49,583
etcetera. ~ so that's a visibility score,

106
00:06:49,583 --> 00:06:53,284
which is also known as pension rate.

107
00:06:53,284 --> 00:06:55,085
Those two are somewhat interchangeable.

108
00:06:55,085 --> 00:06:56,085
~

109
00:06:56,363 --> 00:06:59,486
And then there is ~ sentiment,

110
00:06:59,486 --> 00:07:00,687
right? So like you're appearing,

111
00:07:00,687 --> 00:07:02,249
but is it positive or negative?

112
00:07:02,249 --> 00:07:06,103
What what's the deal? ~ and then position,

113
00:07:06,103 --> 00:07:08,225
right? So you're being listed,

114
00:07:08,225 --> 00:07:10,707
that's fantastic. Are you first or are you
last?

115
00:07:10,707 --> 00:07:16,292
and those are the sort of four KPIs that
that we look at when we think about just

116
00:07:17,801 --> 00:07:22,542
Yeah, that's amazing. A lot of talk
happens around path to purchase,

117
00:07:22,542 --> 00:07:25,183
attribution, MTA, all that stuff.

118
00:07:25,183 --> 00:07:28,584
There's like kind of the brand led
conversations,

119
00:07:28,584 --> 00:07:35,526
AI led, social. You know, where where does
C A CJ see it when with AI growing?

120
00:07:35,526 --> 00:07:40,708
What's like your is there pieces that you
think are kind of more overhyped?

121
00:07:40,708 --> 00:07:44,185
And what where do you kind of stand on
that those paths

122
00:07:44,185 --> 00:07:45,185
to purchase?

123
00:07:45,305 --> 00:07:50,129
Yeah, I mean look data exists that shows

124
00:07:50,189 --> 00:07:56,585
us that AI referred traffic is converting
at higher rates than

125
00:07:56,585 --> 00:08:02,189
non-AI referred traffic. So that's not
that's not made up or overhyped,

126
00:08:02,189 --> 00:08:08,284
that's that's the truth. The the rate at
which it's higher does vary depending

127
00:08:08,284 --> 00:08:09,996
on the time frame or platform,

128
00:08:09,996 --> 00:08:11,907
etc. But that's true.

129
00:08:14,393 --> 00:08:15,393
Tack that

130
00:08:18,201 --> 00:08:23,596
call them answers, right? We used to rely
on search engines to conduct

131
00:08:23,596 --> 00:08:24,820
our own research and

132
00:08:27,945 --> 00:08:28,945
Mm.

133
00:08:30,361 --> 00:08:33,274
~ that's another thing that that's that's
not overhyped,

134
00:08:33,274 --> 00:08:35,036
that's that's just the truth, right?

135
00:08:35,036 --> 00:08:37,849
Even if you're not an AI adopter,

136
00:08:37,849 --> 00:08:40,813
you use Google and Google has AI
overviews.

137
00:08:40,813 --> 00:08:44,767
So ~ it's pretty much ~ reached everyone.

138
00:08:44,767 --> 00:08:51,644
So some of the things that might be
overhyped or around some of the

139
00:08:52,333 --> 00:08:55,455
guess measurement that can be delivered at
this point in time.

140
00:08:55,455 --> 00:08:59,709
But as far as whether or not users are
using it,

141
00:08:59,709 --> 00:09:02,121
if that's not overhyped. Absolutely.

142
00:09:02,121 --> 00:09:04,938
as far as whether or not brands should be
testing,

143
00:09:04,938 --> 00:09:08,076
I I don't think that that's overhyped
because that's happening now

144
00:09:08,076 --> 00:09:13,510
and it's better to be in a position of
being proactive versus reactive.

145
00:09:13,510 --> 00:09:16,433
And there's a lot to learn right now.

146
00:09:16,433 --> 00:09:20,776
~ a lot of the conversations I have with
with brands are around

147
00:09:22,121 --> 00:09:23,843
though can't quite forecast just yet,

148
00:09:23,843 --> 00:09:26,908
even though we can't quite project the
returns,

149
00:09:26,908 --> 00:09:30,623
whether it be, you know, anything further
down the funnel.

150
00:09:32,953 --> 00:09:36,114
Sales or sales revenue. A learning is a
win.

151
00:09:36,114 --> 00:09:38,315
and that's something that you would want
to know.

152
00:09:38,315 --> 00:09:44,137
Like how how does being visible translate
to those downstream ~ APIs?

153
00:09:44,137 --> 00:09:50,099
So I'm not I think that the framing of
some things are overhyped,

154
00:09:50,099 --> 00:09:55,121
but like the fact that you should be doing
it and that it is important is not.

155
00:09:55,467 --> 00:09:56,467
Yeah,

156
00:09:55,727 --> 00:09:57,108
and that's a perfect segue into,

157
00:09:57,108 --> 00:09:58,719
you know, namesake of the pod,

158
00:09:58,719 --> 00:10:01,280
the testing importance. We've clearly
established that like,

159
00:10:01,280 --> 00:10:03,842
hey, you kinda have to be doing these
types of tests.

160
00:10:03,842 --> 00:10:07,964
What's a way you like to see it structured
in a GEO,

161
00:10:07,964 --> 00:10:13,268
answer engine environment? How would you
like to see a test structure appropriately

162
00:10:13,268 --> 00:10:14,588
and set up for success?

163
00:10:15,549 --> 00:10:16,549
So

164
00:10:16,289 --> 00:10:20,281
~ it all goes back to the data.

165
00:10:20,281 --> 00:10:26,394
Yep. ~ we the data's not absolutely
perfect and what I mean by that is

166
00:10:26,394 --> 00:10:32,517
in an ideal world we're getting exact
query or or prompt information from

167
00:10:32,517 --> 00:10:37,519
the PLMs themselves alongside the exact
responses and the sources cited.

168
00:10:37,519 --> 00:10:41,791
In the absence of that we have synthetic
data that we can look at,

169
00:10:41,791 --> 00:10:45,123
we have CDN call signals that we can
incorporate.

170
00:10:45,323 --> 00:10:50,934
a number of things that we can we can look
at to then help help inform what that

171
00:10:50,934 --> 00:10:51,934
test should look like.

172
00:10:51,681 --> 00:10:52,681
Can

173
00:10:51,741 --> 00:10:54,430
you say some credit data, can you share
what that means exactly?

174
00:10:54,369 --> 00:10:55,369
Sure,

175
00:10:54,869 --> 00:10:59,753
so ~ in the absence of receiving the
actual prompt information from platforms,

176
00:10:59,753 --> 00:11:05,566
there have been a number of emerging tools
that have emerged over

177
00:11:05,566 --> 00:11:09,079
the past couple of months, eighteen months
or so,

178
00:11:09,079 --> 00:11:13,322
that run synthetic prompts across those
top LLMs,

179
00:11:13,322 --> 00:11:18,355
capture those responses, and that's what
they then use in order to put together

180
00:11:19,110 --> 00:11:24,324
scores tell you which sources are cited
and how frequently across which topics,

181
00:11:24,324 --> 00:11:25,324
etc etc.

182
00:11:25,939 --> 00:11:26,939
Amazing,

183
00:11:26,590 --> 00:11:31,645
amazing. Are there publisher types that
are starting to really nail it from

184
00:11:31,645 --> 00:11:37,230
a GEO citation perspective? What are are
there are there types of reviews

185
00:11:37,230 --> 00:11:41,244
or types of formats that are really
winning in this Answer Engine environment?

186
00:11:41,673 --> 00:11:48,269
Sure, I mean honestly it's the ones that
type closest to the types

187
00:11:48,269 --> 00:11:53,733
of prompts that consumers ask that are
very structured in a very clear

188
00:11:53,733 --> 00:12:00,339
and succinct way. ~ things like the best X
of you know for whatever

189
00:12:02,873 --> 00:12:06,336
gifts to buy mom, dad, brother,

190
00:12:06,336 --> 00:12:12,241
son, etc etc. ~ these are the kinds of
this is the kind of language that's easily

191
00:12:12,241 --> 00:12:13,241
digested.

192
00:12:19,459 --> 00:12:21,913
Yeah. As far as types of publishers,

193
00:12:21,913 --> 00:12:23,140
it really comes down

194
00:12:29,209 --> 00:12:30,209
industry

195
00:12:33,081 --> 00:12:38,669
Large percentage of affiliate publishers
were being cited or were being cited.

196
00:12:38,669 --> 00:12:43,485
Each individual publisher has taken a
different approach in terms of

197
00:12:43,485 --> 00:12:48,732
how they want to perceive those citations.

198
00:12:48,732 --> 00:12:50,525
Some really appreciate that and I

199
00:12:53,187 --> 00:12:57,739
Optimizing their schema and their
structures and the type of content that
they focus

200
00:12:57,739 --> 00:12:59,263
on. Yeah. And others are

201
00:13:02,213 --> 00:13:04,495
So it's not so much about the type of
publisher,

202
00:13:04,495 --> 00:13:06,926
although I'll tell you can't you know,

203
00:13:06,926 --> 00:13:09,337
publishers that write things on the
internet,

204
00:13:09,337 --> 00:13:12,529
which is everyone, ~ are the types that
could be cited.

205
00:13:12,529 --> 00:13:15,341
It's those that are are really leaning
into it.

206
00:13:15,405 --> 00:13:19,088
I love it. I love it. And we talk a lot
about zero click.

207
00:13:19,088 --> 00:13:21,550
We talk a lot about last click being
flawed.

208
00:13:21,550 --> 00:13:25,233
We talk a lot about the need for more
visibility and exposure.

209
00:13:25,233 --> 00:13:27,765
And we're working on that ourselves.

210
00:13:27,765 --> 00:13:31,538
I mean we we have methodologies to
showcase clients visibility.

211
00:13:31,538 --> 00:13:34,753
We ha we talk and often talk about how
it's a hey,

212
00:13:34,753 --> 00:13:37,554
it's a gift with purchase. You're getting
core partner marketing,

213
00:13:37,554 --> 00:13:42,448
but a massive percentage of this is
showing up in AI citations and GEO.

214
00:13:44,247 --> 00:13:47,578
I've I've talked about this a little bit
and I've written some things about payment

215
00:13:47,578 --> 00:13:51,000
and how do you tie payment and
compensation to this world?

216
00:13:51,000 --> 00:13:54,361
How does CJ I mean how does CJ think about
it?

217
00:13:54,361 --> 00:13:56,222
How do you personally think about it?

218
00:13:56,222 --> 00:14:00,332
~ the compensation models, how do they
evolve in more of

219
00:14:00,332 --> 00:14:02,844
a visibility environment when the click's
being lost?

220
00:14:03,801 --> 00:14:04,801
Sure, so

221
00:14:04,451 --> 00:14:07,993
I mean my personal POV is they absolutely
need to evolve,

222
00:14:07,993 --> 00:14:11,715
right? as search,

223
00:14:11,715 --> 00:14:15,726
I'll just say search, refer traffic
declines for publishers,

224
00:14:15,726 --> 00:14:20,955
they they need different ways to monetize
their influence,

225
00:14:20,955 --> 00:14:26,916
right? ~ and so there are things that we
can use in

226
00:14:26,916 --> 00:14:29,042
the absence of those things that we need,

227
00:14:29,042 --> 00:14:30,541
right? So we talked about those things.

228
00:14:31,553 --> 00:14:37,887
talked about ~ CDN call signals there
there are a lot of things that we can pull
in.

229
00:14:37,887 --> 00:14:39,871
our univers well

230
00:14:44,387 --> 00:14:48,899
This we might need to cut.

231
00:14:48,899 --> 00:14:55,003
CD and call signals, other technologies
you can share

232
00:14:55,003 --> 00:15:00,326
all digital data so that we can analyze
trends,

233
00:15:00,326 --> 00:15:05,849
you know, direct conversion trends and
apply ~ or create methodologies that then

234
00:15:05,849 --> 00:15:11,772
allow us to to make assumptions around
publisher influence ~ to conversions.

235
00:15:11,772 --> 00:15:13,595
And that's something that absolutely needs

236
00:15:14,881 --> 00:15:19,206
and we need to begin testing. I think in
the intern we're focusing

237
00:15:19,206 --> 00:15:22,770
on flat feed engagements as we develop
those,

238
00:15:22,770 --> 00:15:28,739
paper influence or just ~ dynamic
commissioning structures based on that
engagement.

239
00:15:33,773 --> 00:15:37,537
No, it's good. I mean I think that what
I'm hearing is you're thinking about

240
00:15:37,537 --> 00:15:40,841
a lot of ways to compensate. You've got a
number of levers,

241
00:15:40,841 --> 00:15:43,844
you know, kind of within the existing
ecosystem to compensate.

242
00:15:43,844 --> 00:15:48,229
And then there's all these new ways to
measure that we can co tie

243
00:15:48,229 --> 00:15:53,485
to those payment methods that allow you to
have some flexibility around it.

244
00:15:53,485 --> 00:15:55,133
So I think it makes sense too.

245
00:15:55,261 --> 00:16:01,984
But it is an industry priority ~ that we
all need to start thinking about.

246
00:16:01,984 --> 00:16:05,996
~ particularly I would say on the pub
publisher front,

247
00:16:05,996 --> 00:16:09,768
right? ~ we I work really closely with
with publishers,

248
00:16:09,768 --> 00:16:13,690
wanting to hear from them, wanting to hear
that feedback to better understand what

249
00:16:13,690 --> 00:16:16,251
it is that they need. Yeah. Because at the
end of the day,

250
00:16:16,251 --> 00:16:18,142
they're the ones providing the content,

251
00:16:18,142 --> 00:16:20,333
they're the ones driving that influence.

252
00:16:20,475 --> 00:16:21,475
Yeah.

253
00:16:20,889 --> 00:16:25,306
~ and so we wanna make sure that whatever
whatever is developed is

254
00:16:27,093 --> 00:16:30,606
Yeah, absolutely. I mean we often talk
about it like they're the

255
00:16:30,606 --> 00:16:34,038
RD of the ecosystem, they're the currency
of the ecosystem,

256
00:16:34,038 --> 00:16:38,142
they're they're really who is driving a
lot of it in so many ways.

257
00:16:38,142 --> 00:16:40,704
So I love that framing. That's great.

258
00:16:40,704 --> 00:16:44,747
It's mythbuster time. We ready to bust
some myths?

259
00:16:44,747 --> 00:16:45,747
Yeah,

260
00:16:45,307 --> 00:16:47,749
they're what's some

261
00:16:45,505 --> 00:16:46,505
Let's do it.

262
00:16:47,749 --> 00:16:52,733
good what's some good like Geo myths that
we want to just kind of say ~

263
00:16:53,057 --> 00:16:55,041
Not really not really your jam.

264
00:16:55,041 --> 00:16:56,884
meh.

265
00:17:02,445 --> 00:17:08,771
Taking performance budgets and defining
and leveraging those for upper funnel.

266
00:17:08,771 --> 00:17:11,694
If you do that, you're gonna kill
performance.

267
00:17:11,694 --> 00:17:15,217
Like if you remove performance budgets,

268
00:17:15,217 --> 00:17:18,020
then your pr you won't meet your
performance goals.

269
00:17:18,020 --> 00:17:21,684
those th that maybe it's not a myth,

270
00:17:21,684 --> 00:17:22,745
but we'll call it

271
00:17:24,458 --> 00:17:30,596
Is it a mistake or trap that you see
brands saying we're going

272
00:17:30,596 --> 00:17:32,548
to take away

273
00:17:33,356 --> 00:17:34,356
Let's

274
00:17:33,676 --> 00:17:37,309
reallocate performance budgets for the
purposes of GEO,

275
00:17:37,309 --> 00:17:39,456
which then impacts the performance.

276
00:17:39,692 --> 00:17:43,333
Yeah, yeah, you need you need the fuel of
what's working at the very least.

277
00:17:43,333 --> 00:17:47,174
And that should that should be coming from
another experimentation budget or yeah,

278
00:17:47,174 --> 00:17:51,576
you're gonna lose something. One of many
silly claims I

279
00:17:48,821 --> 00:17:52,415
Yeah. The the

280
00:17:56,225 --> 00:17:57,225
Yeah.

281
00:17:56,866 --> 00:18:01,963
Another big myth. GEO is primarily owned.

282
00:18:04,958 --> 00:18:08,524
Or rent it. Yeah, yeah. Yeah.

283
00:18:08,524 --> 00:18:13,393
~ the data shows that the vast majority of
sources cited are third party sources.

284
00:18:13,393 --> 00:18:14,393
Go through

285
00:18:14,179 --> 00:18:15,773
Yeah. If you want GEO

286
00:18:15,773 --> 00:18:19,643
results, you heard it here. Kind of
important.

287
00:18:19,757 --> 00:18:20,757
Kind little bit.

288
00:18:20,738 --> 00:18:24,280
~ big big myth, there are clients we talk
to that I'm like,

289
00:18:24,280 --> 00:18:27,526
all right, let me know. ~ would love to
hear more about what you're doing.

290
00:18:27,526 --> 00:18:29,829
They're like, we got this down.

291
00:18:29,829 --> 00:18:34,976
We are re like ~ rewriting our entire
website.

292
00:18:34,976 --> 00:18:36,637
We are writing so much content.

293
00:18:36,637 --> 00:18:38,720
We are doing all of these things.

294
00:18:41,849 --> 00:18:43,421
I mean you kinda need both, right?

295
00:18:43,421 --> 00:18:48,480
But you don't wanna neglect definitely
don't wanna neglect th third party trusted

296
00:18:48,480 --> 00:18:52,807
content that talks about you in a way that
doesn't feel as biased

297
00:18:52,807 --> 00:18:53,807
as you talking about.

298
00:18:53,999 --> 00:18:56,259
Yeah, but it's not even about the bias,

299
00:18:56,259 --> 00:18:57,293
really, it's about

300
00:18:59,623 --> 00:19:01,037
Let's look at the

301
00:19:01,593 --> 00:19:03,974
Like it just is let's go back to the data.

302
00:19:03,974 --> 00:19:09,176
And that's really what ~ I encourage every
company to do.

303
00:19:09,176 --> 00:19:14,698
let's start with the data and who is being
cited when it comes to your brand.

304
00:19:14,698 --> 00:19:17,579
That's that's then the work that needs to
be done.

305
00:19:17,579 --> 00:19:20,720
And I'm telling you, ~ their analys I

306
00:19:22,393 --> 00:19:24,935
There are analysis that I have read about
on the internet,

307
00:19:24,935 --> 00:19:28,797
~ including analysis that CJ has done and
published,

308
00:19:28,797 --> 00:19:33,721
that shows that over 83% of sources cited
across verticals are

309
00:19:34,341 --> 00:19:38,747
Yeah, that percentage sounds like it keeps
going up.

310
00:19:38,747 --> 00:19:44,494
I kinda like that. good myth busting.

311
00:19:44,494 --> 00:19:49,020
The czar of AI, bespoke. Agentic.

312
00:19:49,020 --> 00:19:52,004
What the heck is it and what's going on
and what's coming?

313
00:19:52,505 --> 00:19:53,505
A

314
00:19:52,745 --> 00:19:56,266
gente. So you know,

315
00:19:56,266 --> 00:20:01,067
I sometimes I worry about hot takes in

316
00:20:01,067 --> 00:20:04,728
AI because they're always like famous last
words.

317
00:20:04,728 --> 00:20:09,039
Yeah. a gen tick in particular,

318
00:20:09,039 --> 00:20:13,431
that one feels it feels a ways away.

319
00:20:13,431 --> 00:20:18,412
Yeah. I sure I'll tell you something
anecdotally.

320
00:20:18,412 --> 00:20:19,952
Yep. Last year.

321
00:20:22,381 --> 00:20:25,524
Last year my friend and I were were
talking about back to school.

322
00:20:25,524 --> 00:20:29,407
She was doing back to school shopping for
her daughters and we built some agents.

323
00:20:29,407 --> 00:20:34,932
We're just like, ~ my god, AI is just
going to simplify all of our lives.

324
00:20:34,932 --> 00:20:38,075
Let's get some agents to get that done and
boom.

325
00:20:40,312 --> 00:20:41,312
About the

326
00:20:40,835 --> 00:20:41,835
wrong stuff.

327
00:20:42,295 --> 00:20:43,295
I listened

328
00:20:42,845 --> 00:20:48,557
to a keynote today ~ that talked about how
either half or over half something like

329
00:20:48,557 --> 00:20:52,018
that of the internet is ~ it's actually
bots.

330
00:20:52,018 --> 00:20:56,929
It's bot traffic. So you know what a lot
of websites have done?

331
00:20:56,929 --> 00:21:01,291
They've implemented ways to block bot
traffic.

332
00:21:01,291 --> 00:21:04,792
And so work agents. Until thank you.

333
00:21:04,792 --> 00:21:07,213
~ so until we really like

334
00:21:08,515 --> 00:21:13,249
solve for that that piece and of course
there's like authentication and and

335
00:21:13,249 --> 00:21:15,750
a lot of but there's a lot of work to be
done

336
00:21:15,897 --> 00:21:17,240
Turns out we're still employed.

337
00:21:16,965 --> 00:21:20,654
it what a

338
00:21:20,914 --> 00:21:25,798
concept but even beyond that I mean
there's there are two different maybe
three

339
00:21:25,798 --> 00:21:31,903
but the two main ~ challenges that we need
to solve for is one consumer trust right

340
00:21:31,903 --> 00:21:35,005
like who I I don't really know that many
people that are handing over

341
00:21:36,257 --> 00:21:40,232
are ready anyway to hand over their credit
card to agents and have them

342
00:21:40,232 --> 00:21:45,909
~ fully make those decisions. And then
there's the the website or internet trust.

343
00:21:45,909 --> 00:21:50,554
The websites don't trust the agents to
come in and and make purchase decisions.

344
00:21:50,554 --> 00:21:55,400
So until those two things are solved for
I'm just I th I think it's

345
00:21:55,481 --> 00:21:56,481
I think it's always

346
00:21:56,082 --> 00:21:58,985
right. Yep, Katarine, head of AI at CJ.

347
00:21:58,985 --> 00:22:02,049
Miss 305, some people say. It's

348
00:22:02,097 --> 00:22:03,097
That one says.

349
00:22:03,030 --> 00:22:04,116
a it's a it's

350
00:22:04,116 --> 00:22:07,435
it's a it's a starting to get traction
here in Santa Barbara.

351
00:22:07,435 --> 00:22:13,001
What's one big takeaway you want to give
the audience in the AI realm that

352
00:22:13,262 --> 00:22:15,821
you recommend they they take action on?

353
00:22:16,055 --> 00:22:17,055
Yeah.

354
00:22:17,589 --> 00:22:18,589
I would say

355
00:22:23,225 --> 00:22:26,586
start collaborating cross channel,

356
00:22:26,586 --> 00:22:30,088
cross cross almost everything really.

357
00:22:30,088 --> 00:22:36,090
~ as we dig into as we dig into

358
00:22:36,090 --> 00:22:39,608
how these LLMs work and and how they
formulate answers,

359
00:22:39,608 --> 00:22:42,493
we're really finding that every single
channel plays role.

360
00:22:42,493 --> 00:22:47,775
And when you are having one channel focus
on all of it,

361
00:22:47,775 --> 00:22:50,697
you're really missing a lot of context and
a lot of

362
00:22:52,137 --> 00:22:55,089
the opportunity. So bring it all together,

363
00:22:55,089 --> 00:22:57,001
create a center of excellence,

364
00:22:57,001 --> 00:22:59,302
create a tiger team, create a strike team,

365
00:22:59,302 --> 00:23:02,865
whatever you want to call it. bring in all
the right stakeholders

366
00:23:03,125 --> 00:23:07,228
and focus on the data.

367
00:23:07,228 --> 00:23:14,013
For the love of God, please look at the
data and make decisions based on on what
it

368
00:23:14,617 --> 00:23:17,659
I love it. Sounds obvious, but it's
clearly helped you navigate all all

369
00:23:17,659 --> 00:23:21,072
the way up through your career with one
great company and you've been able

370
00:23:21,072 --> 00:23:24,795
to lead it now with AI and it's been a
pleasure to get to know you here

371
00:23:24,795 --> 00:23:29,289
and thanks for joining us in a pretty nice
location to to figure it all out.

372
00:23:29,289 --> 00:23:33,052
So until next time.

373
00:23:29,837 --> 00:23:31,373
Thank you so much for having me.

374
00:23:31,373 --> 00:23:32,373
Yeah.

375
00:23:33,077 --> 00:23:35,757
Thanks everybody. I