1
00:00:00,280 --> 00:00:04,240
Speaker 1: More regulatory headaches for Google after a judge yesterday ordered

2
00:00:04,280 --> 00:00:07,920
Speaker 1: Alphabet subsidiary to open up its Android app store as

3
00:00:07,960 --> 00:00:10,799
Speaker 1: part of the Epic Game's anti trust suit, and additional

4
00:00:10,800 --> 00:00:14,080
Speaker 1: headwinds could be ahead today in the DOJ's search trial.

5
00:00:14,520 --> 00:00:18,200
Speaker 2: It is the king of Internet searches, the heavyweight of

6
00:00:18,280 --> 00:00:22,799
Speaker 2: digital advertising, but Google is in trouble literally, as it

7
00:00:22,880 --> 00:00:26,479
Speaker 2: faces multiple anti trust cases in its home market in

8
00:00:26,520 --> 00:00:27,200
Speaker 2: the US.

9
00:00:27,480 --> 00:00:31,080
Speaker 3: With two anti trust case decisions already going against Google

10
00:00:31,120 --> 00:00:34,720
Speaker 3: and a third on its digital advertising business pending a

11
00:00:34,880 --> 00:00:37,839
Speaker 3: major shakeup off, the company seems to be on the cards.

12
00:00:38,240 --> 00:00:40,520
Speaker 2: This week on the Business at Tech Powered by Two

13
00:00:40,560 --> 00:00:44,440
Speaker 2: Degrees Business, we unpack the Google anti trust cases and

14
00:00:44,520 --> 00:00:48,960
Speaker 2: their implications for Google and for advertisers in this part

15
00:00:49,040 --> 00:00:49,720
Speaker 2: of the world.

16
00:00:49,960 --> 00:00:53,320
Speaker 4: This is not about whether Google worked. The reality was

17
00:00:53,320 --> 00:00:55,560
Speaker 4: the products were good, but it is what allowed them

18
00:00:55,640 --> 00:00:58,040
Speaker 4: to make their products good and what allowed them to

19
00:00:58,080 --> 00:01:01,360
Speaker 4: make their products untouchable by editors is the issue here,

20
00:01:01,400 --> 00:01:04,320
Speaker 4: and what would happen if Google didn't control the ad

21
00:01:04,319 --> 00:01:05,920
Speaker 4: supply system like it does now.

22
00:01:06,720 --> 00:01:10,320
Speaker 3: We've got veteran journalist Paul McIntyre joining us. He's been

23
00:01:10,360 --> 00:01:13,800
Speaker 3: covering the advertising and media industries in Australia for decades.

24
00:01:14,120 --> 00:01:16,640
Speaker 3: He's going to talk to us about Google. He's just

25
00:01:16,760 --> 00:01:21,160
Speaker 3: been in Virginia covering one of those antitrust cases. But

26
00:01:21,360 --> 00:01:23,920
Speaker 3: first being a story you've been following that really amounts

27
00:01:23,920 --> 00:01:26,920
Speaker 3: to a cautionary tale for any New Zealand business launching

28
00:01:27,280 --> 00:01:31,000
Speaker 3: a public facing AI chatbot. How on earth did a

29
00:01:31,080 --> 00:01:35,520
Speaker 3: mental health assistance app end up dispensing advice on how

30
00:01:35,560 --> 00:01:38,800
Speaker 3: to perform incest? Walk us through this quagmire.

31
00:01:38,920 --> 00:01:44,039
Speaker 2: Ben so Auckland Company clear Head. They they offer a

32
00:01:44,080 --> 00:01:48,440
Speaker 2: mental health services EAP platform that they sell to businesses

33
00:01:48,480 --> 00:01:52,200
Speaker 2: and that helps businesses offer their staff connections with therapists.

34
00:01:52,200 --> 00:01:54,360
Speaker 2: That's their main kind of bread and butter, and as

35
00:01:54,440 --> 00:01:57,400
Speaker 2: part of that they also have this chatbot. This chatbot

36
00:01:57,440 --> 00:02:00,960
Speaker 2: has been advertised as kind of a something akin to

37
00:02:01,000 --> 00:02:03,560
Speaker 2: a replacement for a GP visit for mental health. It

38
00:02:03,640 --> 00:02:08,720
Speaker 2: can offer quote informal diagnoses. Has been variously mentioned at

39
00:02:08,720 --> 00:02:12,080
Speaker 2: different times, and originally it was just like a logic

40
00:02:12,240 --> 00:02:15,399
Speaker 2: chat bop that just you can go through answer response

41
00:02:16,200 --> 00:02:21,399
Speaker 2: hard coded responses. But they integrated Generative AI into their

42
00:02:21,440 --> 00:02:26,560
Speaker 2: offering to some customers, and as we know, Generative AI

43
00:02:27,120 --> 00:02:29,240
Speaker 2: has a bit of a tendency to go off the

44
00:02:29,320 --> 00:02:33,520
Speaker 2: rails at times, especially if you push it, and one man,

45
00:02:33,560 --> 00:02:37,720
Speaker 2: he pushed it quite hard, and the actual extent to

46
00:02:37,800 --> 00:02:40,720
Speaker 2: how hard that was is somewhat under debate depending on

47
00:02:40,760 --> 00:02:44,639
Speaker 2: who you talk to. But regardless what happened was the

48
00:02:44,760 --> 00:02:50,280
Speaker 2: AI chat Bob said some pretty crazy stuff, regardless of

49
00:02:51,040 --> 00:02:53,160
Speaker 2: the extent to which he pushed it. At the end

50
00:02:53,160 --> 00:02:55,440
Speaker 2: of the day, it still does really raise a lot

51
00:02:55,440 --> 00:02:58,639
Speaker 2: of questions about the use of generative AI in high

52
00:02:58,720 --> 00:03:02,359
Speaker 2: risk areas. And we can go back to my interview

53
00:03:02,400 --> 00:03:06,800
Speaker 2: with doctor Will Ready from Accentua talking about, you know,

54
00:03:06,880 --> 00:03:11,200
Speaker 2: the many concerns that many practitioners have around the use

55
00:03:11,200 --> 00:03:13,400
Speaker 2: of the technology in the medical field.

56
00:03:13,639 --> 00:03:16,560
Speaker 3: Yeah, and just to put this in perspective, so this guy,

57
00:03:16,639 --> 00:03:20,240
Speaker 3: Jim Nightingale, that AI whispery he calls himself, he was

58
00:03:20,280 --> 00:03:25,079
Speaker 3: able to prompt this chatbot to offer detailed sexual advice,

59
00:03:25,639 --> 00:03:29,000
Speaker 3: including how to talk constructively with his parents about being

60
00:03:29,120 --> 00:03:33,200
Speaker 3: voyeurs of his and his partner's sexual exploits. Okay, so

61
00:03:33,880 --> 00:03:36,920
Speaker 3: he didn't hack this chatbot too. This wasn't a prompt

62
00:03:36,920 --> 00:03:40,840
Speaker 3: injection or something like that. He just conversed with this

63
00:03:41,000 --> 00:03:44,760
Speaker 3: chat pot for a long amount of time as to

64
00:03:44,920 --> 00:03:49,080
Speaker 3: crazy questions, and eventually it started giving him pretty crazy answers.

65
00:03:50,000 --> 00:03:55,200
Speaker 2: Yes, and that's that's what he maintains. There is some

66
00:03:55,440 --> 00:03:59,840
Speaker 2: discussion that different people have different perspectives on whether or

67
00:03:59,840 --> 00:04:04,560
Speaker 2: not that you know that is fully accurate. I am

68
00:04:04,600 --> 00:04:07,840
Speaker 2: not in the position to speculate. All I can do

69
00:04:07,960 --> 00:04:11,560
Speaker 2: is report what different people are saying. Clearhead. They are

70
00:04:11,640 --> 00:04:18,120
Speaker 2: maintaining that he used high level prompt injection, really advanced techniques,

71
00:04:18,600 --> 00:04:24,240
Speaker 2: and that no normal person could ever have had such results. Regardless,

72
00:04:24,279 --> 00:04:27,599
Speaker 2: at the end of the day, these results did happen.

73
00:04:28,440 --> 00:04:31,560
Speaker 2: And one of the things that Clearhead is doing as

74
00:04:31,600 --> 00:04:34,840
Speaker 2: a result of this is they've firstly they've rolled it back,

75
00:04:35,000 --> 00:04:37,120
Speaker 2: but as they start to think about rolling it back out,

76
00:04:37,160 --> 00:04:38,840
Speaker 2: what one thing they're going to do is restrict it

77
00:04:39,080 --> 00:04:43,760
Speaker 2: the interactions with users to a much shorter time, which

78
00:04:43,800 --> 00:04:47,599
Speaker 2: seems sensible to me if it is as they have

79
00:04:47,920 --> 00:04:51,080
Speaker 2: otherwise advertised. It just supposed to be a navigator to

80
00:04:51,160 --> 00:04:53,559
Speaker 2: help people figure out what kind of therapist they should

81
00:04:53,560 --> 00:04:56,560
Speaker 2: be seeing, so that's how they said it should be,

82
00:04:56,640 --> 00:04:59,200
Speaker 2: but it also has kind of been coded to ask

83
00:04:59,200 --> 00:05:02,760
Speaker 2: people more questions and get more details about their mental health.

84
00:05:02,839 --> 00:05:06,120
Speaker 2: So it just feels a bit blurry, I think is

85
00:05:06,120 --> 00:05:09,960
Speaker 2: probably the best way to put it overall. And if

86
00:05:09,960 --> 00:05:14,240
Speaker 2: you are going to be using novel technologies in a

87
00:05:14,279 --> 00:05:20,640
Speaker 2: way that supports mental health, then you really should be

88
00:05:20,760 --> 00:05:25,560
Speaker 2: putting them through their paces in a big way. And

89
00:05:25,839 --> 00:05:29,719
Speaker 2: also it also raises the question to me of have

90
00:05:29,880 --> 00:05:33,279
Speaker 2: we really done enough research here about the role of

91
00:05:33,320 --> 00:05:38,000
Speaker 2: generative AI in a clinical mental health setting, because as

92
00:05:38,000 --> 00:05:40,280
Speaker 2: far as I can tell, there hasn't really been any

93
00:05:40,360 --> 00:05:45,560
Speaker 2: clinical studies done on generative AI to support mental health.

94
00:05:45,560 --> 00:05:47,680
Speaker 2: We've seen some stuff saying yes, bots can be good,

95
00:05:47,760 --> 00:05:50,760
Speaker 2: and they can be non judgmental and help people want

96
00:05:50,800 --> 00:05:53,640
Speaker 2: to open up and all these kinds of things, but

97
00:05:54,480 --> 00:05:58,200
Speaker 2: at the end of the day, you know, it's hard

98
00:05:58,200 --> 00:06:01,600
Speaker 2: to say for sure one way or the other, like absolutely,

99
00:06:01,600 --> 00:06:03,560
Speaker 2: they're only good, or they only bad, or they're good

100
00:06:03,600 --> 00:06:06,800
Speaker 2: under these circumstances, or these are the guardrails, or these

101
00:06:06,839 --> 00:06:08,760
Speaker 2: are the ethics that we need to be following, because

102
00:06:08,800 --> 00:06:12,760
Speaker 2: they just haven't really been developed at a high level

103
00:06:12,800 --> 00:06:14,840
Speaker 2: of research yet. And I think that a high level

104
00:06:14,839 --> 00:06:17,680
Speaker 2: of research is something that is important when it comes

105
00:06:17,720 --> 00:06:19,920
Speaker 2: to clinical applications of technology.

106
00:06:20,360 --> 00:06:23,400
Speaker 3: Yeah, so ultimately, clear head to said, oops, you know

107
00:06:23,440 --> 00:06:26,200
Speaker 3: we got this wrong. We've rolled it back. There is

108
00:06:26,960 --> 00:06:30,360
Speaker 3: some dispute and it may go legal between these parties.

109
00:06:31,520 --> 00:06:35,440
Speaker 3: Jim Nightingale has deleted his references to clear Head in

110
00:06:35,800 --> 00:06:37,880
Speaker 3: that medium post, so if you're looking for it, it won't

111
00:06:38,000 --> 00:06:42,520
Speaker 3: refer to clear Head, but sort of maintains his position

112
00:06:42,600 --> 00:06:45,400
Speaker 3: that this was literally asking this chatbot a series of

113
00:06:45,440 --> 00:06:50,159
Speaker 3: prompts to elicit these responses, as you say, you know,

114
00:06:50,200 --> 00:06:52,960
Speaker 3: the clear heads looking at limiting the interaction, the number

115
00:06:53,000 --> 00:06:57,120
Speaker 3: of back and forth sort of chats between a person

116
00:06:57,200 --> 00:06:59,640
Speaker 3: and the bot, which is exactly what Bing did and

117
00:07:00,080 --> 00:07:04,279
Speaker 3: others have done to try and stop or limit hallucinations

118
00:07:04,320 --> 00:07:08,400
Speaker 3: and inappropriate responses for two years of context for this

119
00:07:08,560 --> 00:07:11,680
Speaker 3: really that, particularly if you're doing something as sensitive as

120
00:07:11,760 --> 00:07:15,920
Speaker 3: mental health, you'd think that there are enough disaster stories

121
00:07:15,960 --> 00:07:17,560
Speaker 3: to inform your approach here.

122
00:07:18,000 --> 00:07:21,840
Speaker 2: And the gym that AI whisper it is not a

123
00:07:21,840 --> 00:07:26,080
Speaker 2: AI hater by any stretch of the imagination. He also

124
00:07:26,120 --> 00:07:30,000
Speaker 2: recently published a piece about GOVGPT, the New Zealand government's

125
00:07:30,160 --> 00:07:35,520
Speaker 2: small business support chatbot, and he spoke about how that

126
00:07:35,680 --> 00:07:38,400
Speaker 2: was a really good example of the use and related

127
00:07:38,400 --> 00:07:42,960
Speaker 2: it to retrieval augmented generation, which is specifically giving it

128
00:07:43,160 --> 00:07:46,960
Speaker 2: certain resources and saying these are your resources. Don't go

129
00:07:47,000 --> 00:07:50,560
Speaker 2: beyond the bounds of these resources, and really emphasizing that

130
00:07:50,640 --> 00:07:54,320
Speaker 2: your job is to help people get information from those resources,

131
00:07:54,760 --> 00:07:57,440
Speaker 2: nothing else. I had a play with gov GPT. Have

132
00:07:57,480 --> 00:07:58,600
Speaker 2: you had a play with it yet? Yeah?

133
00:07:58,640 --> 00:08:03,600
Speaker 3: Yeah, I have on specific questions like that, it's good.

134
00:08:03,640 --> 00:08:06,320
Speaker 3: But if it doesn't, you know, if it doesn't know

135
00:08:06,360 --> 00:08:10,360
Speaker 3: the answer from the limited sources I think ten websites

136
00:08:10,440 --> 00:08:13,080
Speaker 3: that have been fed into it, it will just tell

137
00:08:13,120 --> 00:08:15,800
Speaker 3: you all say, look, I don't have the answer to that,

138
00:08:15,960 --> 00:08:18,920
Speaker 3: go to this website for more information. So very quickly

139
00:08:19,040 --> 00:08:21,520
Speaker 3: it can become not very useful, and I guess that

140
00:08:21,640 --> 00:08:25,040
Speaker 3: is the balance we're talking about here. For mental health,

141
00:08:25,120 --> 00:08:28,600
Speaker 3: you want to have the AI chatbot have enough of

142
00:08:28,600 --> 00:08:31,720
Speaker 3: a conversation to figure out what that person's state of

143
00:08:31,800 --> 00:08:35,320
Speaker 3: mind is like at the moment and therefore to tailor

144
00:08:35,360 --> 00:08:39,280
Speaker 3: those responses at a very sensitive time, someone's potentially in crisis.

145
00:08:39,880 --> 00:08:43,720
Speaker 3: So you need enough information and to be able to

146
00:08:43,760 --> 00:08:47,520
Speaker 3: respond with nuance but not go off the rails. And

147
00:08:47,559 --> 00:08:51,600
Speaker 3: so gov GPT is an example a very conservative, closed

148
00:08:51,640 --> 00:08:54,800
Speaker 3: down and clear heads iteration off its Chapbot for a

149
00:08:54,840 --> 00:08:57,520
Speaker 3: while anyway, before they rolled it back sort of went

150
00:08:57,679 --> 00:09:00,160
Speaker 3: a little bit the other way. So I guess every

151
00:09:00,360 --> 00:09:04,840
Speaker 3: company needs to put the guardrails, appropriate guardrails in place.

152
00:09:04,880 --> 00:09:07,679
Speaker 3: They need to decide what level of risk they're willing

153
00:09:07,720 --> 00:09:11,880
Speaker 3: to live with. But ultimately, when it's you know, gov

154
00:09:12,000 --> 00:09:15,120
Speaker 3: GBT is just about finding government information on websites. When

155
00:09:15,120 --> 00:09:18,360
Speaker 3: you're dealing with something really personal and sensitive like mental health,

156
00:09:19,280 --> 00:09:21,840
Speaker 3: RAG should have been employed and it should have been

157
00:09:22,320 --> 00:09:23,360
Speaker 3: a lot more lockdown.

158
00:09:24,080 --> 00:09:26,680
Speaker 2: It does make me wonder, like we need to really think,

159
00:09:26,920 --> 00:09:31,000
Speaker 2: really really hard, what is it that this spot is

160
00:09:31,120 --> 00:09:33,920
Speaker 2: actually supposed to be doing. Is it supposed to be

161
00:09:34,240 --> 00:09:37,360
Speaker 2: getting basic information so that it can pass that on

162
00:09:38,000 --> 00:09:42,520
Speaker 2: to match with the correct therapist. Or is it supposed

163
00:09:42,520 --> 00:09:45,439
Speaker 2: to be offering platitudes and kitten pictures to make people

164
00:09:45,480 --> 00:09:48,720
Speaker 2: feel better about themselves or or or is it supposed

165
00:09:48,720 --> 00:09:50,920
Speaker 2: to be encouraging people to open up more and dig

166
00:09:50,920 --> 00:09:54,800
Speaker 2: into their trauma. But I'm not really comfortable with the

167
00:09:54,880 --> 00:09:58,440
Speaker 2: ladder myself. But yeah, maybe there are some some times

168
00:09:58,480 --> 00:10:01,079
Speaker 2: and places where that will be approp yep. But maybe

169
00:10:01,120 --> 00:10:03,280
Speaker 2: it's going to take us five years of research to

170
00:10:03,320 --> 00:10:06,160
Speaker 2: actually get to the point where we know what the

171
00:10:06,160 --> 00:10:10,360
Speaker 2: safeguards are, what the correct prompting is, what the models

172
00:10:10,400 --> 00:10:13,280
Speaker 2: we should be using, what the training data needs to be.

173
00:10:14,040 --> 00:10:18,160
Speaker 2: All of that stuff done tested, rigorously, done in a

174
00:10:18,200 --> 00:10:23,920
Speaker 2: clinical setting with proper experiments and proper authority and ethics

175
00:10:23,960 --> 00:10:26,160
Speaker 2: approval to get to that end state.

176
00:10:26,440 --> 00:10:30,440
Speaker 3: Yeah, we invited Angela Limon from Clearhead to talk about this.

177
00:10:30,720 --> 00:10:34,640
Speaker 3: She declined our invitations, so they clearly will be doing

178
00:10:34,720 --> 00:10:37,400
Speaker 3: some work behind the scenes to make sure this doesn't

179
00:10:37,440 --> 00:10:39,400
Speaker 3: happen again. You know, it seem like a great sort

180
00:10:39,440 --> 00:10:42,840
Speaker 3: of company, the goals of their business anyway. But definitely

181
00:10:42,840 --> 00:10:44,280
Speaker 3: if you're sitting here in a startup or in a

182
00:10:44,280 --> 00:10:47,720
Speaker 3: business working on a chat pot boy, this is a

183
00:10:47,800 --> 00:10:50,120
Speaker 3: solutary lesson absolutely.

184
00:10:50,920 --> 00:10:53,840
Speaker 2: So we'll keep you updated on any developments with Clearhead

185
00:10:53,880 --> 00:10:57,840
Speaker 2: and it's digital well being service now onto Google and

186
00:10:57,920 --> 00:10:59,960
Speaker 2: the hell is year it has had with the US

187
00:11:00,240 --> 00:11:01,640
Speaker 2: Department of Justice.

188
00:11:01,960 --> 00:11:06,440
Speaker 3: Yeah. Paul McIntyre is a journalist. He's a Kiwi actually

189
00:11:06,480 --> 00:11:09,360
Speaker 3: has been in Australia for a long time. For about

190
00:11:09,440 --> 00:11:13,520
Speaker 3: thirty years. He's covered the advertising sector the media as

191
00:11:13,559 --> 00:11:17,560
Speaker 3: a business journalist for Advertising Age, Cidney Morning Herald. Lots

192
00:11:17,559 --> 00:11:23,120
Speaker 3: of great publications over there. He's currently the executive editor

193
00:11:23,200 --> 00:11:26,840
Speaker 3: of MI three, a really great publication out of Australia

194
00:11:26,920 --> 00:11:30,600
Speaker 3: all about the advertising sector and the media. And he

195
00:11:30,840 --> 00:11:36,640
Speaker 3: has been following sort of Google's travails through the courts

196
00:11:36,720 --> 00:11:39,480
Speaker 3: in America over the last year or so. Has been

197
00:11:39,520 --> 00:11:43,280
Speaker 3: sort of looking at the antitrust space for a long

198
00:11:43,320 --> 00:11:47,000
Speaker 3: time now, and the move as glacial as it has

199
00:11:47,080 --> 00:11:51,000
Speaker 3: been towards holding Google to account for its monopoly power.

200
00:11:51,640 --> 00:11:54,920
Speaker 3: The premise of the US anti trust action against Google

201
00:11:55,040 --> 00:11:58,360
Speaker 3: is that businesses have been paying over the odds for

202
00:11:58,520 --> 00:12:01,599
Speaker 3: years now because of Google's monopoly power. You may be

203
00:12:01,720 --> 00:12:05,200
Speaker 3: reaching an audience and being excited about the results, but

204
00:12:05,360 --> 00:12:07,640
Speaker 3: how much more excited would you be and what better

205
00:12:07,679 --> 00:12:10,400
Speaker 3: results would you get if there was a diversity of

206
00:12:10,480 --> 00:12:14,760
Speaker 3: companies providing these types of advertising services. That's what really

207
00:12:14,760 --> 00:12:16,120
Speaker 3: goes to the heart of this case.

208
00:12:17,160 --> 00:12:19,839
Speaker 2: Yeah, and McIntire seems to know his stuff inside. And

209
00:12:19,880 --> 00:12:22,319
Speaker 2: Natzi has a really interesting interview that you had with.

210
00:12:29,320 --> 00:12:32,520
Speaker 3: Paul McIntyre. Welcome to the business of tech, So good

211
00:12:32,520 --> 00:12:33,320
Speaker 3: to see how you doing.

212
00:12:34,600 --> 00:12:37,920
Speaker 4: Hello, Peter Griffin. Was probably about a month ago we

213
00:12:37,960 --> 00:12:40,440
Speaker 4: might have had a great Californian red in San Francisco.

214
00:12:40,480 --> 00:12:40,920
Speaker 4: I believe.

215
00:12:41,040 --> 00:12:42,880
Speaker 3: Yeah, it's only a few weeks ago. We were at

216
00:12:43,240 --> 00:12:46,360
Speaker 3: the Salesforce conference in San Francisco. I went on to

217
00:12:46,480 --> 00:12:50,000
Speaker 3: Palo Alto. You went to the East Coast. Sounds like

218
00:12:50,080 --> 00:12:52,600
Speaker 3: you had a ball over there. Ended up in Virginia

219
00:12:53,240 --> 00:12:56,679
Speaker 3: where these antitrust cases, or at least one of them

220
00:12:56,679 --> 00:13:01,080
Speaker 3: has been heard against Google. What you actually want to

221
00:13:01,200 --> 00:13:03,360
Speaker 3: go all the way over to the East Coast to

222
00:13:03,400 --> 00:13:08,760
Speaker 3: sit through a very dry, boring and legally dense anti

223
00:13:08,760 --> 00:13:09,400
Speaker 3: trust case.

224
00:13:10,000 --> 00:13:13,080
Speaker 4: Well, you know the answer to that, Peter want mad

225
00:13:13,120 --> 00:13:16,600
Speaker 4: bastard journo. Firstly, in the second it's been it's been

226
00:13:17,480 --> 00:13:20,960
Speaker 4: a big issue for publishers and I've been following this

227
00:13:21,000 --> 00:13:23,880
Speaker 4: for you know, and the competition issues around big tech

228
00:13:23,960 --> 00:13:28,240
Speaker 4: for probably ten years. I was not so liked for

229
00:13:28,240 --> 00:13:30,880
Speaker 4: a while, but now it's sort of coming to that

230
00:13:31,120 --> 00:13:33,440
Speaker 4: there are really big issues. So the answer to your

231
00:13:33,520 --> 00:13:37,800
Speaker 4: question is it was purely context and intrigue.

232
00:13:38,160 --> 00:13:42,840
Speaker 3: And Paul, this is one of three major anti trust cases.

233
00:13:42,880 --> 00:13:46,480
Speaker 3: It's really all coalesced. In twenty twenty four, after years

234
00:13:46,480 --> 00:13:49,960
Speaker 3: of laying the groundwork the FTC and the Department of Justice,

235
00:13:50,280 --> 00:13:53,959
Speaker 3: we had the Search monopoly case that's already gone against Google.

236
00:13:54,000 --> 00:13:57,960
Speaker 3: Back in August, a district court found that Google violated

237
00:13:58,080 --> 00:14:02,240
Speaker 3: antitrust law by maintaining a notly in search. That was

238
00:14:02,320 --> 00:14:05,560
Speaker 3: because they paid billions of dollars every year to the

239
00:14:05,679 --> 00:14:09,040
Speaker 3: likes of Apple and Samsung to make sure the Google

240
00:14:09,120 --> 00:14:13,200
Speaker 3: Search engine is the default search engine in browsers and

241
00:14:13,240 --> 00:14:16,720
Speaker 3: on devices like the iPhone. So that went against them.

242
00:14:16,880 --> 00:14:19,280
Speaker 4: Only about twenty billion dollars worth of deals there too, Peter,

243
00:14:19,360 --> 00:14:20,480
Speaker 4: and those and those of right, I.

244
00:14:20,480 --> 00:14:23,600
Speaker 3: Think twenty six billion in twenty twenty one alone, so

245
00:14:23,680 --> 00:14:27,960
Speaker 3: that's massive. Then we've got this other one that's also

246
00:14:28,040 --> 00:14:32,680
Speaker 3: played out around the Google App Store, so judging that

247
00:14:32,680 --> 00:14:37,200
Speaker 3: that was anti competitive potentially wanting to allow other competitors

248
00:14:37,200 --> 00:14:41,280
Speaker 3: to set up their own app stores on an Android phone.

249
00:14:41,400 --> 00:14:43,080
Speaker 3: So that's a lot of grief for them. And then

250
00:14:43,120 --> 00:14:45,960
Speaker 3: the one that you're particularly interested in as someone who's

251
00:14:46,000 --> 00:14:50,360
Speaker 3: covered the advertising market, the marketing sector in Australia and

252
00:14:50,400 --> 00:14:53,520
Speaker 3: around the world for a long time, the digital advertising case.

253
00:14:54,560 --> 00:14:56,120
Speaker 3: What were they alleging in that one?

254
00:14:56,560 --> 00:14:59,720
Speaker 4: Well, we should probably just for some context, Peter, just

255
00:14:59,800 --> 00:15:03,960
Speaker 4: under stand how big this game is in this particular,

256
00:15:03,960 --> 00:15:07,800
Speaker 4: this third case you talk about with Google and Department

257
00:15:07,840 --> 00:15:11,920
Speaker 4: of Justice. So by twenty twenty eight digital, the global

258
00:15:11,920 --> 00:15:14,960
Speaker 4: digital advertising market will hit about a trillion dollars, right,

259
00:15:15,000 --> 00:15:17,480
Speaker 4: So that's what we're that's that's the pot that we're

260
00:15:17,480 --> 00:15:21,360
Speaker 4: talking about here. And traditionally, obviously that has been a

261
00:15:21,440 --> 00:15:23,960
Speaker 4: game for publishers and a lot of others, but publishers

262
00:15:24,000 --> 00:15:27,320
Speaker 4: sort of were dominant. Clearly, that's that's the ad supplied

263
00:15:27,320 --> 00:15:29,240
Speaker 4: part of it has changed considerably. So if we just

264
00:15:29,320 --> 00:15:31,920
Speaker 4: run through some some top line points out of probably

265
00:15:31,960 --> 00:15:34,560
Speaker 4: some of the all the cases really just for context,

266
00:15:34,920 --> 00:15:40,800
Speaker 4: it's all about the regulators finally clocking on to the

267
00:15:40,840 --> 00:15:45,280
Speaker 4: point that there may be adding competitive or competition issues

268
00:15:45,360 --> 00:15:48,320
Speaker 4: in tech. That's pretty much it. In whatever whatever area

269
00:15:48,440 --> 00:15:52,120
Speaker 4: they play, app stores, advertising, social, just as a as

270
00:15:52,160 --> 00:15:55,160
Speaker 4: a platform, there are there is dominance, and of course

271
00:15:55,280 --> 00:15:57,840
Speaker 4: historically we know that can lead to problems. So the

272
00:15:57,840 --> 00:16:00,760
Speaker 4: regulators have got over we've all got over the I guess,

273
00:16:00,800 --> 00:16:03,560
Speaker 4: the novelty and the romance of big tech for the

274
00:16:03,640 --> 00:16:05,840
Speaker 4: last fifteen years and now realizing, Okay, there's some good

275
00:16:05,840 --> 00:16:09,240
Speaker 4: stuff that happens, but there's also some shit that happens,

276
00:16:09,280 --> 00:16:11,840
Speaker 4: and sometimes it's not great for lots of people. But

277
00:16:11,960 --> 00:16:14,200
Speaker 4: it's they're still stars. So that's the context. And by

278
00:16:14,240 --> 00:16:16,320
Speaker 4: the way, it's just not the US regulators around the

279
00:16:16,320 --> 00:16:21,200
Speaker 4: world jumping on this. So the UK Competition and Market's

280
00:16:21,200 --> 00:16:26,560
Speaker 4: Authority is already designated pride to the US legal system

281
00:16:26,640 --> 00:16:31,200
Speaker 4: designating it last month that Google's ad supply chain is monopolistic,

282
00:16:31,600 --> 00:16:34,920
Speaker 4: monopolistic and anti competitive. So it is happening everywhere. It's

283
00:16:34,920 --> 00:16:37,760
Speaker 4: just that in the home markets of these big tech companies.

284
00:16:37,760 --> 00:16:40,240
Speaker 4: It gets super interesting because obviously the US has been

285
00:16:40,600 --> 00:16:43,440
Speaker 4: less inclined to want to regulate their star performance. So

286
00:16:43,640 --> 00:16:46,360
Speaker 4: a couple of let's go through sort of some big points,

287
00:16:46,400 --> 00:16:48,400
Speaker 4: I think, just to set up the context of this.

288
00:16:48,560 --> 00:16:52,760
Speaker 4: So Google dominates the global ad tech system. It dominates search,

289
00:16:52,880 --> 00:16:56,360
Speaker 4: it dominates app stores, on the Android system, and the

290
00:16:56,400 --> 00:16:58,680
Speaker 4: browser market. So you think about Chrome, it's at all

291
00:16:58,800 --> 00:17:01,880
Speaker 4: those things. We know that Google Search, its ad tech

292
00:17:02,000 --> 00:17:05,439
Speaker 4: is so dominant that out of this court case that

293
00:17:05,480 --> 00:17:11,360
Speaker 4: we're about to cover, some evidence emerged where its executives

294
00:17:11,480 --> 00:17:15,320
Speaker 4: said that only an act of God could dislodge Google

295
00:17:15,400 --> 00:17:20,200
Speaker 4: from controlling the global advertising supply system. Its Chrome browser

296
00:17:20,200 --> 00:17:25,520
Speaker 4: has about sixty six percent share of global browsers. It

297
00:17:25,560 --> 00:17:28,760
Speaker 4: has ninety percent share of global searches and ninety five

298
00:17:28,800 --> 00:17:32,400
Speaker 4: percent on mobile. Google AD Manager, which is the way

299
00:17:32,440 --> 00:17:37,760
Speaker 4: in which auctions are facilitated for its broader ecosystem of

300
00:17:37,800 --> 00:17:42,080
Speaker 4: publishers that ad exchange. It controls ninety one percent of

301
00:17:42,119 --> 00:17:47,520
Speaker 4: the global ad server market. Eight of the ten top

302
00:17:47,800 --> 00:17:50,240
Speaker 4: advertisers in the world rely on it. Nine of ten

303
00:17:50,320 --> 00:17:54,240
Speaker 4: global publishers will rely on it. Google Ads controls eighty

304
00:17:54,240 --> 00:17:58,840
Speaker 4: seven percent of global ad demand. Google's ad exchange trades

305
00:17:58,880 --> 00:18:02,119
Speaker 4: fifty six percent of the world ads and eighty seven

306
00:18:02,119 --> 00:18:05,800
Speaker 4: percent of its entire revenue comes from ads and thirteen

307
00:18:06,000 --> 00:18:08,920
Speaker 4: billion publisher ads per day delivered by Google. But it's

308
00:18:09,040 --> 00:18:10,959
Speaker 4: nothing compared to its own asset. So there's this some

309
00:18:11,000 --> 00:18:13,280
Speaker 4: of the top line stuff. Now, what came out of

310
00:18:13,280 --> 00:18:16,119
Speaker 4: that is that the dominance and this is with the

311
00:18:16,160 --> 00:18:19,439
Speaker 4: Department of Justice, We're going, okay, this is now a

312
00:18:19,560 --> 00:18:22,760
Speaker 4: dominant monopolistic behavior and player, what are we going to

313
00:18:22,760 --> 00:18:25,080
Speaker 4: do about it? They take them to court in this

314
00:18:25,280 --> 00:18:28,159
Speaker 4: really interesting case. And this is what happens with I

315
00:18:28,200 --> 00:18:31,800
Speaker 4: think tech early on Peter from both the socials and

316
00:18:31,840 --> 00:18:36,000
Speaker 4: the likes of Google, so Meta and Google, the early

317
00:18:36,080 --> 00:18:41,720
Speaker 4: gains for these tech players came from publishing where most

318
00:18:41,720 --> 00:18:45,120
Speaker 4: of the content was being consumed by publishers and their

319
00:18:45,160 --> 00:18:47,960
Speaker 4: sites that moved, of course, but the early advantage was

320
00:18:47,960 --> 00:18:50,480
Speaker 4: hooking up with the publishers to get that momentum and

321
00:18:50,520 --> 00:18:53,040
Speaker 4: then siphoning away the audience, which exactly what's happened. So

322
00:18:53,040 --> 00:18:55,560
Speaker 4: they don't need publishers anymore. They did at the start,

323
00:18:55,600 --> 00:18:57,840
Speaker 4: and the point here is that we, the publishers, fell

324
00:18:57,880 --> 00:19:00,439
Speaker 4: for that. We got a bit carried and did some

325
00:19:00,480 --> 00:19:04,560
Speaker 4: silly stuff really, and so what happens now is they dominated.

326
00:19:04,640 --> 00:19:07,080
Speaker 4: So you had a piece of evidence that came out

327
00:19:07,359 --> 00:19:09,960
Speaker 4: in the case in Virginia a couple of weeks ago

328
00:19:10,400 --> 00:19:12,400
Speaker 4: where a guy called David Rosenblad, who was the head

329
00:19:12,400 --> 00:19:16,320
Speaker 4: of Google's ads, basically said at the start, when in

330
00:19:16,320 --> 00:19:18,760
Speaker 4: two thousand and eight they bought this company called double Click,

331
00:19:18,760 --> 00:19:23,000
Speaker 4: which let them represent other publishers and the long tail,

332
00:19:23,080 --> 00:19:26,200
Speaker 4: really small sites that don't have the ability to monetize

333
00:19:26,240 --> 00:19:29,880
Speaker 4: themselves or have sales teams. Google developed this exchange where

334
00:19:29,880 --> 00:19:32,280
Speaker 4: it could sell on behalf of these small long tail publishers.

335
00:19:32,320 --> 00:19:35,600
Speaker 4: So in the start, what the strategy was that they

336
00:19:35,680 --> 00:19:39,040
Speaker 4: needed to keep the publishers on side. And the quote

337
00:19:39,040 --> 00:19:40,920
Speaker 4: was you have to understand this is from David Rosenblad.

338
00:19:40,920 --> 00:19:43,720
Speaker 4: You have to understand that while bit the big advertisers

339
00:19:44,680 --> 00:19:48,760
Speaker 4: account for the revenue, the business depends on relationships with publishers.

340
00:19:48,800 --> 00:19:51,159
Speaker 4: If you don't have access to inventory, you have nothing

341
00:19:51,240 --> 00:19:54,320
Speaker 4: to say to advertisers. And then later on he said

342
00:19:54,840 --> 00:19:58,000
Speaker 4: the mission was to crush competition and boast that leaving

343
00:19:58,000 --> 00:20:01,200
Speaker 4: Google's ad tech stack was an mayor for publishers and

344
00:20:01,280 --> 00:20:03,399
Speaker 4: required an active god. So they got them in early

345
00:20:03,640 --> 00:20:06,440
Speaker 4: and then changed it around to work their business model.

346
00:20:06,600 --> 00:20:09,600
Speaker 4: And you see it with what's happened now with the

347
00:20:09,600 --> 00:20:12,359
Speaker 4: Global at Exchange. So look, that's I'm going to stop

348
00:20:12,359 --> 00:20:14,159
Speaker 4: there because we can go down a rabbit warrem what

349
00:20:14,160 --> 00:20:15,680
Speaker 4: do you want to know, Peter Griffin.

350
00:20:15,680 --> 00:20:18,600
Speaker 3: Well, I mean some of those status you read out

351
00:20:18,680 --> 00:20:22,640
Speaker 3: really painted in stark terms. I mean it's big retailers

352
00:20:23,000 --> 00:20:25,399
Speaker 3: Cols or in New Zealand, the warehouse or in New

353
00:20:25,480 --> 00:20:30,480
Speaker 3: Zealand every time they place adverts on the Google ad platform.

354
00:20:30,960 --> 00:20:36,600
Speaker 3: According to the US government, they've basically been fleeced collectively

355
00:20:36,720 --> 00:20:40,200
Speaker 3: over the years to the tune of billions of dollars. Collectively.

356
00:20:40,240 --> 00:20:43,880
Speaker 3: These advertisers by paying over the odds now the other

357
00:20:43,920 --> 00:20:46,920
Speaker 3: two cases we touched on the search monopoly a little

358
00:20:46,920 --> 00:20:49,720
Speaker 3: bit more clear cut about how this went about. The

359
00:20:49,760 --> 00:20:54,840
Speaker 3: app store obviously is a walled garden that Google controls

360
00:20:55,040 --> 00:20:58,760
Speaker 3: within the play Store. But this digital advertising case is

361
00:20:58,800 --> 00:21:02,439
Speaker 3: the most complex one because what they're alleging there is

362
00:21:02,480 --> 00:21:05,880
Speaker 3: that Google did all these sorts of things with auctions

363
00:21:06,200 --> 00:21:09,800
Speaker 3: and with ad tech and with those sort of acquisitions

364
00:21:09,800 --> 00:21:14,440
Speaker 3: it did, like double click to really create a stranglehold

365
00:21:14,680 --> 00:21:17,280
Speaker 3: on the digital ad market. What are some of the

366
00:21:17,280 --> 00:21:19,600
Speaker 3: things that you heard talked about there that there are

367
00:21:19,640 --> 00:21:24,240
Speaker 3: actually alleged to be doing that really cemented that monopoly

368
00:21:24,400 --> 00:21:25,760
Speaker 3: in digital advertising.

369
00:21:26,280 --> 00:21:28,800
Speaker 4: It's so mind bending that we're not going to bore

370
00:21:28,840 --> 00:21:31,359
Speaker 4: your listers with the detail. They'll just have to trust

371
00:21:31,440 --> 00:21:34,359
Speaker 4: us that the top line directional is right because we

372
00:21:34,359 --> 00:21:35,680
Speaker 4: can get into the weeds here and it's going to

373
00:21:35,680 --> 00:21:37,399
Speaker 4: get ugly and we don't want to do that, Peter So.

374
00:21:37,640 --> 00:21:41,040
Speaker 4: But essentially, in the end, it built, it built such

375
00:21:41,080 --> 00:21:43,439
Speaker 4: a system in the in the ad tech and the

376
00:21:43,480 --> 00:21:47,159
Speaker 4: ad server market that represented the lines heare of the

377
00:21:47,200 --> 00:21:50,600
Speaker 4: long tail of publishers, smaller publishers, but then got the

378
00:21:50,600 --> 00:21:53,480
Speaker 4: big ones involved as well. And it's the infrastructure so

379
00:21:53,480 --> 00:21:56,879
Speaker 4: it's essentially built itself. It's the incumbent infrastructure for the world,

380
00:21:56,880 --> 00:22:02,879
Speaker 4: if you like, for distributing advertising, matching audiences with ads,

381
00:22:03,119 --> 00:22:07,719
Speaker 4: so it's essentially the proxy gateway. And so by doing that,

382
00:22:07,800 --> 00:22:12,399
Speaker 4: it's been able to overcharge's and it's taking a huge clip,

383
00:22:12,480 --> 00:22:14,679
Speaker 4: so it's ten times. So if you were using if

384
00:22:14,720 --> 00:22:18,040
Speaker 4: you were able to use Arrival before they essentially knocked

385
00:22:18,040 --> 00:22:21,040
Speaker 4: most of them off, Google costs ten times more than

386
00:22:21,080 --> 00:22:24,240
Speaker 4: those other rivals now that they control the supply chain,

387
00:22:24,320 --> 00:22:26,600
Speaker 4: so there's a cost play. You think about like a

388
00:22:26,680 --> 00:22:29,200
Speaker 4: stock market, and you've got buyers and sellers. So that's

389
00:22:29,240 --> 00:22:32,360
Speaker 4: what Google controls in terms of the biggest ad exchange

390
00:22:32,359 --> 00:22:34,960
Speaker 4: in the world at fifty six percent, so it controls

391
00:22:34,960 --> 00:22:37,680
Speaker 4: a stock market, you know, trading floor if you like,

392
00:22:38,240 --> 00:22:40,560
Speaker 4: for advertising. And what we see is that as it

393
00:22:40,600 --> 00:22:43,959
Speaker 4: got publishes in built momentum, it also has its own assets, right,

394
00:22:44,000 --> 00:22:48,000
Speaker 4: so it owns YouTube and an owned search and so

395
00:22:48,040 --> 00:22:51,359
Speaker 4: what happens is it starts in the auction system. It

396
00:22:51,440 --> 00:22:55,800
Speaker 4: was favoring its own assets over the third party publish

397
00:22:55,840 --> 00:22:58,600
Speaker 4: a supply chain, so it was manipulating the auctions. It

398
00:22:58,640 --> 00:23:04,199
Speaker 4: was loading the the ad tech tax it bought competitors

399
00:23:04,240 --> 00:23:07,520
Speaker 4: shut them down so they couldn't they couldn't be arrival,

400
00:23:07,600 --> 00:23:11,119
Speaker 4: couldn't have any competition. So all those consistent behaviors we

401
00:23:11,160 --> 00:23:16,560
Speaker 4: see with how you grow and control the market is

402
00:23:16,560 --> 00:23:22,320
Speaker 4: what we're seeing Google do through these really complicated interactions

403
00:23:22,359 --> 00:23:25,400
Speaker 4: between different tech stacks and the market.

404
00:23:25,200 --> 00:23:30,080
Speaker 3: And also accusations of collusion with Facebook, one of the

405
00:23:30,119 --> 00:23:33,920
Speaker 3: other big digital advertising players. Tell us about that.

406
00:23:34,640 --> 00:23:36,640
Speaker 4: In about twenty eighteen, I think there's a thing called

407
00:23:36,920 --> 00:23:38,800
Speaker 4: head of bidding emerged where a publisher like The New

408
00:23:38,840 --> 00:23:42,760
Speaker 4: Zealand Herald could send out its inventory to multiple at

409
00:23:42,840 --> 00:23:47,040
Speaker 4: exchanges at once and choose the best, the highest, the

410
00:23:47,040 --> 00:23:51,000
Speaker 4: best margin, the best play right in real time. So

411
00:23:51,040 --> 00:23:53,040
Speaker 4: that it was essentially a move to be able to

412
00:23:53,119 --> 00:23:55,360
Speaker 4: do that. That was head of bidding. Google and Facebook

413
00:23:55,400 --> 00:23:57,840
Speaker 4: saw that as a major threat to their system of

414
00:23:57,880 --> 00:24:04,360
Speaker 4: controlling actions and icing and the gateway to inventory, so

415
00:24:04,640 --> 00:24:08,600
Speaker 4: they colluded to try and offset this and stop it.

416
00:24:08,680 --> 00:24:11,439
Speaker 4: And you would have seen that new initiative that Google

417
00:24:11,440 --> 00:24:14,320
Speaker 4: brought out called AMP you know, so it was essentially

418
00:24:14,720 --> 00:24:17,280
Speaker 4: under the guise of loading your pages faster and so forth,

419
00:24:17,320 --> 00:24:19,359
Speaker 4: a MP it was accounted ahead of bidding, but it

420
00:24:19,400 --> 00:24:22,000
Speaker 4: allowed them to control the inventory. It kept it inside

421
00:24:22,000 --> 00:24:25,280
Speaker 4: their system again, right, And that's essentially the main play

422
00:24:25,320 --> 00:24:27,119
Speaker 4: of this is that instead of publishers be able to

423
00:24:27,160 --> 00:24:29,840
Speaker 4: go to Google and other exchanges and go who's going

424
00:24:29,880 --> 00:24:32,120
Speaker 4: to give me the best deal, Google said, no, you're

425
00:24:32,160 --> 00:24:33,679
Speaker 4: not going to do that. If you use head a bidding,

426
00:24:33,920 --> 00:24:35,720
Speaker 4: We're not going to supply you the data that you

427
00:24:35,800 --> 00:24:40,000
Speaker 4: need from us to compare what we've done versus competitors.

428
00:24:40,080 --> 00:24:42,480
Speaker 4: So it shut down a data supply and therefore no

429
00:24:42,560 --> 00:24:45,160
Speaker 4: transparency in the market to say, okay, who was better

430
00:24:45,240 --> 00:24:48,880
Speaker 4: Google or alternative XT. So those sorts of things were

431
00:24:48,960 --> 00:24:53,040
Speaker 4: anything that was basically threatening the control or the opening

432
00:24:53,080 --> 00:24:56,159
Speaker 4: up of the system to other players. It was working

433
00:24:56,800 --> 00:24:59,800
Speaker 4: quite passionately about Peter to stop.

434
00:25:00,920 --> 00:25:03,960
Speaker 3: So some pretty compelling evidence. Obviously, they've lost the search

435
00:25:04,040 --> 00:25:07,920
Speaker 3: monopoly case. They are appealing that, but already the court

436
00:25:07,960 --> 00:25:11,119
Speaker 3: is looking at what the remedy is there. They've lost

437
00:25:11,320 --> 00:25:15,439
Speaker 3: the app store case that Epic brought against Google, so

438
00:25:15,960 --> 00:25:19,879
Speaker 3: they're looking at remedy there as well. The digital advertising

439
00:25:20,119 --> 00:25:23,520
Speaker 3: case ended prematurely, so we should have a judgment maybe

440
00:25:23,560 --> 00:25:25,240
Speaker 3: before the end of the year or early next year

441
00:25:25,280 --> 00:25:26,399
Speaker 3: on that one.

442
00:25:26,880 --> 00:25:30,120
Speaker 4: Yes, end of the year, you'll probably see something from

443
00:25:31,119 --> 00:25:34,920
Speaker 4: from the judge in the District Court US District Court

444
00:25:35,000 --> 00:25:37,760
Speaker 4: of Virginia, and by the way, just on that. So

445
00:25:37,800 --> 00:25:40,840
Speaker 4: you're right, Google has lost every case so far, and

446
00:25:40,880 --> 00:25:43,359
Speaker 4: there's more coming to by the way, with Apple on

447
00:25:43,400 --> 00:25:46,760
Speaker 4: its app stored monopoly and dominance Apple and its claims

448
00:25:46,800 --> 00:25:51,640
Speaker 4: around privacy and perhaps some duplicitus or some hypocrisy there

449
00:25:51,640 --> 00:25:53,880
Speaker 4: and the claims versus what it's doing because Apples moving

450
00:25:53,880 --> 00:25:56,120
Speaker 4: into the AD system surprisingly and there's a whole bunch

451
00:25:56,160 --> 00:25:58,400
Speaker 4: to come over there. So there is a broad sweep

452
00:25:58,520 --> 00:26:02,200
Speaker 4: of k is coming just not at Google, but at

453
00:26:02,240 --> 00:26:06,320
Speaker 4: Meta as well. And there's another really really interesting case

454
00:26:06,359 --> 00:26:11,119
Speaker 4: coming up with by eighteen attorneys state attorneys general in

455
00:26:11,160 --> 00:26:13,679
Speaker 4: the US, which is even you know, even racier with

456
00:26:13,720 --> 00:26:15,920
Speaker 4: some of the evidents they've got. So there's more to come.

457
00:26:16,119 --> 00:26:19,600
Speaker 4: But essentially the three cases we've seen this year, both

458
00:26:19,760 --> 00:26:23,840
Speaker 4: as you say, apps and search, Google has been ruled monopolistic.

459
00:26:24,320 --> 00:26:28,919
Speaker 4: And in this final case of all, the commentary and

460
00:26:29,040 --> 00:26:31,720
Speaker 4: expectation is and certainly some of the comments from and

461
00:26:31,760 --> 00:26:35,480
Speaker 4: how the judge judges line of questioning would suggest that

462
00:26:35,640 --> 00:26:39,040
Speaker 4: Google's in some trouble in around. It's control of the

463
00:26:39,080 --> 00:26:40,040
Speaker 4: ourt tech system as well.

464
00:26:40,760 --> 00:26:43,080
Speaker 3: Just want to sort of step back a little bit

465
00:26:43,280 --> 00:26:47,520
Speaker 3: and you know you've covered this market advertising in Australia,

466
00:26:47,600 --> 00:26:52,360
Speaker 3: big advertising market, digital advertising, digital marketing is huge there.

467
00:26:52,440 --> 00:26:54,280
Speaker 3: Google is really at the heart of that we're talking

468
00:26:54,320 --> 00:26:56,800
Speaker 3: about in New Zealand a loan. It's a billion dollar

469
00:26:56,880 --> 00:27:01,639
Speaker 3: business for Google. They don't report that as revenue because

470
00:27:01,640 --> 00:27:05,359
Speaker 3: they transfer a lot of that revenue to other jurisdictions,

471
00:27:05,480 --> 00:27:08,200
Speaker 3: but they're basically turning over a billion here. I'm guessing

472
00:27:08,240 --> 00:27:10,160
Speaker 3: it's more like four billion in Australia.

473
00:27:10,200 --> 00:27:12,840
Speaker 4: We think it's about six billion in Australia, and I

474
00:27:12,880 --> 00:27:15,560
Speaker 4: think some numbers at the age we'll see that our

475
00:27:15,560 --> 00:27:19,200
Speaker 4: competition regulator did with its digital platforms inquiry over the

476
00:27:19,280 --> 00:27:22,760
Speaker 4: last five or three four years bears out about six

477
00:27:22,800 --> 00:27:23,760
Speaker 4: billion a bit more. So.

478
00:27:23,800 --> 00:27:27,679
Speaker 3: It's incredible. So it's a huge business. When did it

479
00:27:27,760 --> 00:27:30,560
Speaker 3: start to occur to you in this whole process, the

480
00:27:30,640 --> 00:27:37,040
Speaker 3: rise of Google. It's something was fundamentally tipped in Google's favor.

481
00:27:36,359 --> 00:27:39,080
Speaker 4: Probably about ten years ago. I started I started arcing

482
00:27:39,160 --> 00:27:42,080
Speaker 4: up ten years ago, Peter, both on Meta and Google.

483
00:27:43,000 --> 00:27:48,639
Speaker 4: I had been I'd seen the early wave of the

484
00:27:48,680 --> 00:27:52,480
Speaker 4: ad tech startup community coming out of out of the

485
00:27:52,520 --> 00:27:56,680
Speaker 4: West Coast in probably two thousand and nine ten, And interestingly,

486
00:27:56,720 --> 00:27:59,439
Speaker 4: they were all coming out of the stock equity markets

487
00:27:59,480 --> 00:28:03,000
Speaker 4: and finance markets on the East coast, seeing an opportunity

488
00:28:03,080 --> 00:28:09,440
Speaker 4: to this this big laboring, lumbering, bloated, non tech, non

489
00:28:09,480 --> 00:28:13,719
Speaker 4: automated sort of industry, and saw where we can light

490
00:28:13,800 --> 00:28:16,480
Speaker 4: this up with some automation and programmatic advertising, is what

491
00:28:16,480 --> 00:28:19,200
Speaker 4: what what became? So I saw I was, you know,

492
00:28:19,200 --> 00:28:21,000
Speaker 4: at the time, I was obviously covering the media sector,

493
00:28:21,040 --> 00:28:22,919
Speaker 4: had done for a long time before that, and I

494
00:28:23,000 --> 00:28:25,960
Speaker 4: was getting to pitch from these these startups out of

495
00:28:25,960 --> 00:28:28,480
Speaker 4: the out of Silicon Valley Valley saying this is great

496
00:28:28,480 --> 00:28:30,639
Speaker 4: for publishing, Paul. This is going to increase margins and

497
00:28:30,680 --> 00:28:32,639
Speaker 4: publishers are going it's going to drop the cost of

498
00:28:33,440 --> 00:28:37,879
Speaker 4: doing business and take out take out operating costs and

499
00:28:37,960 --> 00:28:40,360
Speaker 4: allow you and allow publishers to set a floor on

500
00:28:40,400 --> 00:28:42,960
Speaker 4: their on their pricing and go up from there. And

501
00:28:43,000 --> 00:28:45,000
Speaker 4: it's all good and life is going to be so

502
00:28:45,120 --> 00:28:48,040
Speaker 4: much better, and you should write good things about us,

503
00:28:48,400 --> 00:28:50,200
Speaker 4: which of course I did for the first you know,

504
00:28:50,680 --> 00:28:53,520
Speaker 4: two years. Where you know, publishers, we're all we're all going.

505
00:28:53,800 --> 00:28:55,360
Speaker 4: You know, there was there was always a sense of

506
00:28:56,600 --> 00:28:59,880
Speaker 4: be careful here. But you know, the tech was such untouched,

507
00:29:00,120 --> 00:29:03,680
Speaker 4: wasn't it. Remember it was everything from I pos to startups.

508
00:29:03,680 --> 00:29:06,840
Speaker 4: There were gods, demigods and how dare you question any

509
00:29:06,880 --> 00:29:09,760
Speaker 4: of this stuff? So I you know, reluctantly or carefully

510
00:29:09,800 --> 00:29:11,800
Speaker 4: went along with it. But after about two years realized

511
00:29:12,920 --> 00:29:16,800
Speaker 4: that it was a crunch of you know, rotten potatoes. Really.

512
00:29:16,920 --> 00:29:18,760
Speaker 4: In fact, what was happening was that all it did

513
00:29:18,960 --> 00:29:22,280
Speaker 4: was that the automation and of the supply chain and

514
00:29:22,320 --> 00:29:27,440
Speaker 4: advertising was deflationary. So it essentially put prices down, and

515
00:29:27,480 --> 00:29:30,360
Speaker 4: for a whole bunch of reasons, including the fact that

516
00:29:30,360 --> 00:29:34,720
Speaker 4: there was all sorts of nefarious activities going on inside

517
00:29:34,720 --> 00:29:37,960
Speaker 4: the digital advertising system. There was players. I mean, just

518
00:29:37,960 --> 00:29:40,920
Speaker 4: just so everybody knows, it's important to know that the

519
00:29:41,880 --> 00:29:46,640
Speaker 4: digital advertising market is the second largest most lucrative for

520
00:29:46,960 --> 00:29:51,520
Speaker 4: criminals in the world behind narcotics. That's like ad fraud,

521
00:29:51,760 --> 00:29:55,120
Speaker 4: and the ability to gain money out of the system

522
00:29:55,280 --> 00:29:58,960
Speaker 4: by producing all sorts of really interesting fraudulent activities is

523
00:29:59,040 --> 00:30:01,960
Speaker 4: massive to the of you know, estimates range between ten

524
00:30:02,000 --> 00:30:05,280
Speaker 4: and twenty five billion in fraud going to criminal syndicates,

525
00:30:05,520 --> 00:30:09,680
Speaker 4: and we're talking about you know, everything from GEO you know, regimes,

526
00:30:09,960 --> 00:30:15,840
Speaker 4: political regimes that have alternative agendas, to just sort of

527
00:30:16,400 --> 00:30:20,520
Speaker 4: great mad hackers who can make money out of gaming

528
00:30:20,560 --> 00:30:24,600
Speaker 4: the system. So there's a lot in that. And I

529
00:30:24,680 --> 00:30:27,520
Speaker 4: started to see very early on that there was what

530
00:30:27,560 --> 00:30:30,840
Speaker 4: the claims were in programmatic wasn't happening. And then when

531
00:30:30,920 --> 00:30:33,240
Speaker 4: we started to see, you know how Google start to

532
00:30:33,280 --> 00:30:36,160
Speaker 4: dominate and link up the power wasn't linking up its

533
00:30:36,440 --> 00:30:41,040
Speaker 4: enterprise stack. So it had Search, it had YouTube, it

534
00:30:41,120 --> 00:30:46,160
Speaker 4: had Gmail, it has enterprise all the way through. It's

535
00:30:46,200 --> 00:30:51,240
Speaker 4: loaded into the enterprise software systems as well, so it

536
00:30:51,280 --> 00:30:54,080
Speaker 4: was really powerful and being embedded. And you could see

537
00:30:54,080 --> 00:30:57,520
Speaker 4: that happening, and you could see that it wasn't The

538
00:30:57,520 --> 00:31:01,120
Speaker 4: publishers were struggling. The journals were being off low, revenues

539
00:31:01,560 --> 00:31:04,280
Speaker 4: were declining, and of course you know they'd also lost

540
00:31:04,280 --> 00:31:06,440
Speaker 4: a lot of their classified stuff which was underpinning and

541
00:31:06,520 --> 00:31:10,080
Speaker 4: underwriting journalism per se. So long answer to say, probably

542
00:31:10,120 --> 00:31:10,560
Speaker 4: ten years.

543
00:31:16,920 --> 00:31:21,280
Speaker 3: You know, you talk to big Australian advertisers every day.

544
00:31:21,400 --> 00:31:25,200
Speaker 3: You know, we're talking about big retailers, you know, Quantus Telstra.

545
00:31:26,280 --> 00:31:29,160
Speaker 3: I mean they Yeah, you don't hear them railing against Google.

546
00:31:29,520 --> 00:31:33,080
Speaker 3: It's just sort of been a silent takeover. Really, do

547
00:31:33,120 --> 00:31:36,040
Speaker 3: you think that advertisers, why hasn't there been a concerned

548
00:31:36,120 --> 00:31:39,920
Speaker 3: pushback against Google. They would have seen internally what was happening,

549
00:31:40,400 --> 00:31:43,080
Speaker 3: that their options were getting more and more limited, that

550
00:31:43,120 --> 00:31:46,760
Speaker 3: they're fundamentally paying more in the long run. Why hasn't

551
00:31:46,760 --> 00:31:49,240
Speaker 3: the industry been sort of saying, we're not going to

552
00:31:49,240 --> 00:31:50,000
Speaker 3: take this anymore.

553
00:31:50,600 --> 00:31:53,720
Speaker 4: Well, it's probably only starting to happen now, Peter, in

554
00:31:53,760 --> 00:31:57,160
Speaker 4: that the costs for instance, of search, now it was

555
00:31:57,200 --> 00:32:00,680
Speaker 4: good times. Don't like this is not about whether Google

556
00:32:00,760 --> 00:32:04,160
Speaker 4: worked and whether Google worked for advertisers or worked for

557
00:32:04,200 --> 00:32:07,480
Speaker 4: the consumer, because clearly consumers said, you know, we're all

558
00:32:07,520 --> 00:32:11,760
Speaker 4: doing search Google Search. I mean, that's a that's that's

559
00:32:11,800 --> 00:32:14,560
Speaker 4: got some big clouds over it, you know, in the

560
00:32:14,560 --> 00:32:16,880
Speaker 4: next couple of years. But the reality was the products

561
00:32:16,920 --> 00:32:19,120
Speaker 4: were good. You know, this is not about but it's

562
00:32:19,160 --> 00:32:21,480
Speaker 4: what allowed them to make their products good and what

563
00:32:21,520 --> 00:32:25,280
Speaker 4: allowed them to make their products untouchable by competitors is

564
00:32:25,680 --> 00:32:29,680
Speaker 4: the issue here, And what would happen if if Google

565
00:32:29,800 --> 00:32:33,720
Speaker 4: didn't control the ad supply system like like it does now,

566
00:32:33,880 --> 00:32:35,680
Speaker 4: would there be more innovation. And this is the question

567
00:32:35,680 --> 00:32:38,040
Speaker 4: that goes right back to the Microsoft, you know in case,

568
00:32:38,080 --> 00:32:40,040
Speaker 4: the anti trust case back in the late nineties, which

569
00:32:40,080 --> 00:32:42,400
Speaker 4: is the pretty much the last time we've seen anything

570
00:32:42,480 --> 00:32:46,040
Speaker 4: of this size happen in tech or full stop industry,

571
00:32:46,080 --> 00:32:49,040
Speaker 4: full stop in terms of anti trade or as the

572
00:32:49,160 --> 00:32:51,840
Speaker 4: US call it, or do we talk about you know, competition.

573
00:32:52,600 --> 00:32:57,000
Speaker 4: So the the issue is what would happen if Google

574
00:32:57,040 --> 00:33:00,000
Speaker 4: didn't have that and we saw, you know, the eye

575
00:33:00,000 --> 00:33:03,680
Speaker 4: coy of the Microsoft anti trust case was when when

576
00:33:03,800 --> 00:33:08,280
Speaker 4: Microsoft couldn't have Explorer I think it was at the time,

577
00:33:08,360 --> 00:33:12,680
Speaker 4: is this default browser in its entire suite? What happened? Well,

578
00:33:12,800 --> 00:33:16,120
Speaker 4: the ruling said you can't do that, and ironically, what

579
00:33:16,200 --> 00:33:18,880
Speaker 4: are to do? While it spawned Google? And Google has

580
00:33:18,960 --> 00:33:22,239
Speaker 4: now basically that anti trust spawned a new innovator. And

581
00:33:22,280 --> 00:33:24,600
Speaker 4: now we're going to say, okay, does his to repeat

582
00:33:24,680 --> 00:33:27,000
Speaker 4: if the same thing that happened with Microsoft' up mean,

583
00:33:27,000 --> 00:33:28,240
Speaker 4: now do we open it up? Well?

584
00:33:28,440 --> 00:33:31,960
Speaker 3: Yeah, it worked. That's why you didn't hear these hells

585
00:33:31,960 --> 00:33:34,560
Speaker 3: of outrage. I'm sure they were having those conversations with

586
00:33:34,640 --> 00:33:39,120
Speaker 3: Google account managers in Australia and New Zealand going hang on,

587
00:33:39,240 --> 00:33:41,640
Speaker 3: you know, this doesn't seem right to me. But ultimately

588
00:33:41,640 --> 00:33:44,840
Speaker 3: it was working, and it was working fundamentally better than

589
00:33:45,600 --> 00:33:48,240
Speaker 3: digital classified or whatever came before.

590
00:33:48,320 --> 00:33:50,680
Speaker 4: And that's where will there's that, yes, and the thing

591
00:33:50,680 --> 00:33:54,200
Speaker 4: about the reluctance by corporates to want to play this

592
00:33:54,280 --> 00:33:56,959
Speaker 4: and it's a it's a it's a very smart I mean,

593
00:33:57,160 --> 00:33:59,800
Speaker 4: they're embedded. Google is embedded right through the enterprise stack.

594
00:33:59,840 --> 00:34:04,320
Speaker 4: But in the end, the thing that corporates had from

595
00:34:04,400 --> 00:34:06,840
Speaker 4: Google that they'd never had before from any other media

596
00:34:06,920 --> 00:34:10,080
Speaker 4: company was a return path data flow, so they could

597
00:34:10,120 --> 00:34:15,000
Speaker 4: see who was buying. They could track someone responding to

598
00:34:15,040 --> 00:34:17,480
Speaker 4: a search ad. You could track it through to the advertiser,

599
00:34:17,560 --> 00:34:20,760
Speaker 4: could see it was coming through search. You could see

600
00:34:21,040 --> 00:34:24,560
Speaker 4: that when they did YouTube, there was search related results,

601
00:34:24,600 --> 00:34:27,120
Speaker 4: there was data on the profile, this profile on the individual.

602
00:34:27,320 --> 00:34:29,279
Speaker 4: So it was as much about the product as it

603
00:34:29,360 --> 00:34:31,399
Speaker 4: was about the data flow that these companies had never

604
00:34:31,440 --> 00:34:34,200
Speaker 4: had before. And that was very clever and it was

605
00:34:34,200 --> 00:34:36,680
Speaker 4: a very tech thing to do and very analytical. Now

606
00:34:37,400 --> 00:34:40,719
Speaker 4: now what we're seeing is great debate on the veracity

607
00:34:40,760 --> 00:34:43,960
Speaker 4: of that data, and it was essentially sometimes misleading. For instance,

608
00:34:44,000 --> 00:34:46,480
Speaker 4: there's a lot of people that go they buying search ads,

609
00:34:47,080 --> 00:34:50,840
Speaker 4: and companies don't need to buy search ads, for instance,

610
00:34:50,840 --> 00:34:53,400
Speaker 4: if you've got your organic plane because your brand holds up.

611
00:34:53,440 --> 00:34:55,640
Speaker 4: And in fact, when you take search ads out and

612
00:34:55,680 --> 00:34:59,400
Speaker 4: you don't have your paid search listing for whatever company

613
00:34:59,600 --> 00:35:02,200
Speaker 4: you find, nothing happens. There's been trials done, nothing changes

614
00:35:02,239 --> 00:35:04,680
Speaker 4: in the business, but such as the momentum that the

615
00:35:04,800 --> 00:35:08,759
Speaker 4: data flow across its network and across the ecosystem pulling in.

616
00:35:08,800 --> 00:35:11,080
Speaker 4: By the way, what we're now about to talk about,

617
00:35:11,120 --> 00:35:13,719
Speaker 4: which is the ad tech stack. Remember the ad tech

618
00:35:14,560 --> 00:35:18,960
Speaker 4: open web system allows Google to plant cookies on billions

619
00:35:18,960 --> 00:35:23,799
Speaker 4: of people and then have either aggregate the understanding and

620
00:35:23,880 --> 00:35:26,560
Speaker 4: behavior of those users and what their preferences and where

621
00:35:26,600 --> 00:35:29,120
Speaker 4: they're going, and build out models say to back to

622
00:35:29,160 --> 00:35:31,759
Speaker 4: the advertiser, we think, you know, there's x percentage of

623
00:35:31,760 --> 00:35:35,279
Speaker 4: people going to buy cars in July twenty twenty five

624
00:35:35,719 --> 00:35:38,400
Speaker 4: and bank on it. And of course all that data

625
00:35:38,520 --> 00:35:41,880
Speaker 4: collection helps. So it's a combination of things as to

626
00:35:42,080 --> 00:35:43,439
Speaker 4: and this is the thing. It's not black and white,

627
00:35:43,440 --> 00:35:46,200
Speaker 4: this one Peter, which makes it really tricky to discuss.

628
00:35:46,360 --> 00:35:49,919
Speaker 4: You know, the where's the smoking gun, Well, there's about

629
00:35:50,560 --> 00:35:51,439
Speaker 4: five hundred of them.

630
00:35:52,000 --> 00:35:56,400
Speaker 3: There are and been outlined. Yeah, it's quite confronting how

631
00:35:56,400 --> 00:35:59,520
Speaker 3: they've outlined it, but nevertheless, here we are. Now the

632
00:35:59,640 --> 00:36:01,440
Speaker 3: last thing we sort of want to do. I guess

633
00:36:01,920 --> 00:36:05,040
Speaker 3: a typical advertiser who's spending thirty to forty million dollars

634
00:36:05,040 --> 00:36:08,360
Speaker 3: a year on Google advertising, they don't want to break

635
00:36:08,440 --> 00:36:12,000
Speaker 3: what's working. But obviously they don't want to be paying

636
00:36:12,360 --> 00:36:15,480
Speaker 3: a levy to Google because they're the only player, you know,

637
00:36:15,640 --> 00:36:20,480
Speaker 3: sewing up the market. So you know what's ahead now. Obviously,

638
00:36:20,520 --> 00:36:23,239
Speaker 3: the you know, on the search monopoly case, they're looking

639
00:36:23,280 --> 00:36:28,040
Speaker 3: at remedies such as stopping that process of Google paying

640
00:36:28,160 --> 00:36:31,200
Speaker 3: billions of dollars to Apple and Samsung and others every year.

641
00:36:31,200 --> 00:36:33,960
Speaker 3: That's a relatively easy thing to do. But on this

642
00:36:34,040 --> 00:36:37,719
Speaker 3: digital advertising stuff, how do you keep what is great

643
00:36:37,719 --> 00:36:41,600
Speaker 3: about Google without breaking the whole thing up? Or you know,

644
00:36:41,840 --> 00:36:45,920
Speaker 3: is that potentially the only way to allow new competitors

645
00:36:45,920 --> 00:36:48,480
Speaker 3: to come in and actually give Google a run for

646
00:36:48,520 --> 00:36:49,960
Speaker 3: its money and digital advertising?

647
00:36:50,600 --> 00:36:54,360
Speaker 4: Yeah, this is the grand question really, And look, the

648
00:36:54,920 --> 00:36:57,560
Speaker 4: thing is that you you can do a few things,

649
00:36:58,040 --> 00:36:59,960
Speaker 4: and some of the things that they are talking about

650
00:37:00,080 --> 00:37:03,080
Speaker 4: is essentially, yeah, busting up the breaking up the ad

651
00:37:03,120 --> 00:37:05,080
Speaker 4: tech system. So it opens it up. So at least

652
00:37:05,080 --> 00:37:08,080
Speaker 4: you have, you know, an AD server that doesn't control

653
00:37:08,280 --> 00:37:10,920
Speaker 4: you know, ninety percent of where you know of the

654
00:37:10,960 --> 00:37:15,640
Speaker 4: mechanism of distributing ads in microseconds. So the ad server

655
00:37:16,760 --> 00:37:22,520
Speaker 4: is a massive development in that if Google is forced

656
00:37:22,719 --> 00:37:26,960
Speaker 4: to open up in some way that particular part of

657
00:37:26,960 --> 00:37:29,680
Speaker 4: the market, then it brings in other players. And look,

658
00:37:29,680 --> 00:37:32,040
Speaker 4: there's been three or four five others that are that

659
00:37:32,080 --> 00:37:35,120
Speaker 4: are that have gone by the wayside. You know, Microsoft

660
00:37:35,160 --> 00:37:36,959
Speaker 4: is still in there, but you know it's it's even

661
00:37:37,000 --> 00:37:39,399
Speaker 4: Microsoft and it's great. Mic is struggling on that front.

662
00:37:39,440 --> 00:37:42,000
Speaker 4: So yes, you break up, You break up the ad

663
00:37:42,040 --> 00:37:46,120
Speaker 4: tech stack and open it up. And essentially, you know,

664
00:37:46,239 --> 00:37:48,760
Speaker 4: people are talking about the trade deskers, which is probably

665
00:37:48,960 --> 00:37:53,480
Speaker 4: the great hope for a competition, and that's a sixty

666
00:37:53,480 --> 00:37:56,239
Speaker 4: billion dollar business, right, So breaking up the ad tech

667
00:37:56,280 --> 00:37:58,600
Speaker 4: stack is probably the full thing and first thing, and

668
00:37:58,760 --> 00:38:01,880
Speaker 4: breaking up probably the most important thing, Peter, if something's

669
00:38:01,920 --> 00:38:05,280
Speaker 4: going to happen, is breaking up the data data links.

670
00:38:05,320 --> 00:38:08,160
Speaker 4: So the unified profile or the unified view you can

671
00:38:08,160 --> 00:38:12,080
Speaker 4: get of a user across Google's entire system, whether it

672
00:38:12,120 --> 00:38:14,759
Speaker 4: be the open ad exchanges, which is other publishers or

673
00:38:14,800 --> 00:38:18,280
Speaker 4: its owned assets. That that flow is what's really powerful,

674
00:38:18,320 --> 00:38:22,040
Speaker 4: and it's what brings It's what enterprise level companies like

675
00:38:22,120 --> 00:38:24,600
Speaker 4: because they can data match their customers. They can get

676
00:38:24,640 --> 00:38:27,120
Speaker 4: a profile, they have it, they have a view on

677
00:38:27,200 --> 00:38:29,319
Speaker 4: their own customers, and they can match that with the

678
00:38:29,360 --> 00:38:33,719
Speaker 4: Google profile and target accordingly and sort of really get

679
00:38:33,719 --> 00:38:37,320
Speaker 4: into some pretty interesting stuff that gives them, in the

680
00:38:37,440 --> 00:38:40,960
Speaker 4: end a data a feedback flow on what's happening with

681
00:38:41,040 --> 00:38:43,680
Speaker 4: their ads, which of course a lot of traditional media

682
00:38:43,719 --> 00:38:44,080
Speaker 4: can't do.

683
00:38:44,560 --> 00:38:48,480
Speaker 3: Yeah, so this really is sort of an existential threat

684
00:38:48,520 --> 00:38:51,760
Speaker 3: to Google as we know it and as it currently

685
00:38:51,800 --> 00:38:55,960
Speaker 3: makes profit from digital advertising. I mean, all of this,

686
00:38:55,960 --> 00:38:58,640
Speaker 3: this is three fronts and there are more cases to come.

687
00:38:59,280 --> 00:39:01,920
Speaker 3: Something is going to change here that ultimately is going

688
00:39:02,000 --> 00:39:04,160
Speaker 3: to hit their bottom line, is going to make them

689
00:39:04,320 --> 00:39:08,120
Speaker 3: potentially a less profitable company with you know, less of

690
00:39:08,160 --> 00:39:10,200
Speaker 3: a market share and digital advertising.

691
00:39:10,840 --> 00:39:14,480
Speaker 4: Well, we haven't even touched on its move from Google

692
00:39:14,520 --> 00:39:19,480
Speaker 4: Search to AI overviews and the AI based search function

693
00:39:19,560 --> 00:39:21,279
Speaker 4: that it's trying to build in with its with its

694
00:39:22,120 --> 00:39:27,640
Speaker 4: with its tech. But that's that's another you know, subject altogether.

695
00:39:27,719 --> 00:39:30,680
Speaker 4: But yeah, no, it is, it is. If this stuff happens,

696
00:39:30,719 --> 00:39:35,200
Speaker 4: there is there is enormous change for both Google and

697
00:39:35,280 --> 00:39:40,360
Speaker 4: the market itself because we are essentially we're cows of

698
00:39:40,400 --> 00:39:42,520
Speaker 4: the dairy farm and we're in a routine and we're

699
00:39:42,560 --> 00:39:43,880
Speaker 4: used to We've got a whole bunch of systems, a

700
00:39:43,920 --> 00:39:46,560
Speaker 4: whole bunch of people. There's whole bunch of capabilities, people

701
00:39:46,560 --> 00:39:48,799
Speaker 4: that have been trained and working through how to use

702
00:39:48,840 --> 00:39:51,560
Speaker 4: the back end of the entire Google stack. So it's

703
00:39:51,600 --> 00:39:56,239
Speaker 4: a massive, massive industry. What what some say is likely

704
00:39:56,280 --> 00:39:59,040
Speaker 4: to happen out of this is that you've heard it before,

705
00:39:59,080 --> 00:40:02,760
Speaker 4: but it may open up again the well. It forces

706
00:40:02,800 --> 00:40:07,080
Speaker 4: the understanding that there is premium inventory, there is premium publishers,

707
00:40:07,120 --> 00:40:09,640
Speaker 4: and there is the long tail of publishers. They do

708
00:40:09,719 --> 00:40:12,279
Speaker 4: a different role, they have a different purpose, and there

709
00:40:12,320 --> 00:40:17,080
Speaker 4: may be that, you know, premium audiences or professionally curated

710
00:40:17,520 --> 00:40:20,040
Speaker 4: news content, and beyond news, it's just not about news either,

711
00:40:20,040 --> 00:40:23,360
Speaker 4: by the way, but they there's a bet that's a

712
00:40:23,360 --> 00:40:27,960
Speaker 4: better buy. And so there will be alternatives and new

713
00:40:28,000 --> 00:40:30,719
Speaker 4: exchanges and new ways to buy this stuff which won't

714
00:40:30,760 --> 00:40:35,080
Speaker 4: be as consolidated globally into one platform, and it breaks down,

715
00:40:35,160 --> 00:40:39,319
Speaker 4: makes it more complex. But that's essentially where I think,

716
00:40:40,280 --> 00:40:42,080
Speaker 4: you know, if we're thinking about things, that's where we've

717
00:40:42,080 --> 00:40:42,400
Speaker 4: got to go.

718
00:40:42,640 --> 00:40:47,759
Speaker 3: Yeah. Hey, just finally, Paul, you mentioned briefly cookies. You know,

719
00:40:48,080 --> 00:40:52,839
Speaker 3: we're leading up to, in Google's world, removal of doing

720
00:40:52,840 --> 00:40:56,359
Speaker 3: the way with cookies. Essentially they're backtracked on that. How

721
00:40:56,400 --> 00:41:01,040
Speaker 3: important does this cookies are essential to tracking people across

722
00:41:01,080 --> 00:41:04,480
Speaker 3: the web? Is does does that decision tie into do

723
00:41:04,560 --> 00:41:06,400
Speaker 3: you think they put the brakes on that? To see

724
00:41:06,560 --> 00:41:07,640
Speaker 3: where this all ends up?

725
00:41:09,760 --> 00:41:14,920
Speaker 4: So basically, the the the cookie thing is about perhaps

726
00:41:15,120 --> 00:41:17,200
Speaker 4: jazz hands a bit of a distraction whilst they build

727
00:41:17,239 --> 00:41:20,440
Speaker 4: out their owned media play and even you know, prepare

728
00:41:20,480 --> 00:41:22,840
Speaker 4: preparing the some that even in this court case I

729
00:41:22,880 --> 00:41:27,799
Speaker 4: was at in in Virginia last week last month, they

730
00:41:27,840 --> 00:41:31,000
Speaker 4: believe that, you know, essentially, even this court cake Googles,

731
00:41:31,040 --> 00:41:33,560
Speaker 4: Google's sort of quietly nodding to the fact that they've

732
00:41:33,640 --> 00:41:37,680
Speaker 4: they've lost this one. And where they're reframing to rewiring

733
00:41:37,719 --> 00:41:40,799
Speaker 4: to is owned assets only so that gets them out

734
00:41:40,840 --> 00:41:44,000
Speaker 4: of the out of the sort of the broader monopolistic behavior.

735
00:41:44,080 --> 00:41:46,719
Speaker 3: So well, I mean, it makes sense, and you know

736
00:41:46,760 --> 00:41:51,440
Speaker 3: it'd be much you know, straightforward business. You've got YouTube there,

737
00:41:51,440 --> 00:41:54,240
Speaker 3: which which is massive. You've got search. You've got Gmail

738
00:41:55,040 --> 00:41:57,840
Speaker 3: and all these other things you know, draw people to

739
00:41:57,960 --> 00:42:00,960
Speaker 3: those advertised to them around there. Do you really need

740
00:42:01,040 --> 00:42:04,520
Speaker 3: to control everything else? Well, they're not going to be

741
00:42:04,560 --> 00:42:06,239
Speaker 3: able to, that's right.

742
00:42:06,280 --> 00:42:08,719
Speaker 4: Well, I mean, in the end, the cookie thing is

743
00:42:10,239 --> 00:42:13,000
Speaker 4: the market has now moved is prepping beyond cookies now anyway,

744
00:42:13,040 --> 00:42:14,960
Speaker 4: it's all going to first party data. They realized that

745
00:42:14,960 --> 00:42:17,120
Speaker 4: the writings on the wolf of that, whether Google slows

746
00:42:17,120 --> 00:42:20,040
Speaker 4: it down, there is convenience for the for the bean

747
00:42:20,080 --> 00:42:22,440
Speaker 4: counters about how much they spend in prep in the

748
00:42:22,440 --> 00:42:24,880
Speaker 4: next twelve months and two years, So there's order of

749
00:42:25,040 --> 00:42:26,920
Speaker 4: you know that there's the quantum of budget and an

750
00:42:26,920 --> 00:42:30,080
Speaker 4: investment to prep. But I think most most companies now

751
00:42:30,200 --> 00:42:33,359
Speaker 4: are okay, first party's, first party data, own your own

752
00:42:33,400 --> 00:42:37,560
Speaker 4: customer data and know that is the priority now versus

753
00:42:37,680 --> 00:42:40,720
Speaker 4: you know, the third the third party cooking tracking, third

754
00:42:40,719 --> 00:42:43,960
Speaker 4: party tracking. They're still doing it, by the way, it's

755
00:42:43,960 --> 00:42:46,560
Speaker 4: still massive. It's not like it's finished now, but they

756
00:42:46,640 --> 00:42:50,359
Speaker 4: are now in the transition over to first party. Wow.

757
00:42:50,560 --> 00:42:53,239
Speaker 3: Okay, mate, Well, thank you so much for taking us

758
00:42:53,280 --> 00:42:56,560
Speaker 3: through all of that. A lot of great coverage in

759
00:42:56,560 --> 00:42:59,520
Speaker 3: recent years on m I three about this whole issue,

760
00:42:59,560 --> 00:43:02,000
Speaker 3: So we'll linked to a lot of those stories and

761
00:43:02,239 --> 00:43:04,120
Speaker 3: twenty twenty five is going to be massive. Love to

762
00:43:04,120 --> 00:43:06,680
Speaker 3: come back to you once this is all shaken out

763
00:43:06,680 --> 00:43:08,520
Speaker 3: and we actually see what it's going to mean for

764
00:43:08,600 --> 00:43:11,120
Speaker 3: Google and for advertisers in this part of the world.

765
00:43:11,480 --> 00:43:14,120
Speaker 4: Yeah, thanks, Peter, It really got conversation.

766
00:43:14,239 --> 00:43:14,640
Speaker 2: Enjoyed it.

767
00:43:17,080 --> 00:43:18,960
Speaker 3: So that was poorly Obviously knows a heck of a

768
00:43:19,000 --> 00:43:22,359
Speaker 3: lot about all of this stuff. And he pointed out

769
00:43:22,520 --> 00:43:25,719
Speaker 3: after the interview we didn't mention it there, but an

770
00:43:25,760 --> 00:43:29,520
Speaker 3: important point, which was that Google called the star witness

771
00:43:29,719 --> 00:43:32,880
Speaker 3: in its case. In its defense, a guy called Mike Israel,

772
00:43:33,120 --> 00:43:37,080
Speaker 3: an economist who pointed out that the government was too

773
00:43:37,200 --> 00:43:42,760
Speaker 3: narrowly defining advertising technology. It was mainly looking at open

774
00:43:42,840 --> 00:43:46,360
Speaker 3: web display advertising, the sort of rectangular ads that appear

775
00:43:47,320 --> 00:43:50,360
Speaker 3: along the top and down the side of a web page.

776
00:43:50,880 --> 00:43:52,880
Speaker 3: So he was trying to argue that, you know, if

777
00:43:52,960 --> 00:43:56,920
Speaker 3: you look much more broadly at digital advertising, Google doesn't

778
00:43:56,960 --> 00:43:59,839
Speaker 3: have as much of a monopoly as being made out.

779
00:44:00,680 --> 00:44:03,319
Speaker 3: So I don't think that has held much sway that

780
00:44:03,440 --> 00:44:07,160
Speaker 3: argument so far, but we will see obviously how this

781
00:44:07,400 --> 00:44:11,120
Speaker 3: case plays out. It's still pending a decision, and he

782
00:44:11,160 --> 00:44:14,600
Speaker 3: also said, you know, there's some great sources if you

783
00:44:14,640 --> 00:44:17,239
Speaker 3: want to follow this as it sort of goes into

784
00:44:17,280 --> 00:44:19,920
Speaker 3: the next crucial phase, as at least two of these

785
00:44:19,960 --> 00:44:22,759
Speaker 3: cases have had decisions go against Google, so now they'll

786
00:44:22,800 --> 00:44:24,440
Speaker 3: be looking at remedy, and then we have the digital

787
00:44:24,520 --> 00:44:27,520
Speaker 3: advertising one. We will get a decision early next year.

788
00:44:28,080 --> 00:44:33,080
Speaker 3: Ricky Sutton from Future Media has a fantastic substack all

789
00:44:33,120 --> 00:44:36,160
Speaker 3: about this and Ariel go see it at watch my ads.

790
00:44:36,200 --> 00:44:39,520
Speaker 3: Those are to go to sources poor recommends if you're

791
00:44:39,520 --> 00:44:41,480
Speaker 3: really interested in following this stuff.

792
00:44:41,760 --> 00:44:46,520
Speaker 2: Yeah. The pressure on Google to show what's up at sleeves, really,

793
00:44:46,560 --> 00:44:49,359
Speaker 2: I think is the most interesting part of all of this,

794
00:44:49,520 --> 00:44:52,160
Speaker 2: because quite a lot of the time, these big tech companies,

795
00:44:52,200 --> 00:44:56,520
Speaker 2: when they are taking on a massive industry as a

796
00:44:56,920 --> 00:45:00,960
Speaker 2: huge kind of monopolistic force, they aren't really asked to

797
00:45:01,040 --> 00:45:04,719
Speaker 2: actually prove it, prove that they're not. Especially since the

798
00:45:04,760 --> 00:45:09,080
Speaker 2: kind of landmark late eighties early nineties cases against Microsoft

799
00:45:09,960 --> 00:45:11,640
Speaker 2: that set the standard for a lot of how we

800
00:45:11,760 --> 00:45:14,920
Speaker 2: view tech monopolies today. I think it's a really good

801
00:45:15,000 --> 00:45:18,440
Speaker 2: thing that we're getting some pressure to on Google to

802
00:45:18,520 --> 00:45:22,520
Speaker 2: actually show some evidence, you know, not just not just saying,

803
00:45:22,560 --> 00:45:24,759
Speaker 2: oh you know, this is what tech companies do, and

804
00:45:25,760 --> 00:45:29,080
Speaker 2: treating them like the omnipotent creatures that we have of

805
00:45:29,120 --> 00:45:33,080
Speaker 2: the past, as McIntyre mentioned, So yeah, that's that's really

806
00:45:33,120 --> 00:45:36,480
Speaker 2: my biggest takeaway is that regardless of the outcome, it's

807
00:45:36,520 --> 00:45:39,000
Speaker 2: a good thing. We're getting more scrutiny and long may

808
00:45:39,040 --> 00:45:39,480
Speaker 2: it last.

809
00:45:40,200 --> 00:45:44,080
Speaker 3: Yeah, And you know, as Paul pointed out, for years,

810
00:45:44,520 --> 00:45:49,000
Speaker 3: there wasn't that scrutiny. We became enamored with all the

811
00:45:49,040 --> 00:45:52,359
Speaker 3: great new, shiny things these big companies were delivering to us,

812
00:45:52,360 --> 00:45:56,680
Speaker 3: and Google in particularly the utility it gives us through Gmail,

813
00:45:56,800 --> 00:46:00,680
Speaker 3: you know, free to use product, Google Maps, Google Earth,

814
00:46:01,040 --> 00:46:03,359
Speaker 3: the Google Search Engine, all of this stuff. It has

815
00:46:03,440 --> 00:46:08,200
Speaker 3: made our lives better. There's no disputing that. But in

816
00:46:08,239 --> 00:46:10,280
Speaker 3: the back room there are a lot of clever people

817
00:46:10,320 --> 00:46:12,560
Speaker 3: going this is going exactly the way we wanted. The

818
00:46:12,600 --> 00:46:16,160
Speaker 3: government is ignoring us. It's going after more traditional sorts

819
00:46:16,160 --> 00:46:18,600
Speaker 3: of companies when it comes to antitrust. All the while

820
00:46:19,239 --> 00:46:22,520
Speaker 3: we are connecting together and all of these services buying

821
00:46:22,600 --> 00:46:25,719
Speaker 3: up the companies that are helping shore up what is

822
00:46:25,719 --> 00:46:30,160
Speaker 3: becoming a monopoly. So very shrewd move and really exploiting

823
00:46:30,239 --> 00:46:33,479
Speaker 3: that sort of innovation is everything, don't slow this down

824
00:46:34,160 --> 00:46:37,360
Speaker 3: sort of mentality that laisse faire approach from the government

825
00:46:37,400 --> 00:46:42,360
Speaker 3: that is very quickly now changing under the Biden administration

826
00:46:42,760 --> 00:46:45,399
Speaker 3: and in the EU and other places around the world.

827
00:46:45,680 --> 00:46:48,319
Speaker 3: It's coming at a time when a new disruptive force,

828
00:46:48,440 --> 00:46:53,000
Speaker 3: AI is arriving, which is probably the biggest prospect really

829
00:46:53,160 --> 00:46:57,680
Speaker 3: off Google's monopoly being cracked open. There will be and

830
00:46:57,719 --> 00:46:59,080
Speaker 3: I don't know how they're going to do it. This

831
00:46:59,120 --> 00:47:02,040
Speaker 3: is going to be so compli to try and break

832
00:47:02,120 --> 00:47:05,480
Speaker 3: up Google or disclose all the inner workings off its

833
00:47:05,480 --> 00:47:08,439
Speaker 3: so everyone's on a level playing field. But then you've

834
00:47:08,440 --> 00:47:12,400
Speaker 3: got all of these competitors coming in using AI to

835
00:47:12,480 --> 00:47:16,200
Speaker 3: do smart things and search and advertising and everything. So

836
00:47:16,840 --> 00:47:19,400
Speaker 3: it sort of seems like we got you know, the

837
00:47:19,440 --> 00:47:21,560
Speaker 3: horse has bolted a little bit. Even though this is

838
00:47:21,600 --> 00:47:24,440
Speaker 3: still a major player. How much longer is it going

839
00:47:24,440 --> 00:47:28,200
Speaker 3: to be a major player in some of these categories

840
00:47:28,280 --> 00:47:32,600
Speaker 3: because AI is changing everything. Yeah, anyway, that's it for

841
00:47:32,719 --> 00:47:35,960
Speaker 3: this week. Thanks so much to Paul McIntyre for taking

842
00:47:36,000 --> 00:47:39,040
Speaker 3: us through the complex cases Google is involved and lots

843
00:47:39,040 --> 00:47:41,279
Speaker 3: of background reading in the show notes, which you can

844
00:47:41,280 --> 00:47:44,319
Speaker 3: find in the podcast section at Business deesk dot Co.

845
00:47:44,520 --> 00:47:44,960
Speaker 4: Dot nz.

846
00:47:45,360 --> 00:47:47,880
Speaker 2: The Business Attach can be streamed there. You can stream

847
00:47:47,880 --> 00:47:50,440
Speaker 2: it from iHeartRadio. You can stream it from your podcast

848
00:47:50,480 --> 00:47:54,480
Speaker 2: platform of choice, especially if you're subscribed, so do that

849
00:47:54,600 --> 00:47:56,600
Speaker 2: and get in touch with your feedback. We'd love to

850
00:47:56,600 --> 00:47:59,320
Speaker 2: hear your suggestions for upcoming guests too. You can email

851
00:47:59,360 --> 00:48:02,839
Speaker 2: me Ben at Business Desk dot co, dot and z, and.

852
00:48:02,760 --> 00:48:04,960
Speaker 3: You can find both of us on x and LinkedIn,

853
00:48:05,000 --> 00:48:07,200
Speaker 3: where you can follow the Business of Tech page for

854
00:48:07,360 --> 00:48:08,360
Speaker 3: all of our updates.

855
00:48:08,560 --> 00:48:09,600
Speaker 2: That's it for another week.

856
00:48:09,760 --> 00:48:12,640
Speaker 3: Catch you next time for another episode of the Business

857
00:48:12,719 --> 00:48:13,080
Speaker 3: of Tech.