WEBVTT

1
00:00:02.720 --> 00:00:13.720
It's, looks much worse in this environment than it might at Tesla, for example, because this is an existing social network where breaking eggs means, you know, literal people and companies get damaged.

2
00:00:13.800 --> 00:00:23.540
And so somewhere in between those kind of five explanations for what's going on is, uh, is the truth, I think. Or there might just be- Well, what's your, what's your vote? What is your...

3
00:00:23.560 --> 00:00:30.580
First of all, what one, and then but what did, what did you vote for? Well, the poll is still going on. Um- Oh my God, Ari, this is going up on Tuesday.

4
00:00:30.590 --> 00:00:33.020
[laughs] I don't wanna, like, muddy the waters in your Twitter poll.

5
00:00:33.140 --> 00:00:50.920
[upbeat music] Welcome to the Rebooting show.

6
00:00:50.960 --> 00:01:01.240
I'm Brian Morrissey. This is a show where I speak to people operating media businesses to get under the hood of their business models. If you like this show, and I hope you do, please check out another podcast I'm doing.

7
00:01:01.280 --> 00:01:10.900
This one is with former Hearst Magazines president, Troy Young. It's called The People vs. Algorithms podcast, where we take a bit of a broader view of the patterns emerging in media, tech, and culture.

8
00:01:11.500 --> 00:01:18.540
This week we spent a fair bit of time dissecting why charismatic billionaires end up getting things very, very disastrously wrong.

9
00:01:18.620 --> 00:01:29.300
As we've seen in the cases of Sam Bankman-Fried and Elon Musk, very chaotic tenure owning Twitter over the past few weeks. You can find it on Apple or s- Spotify or wherever you get your podcasts.

10
00:01:29.540 --> 00:01:40.520
Um, it's a good show, so please do check it out. This week on the Rebooting show, I spoke to Ari Paparo. Ari is a longtime ad tech exec, and you know if someone is an OG in this space when they go back to DoubleClick.

11
00:01:40.840 --> 00:01:52.700
He's got a very deep understanding of this fast-changing sector. I've relied on Ari many times over the years as I've been lost in a sea of acronyms, and he m- mostly kept out of the acronym soup during our discussion.

12
00:01:53.190 --> 00:01:58.600
But we had a talk about the mess at Twitter and why the advertisers most likely are not coming back anytime soon.

13
00:01:58.640 --> 00:02:08.479
And then we got into what the future of ad tech looks like in an era where targeting specific audiences is getting harder and more expensive. I hope you enjoy this episode, and please send me your feedback.

14
00:02:08.660 --> 00:02:20.560
My email is bmorrissey@gmail.com. Thanks to Jay Sparks of Podhelpus for producing this podcast. If you have any podcasting needs or are thinking about making one, please do get in touch with Jay.

15
00:02:20.620 --> 00:02:32.310
You can find out, uh, more about his services at podhelp.us. [upbeat music] All right. Uh- I wanna have you on to talk about, like, two main things. W- Uh-huh. And, and they're both pressing to the future of- Yeah...

16
00:02:32.340 --> 00:02:38.660
humanity. One is what the hell is going on at Twitter, 'cause it's everyone's favorite topic, and like- Yeah, yeah... you just have to talk about it.

17
00:02:38.980 --> 00:02:45.100
And then we're gonna talk about what the hell is going on with ad tech, and we're gonna try to tie them together. That sounds- So first, like, you're one of my...

18
00:02:45.420 --> 00:02:53.170
You're, y- I think you're, you're my favorite, like, ad tech related Twitter follow. Okay. Not much of a competition- [laughs]... but, uh, I'm doing my best there.

19
00:02:53.320 --> 00:03:02.040
Well, that's like, I, like, joke, like, I, I sometimes joke, probably not to you, but, like, that you're the funniest person in ad tech, but it's not, like, not that much of a compliment. [laughs] Yeah, it's true.

20
00:03:02.100 --> 00:03:09.180
Oh, you know, I, sometimes people, like, stop me on the street and be like, "I love you on Twitter." And my wife is like, "What the fuck just happened?" Like, "Are you famous or something?"

21
00:03:09.820 --> 00:03:15.580
[laughs] I could assure her, no, you are not. I am famous in a very small group that happens to live within a- A very small group...

22
00:03:15.720 --> 00:03:25.120
within a mile- But for those who do not know, walk down, because you have been, like, an ad tech aficionado going back to, like, the original... Not the original, but the old school days.

23
00:03:25.220 --> 00:03:36.700
Bef- well, back when it was just, like, you know, juiced up ad networks and whatnot. Yeah. I started in ad tech in 2004, so I wasn't in the first wave, like the dotcom wave and the DoubleClick craziness.

24
00:03:37.000 --> 00:03:44.140
I started at DoubleClick in 2004 when it was already, like, past its prime. And, and I got- Oh my God, DoubleClick was past its prime in 2004?

25
00:03:44.640 --> 00:03:56.180
In 2004, DoubleClick was a incredibly poorly performing public company filled with people who didn't wanna work there and customers who hated them. Uh, it was a really messed up company.

26
00:03:56.320 --> 00:04:04.780
Uh, and I was lucky enough to join right when they were taken private by a private equity firm. Oh, yeah. Uh, and then cleaned up and fixed.

27
00:04:04.900 --> 00:04:11.700
So I got lucky t- Yeah, they were, they were tarted up, and then they just got the timing totally right with Google. [laughs] Tarted up. They were tarted up really significantly.

28
00:04:11.780 --> 00:04:25.480
The, uh, uh, DoubleClick went from being sold off the public markets for 300 million to being sold to Google less than two years later for 3 billion. Yeah. So 10X return in less than two years, that's quite a tarting.

29
00:04:26.100 --> 00:04:33.090
Um- And then a, and then Aquent the, the, like- [laughs]... it's still possibly one of the worst deals that, like, took place in this industry.

30
00:04:33.110 --> 00:04:45.130
Aquent is competing with Elon Musk, I think, for the worst use of capital, uh, in the history of digital media. Uh, yeah, so I got lucky twice though in terms of how I got to where I am, uh, or maybe three times.

31
00:04:45.160 --> 00:04:52.380
The first was I got in DoubleClick at a perfect time when it was ready to be fixed. Second was my job at DoubleClick was really interesting.

32
00:04:52.450 --> 00:04:59.460
I was kind of the product manager for miscellaneous, so whenever anyone had some crazy idea that was not a banner ad, they would come to me.

33
00:04:59.980 --> 00:05:10.860
So that was video, that was mobile, that was billboards, that was, like, anything. Like, all the strange entrepreneurs called me, uh, and tried to get me to integrate with them. So that was a really fun job.

34
00:05:10.880 --> 00:05:17.850
And I guess the third thing I got lucky on was that Business Insider, I became friends with Business Insider 'cause I was, like, the only person at Google who would ever talk to them.

35
00:05:17.860 --> 00:05:19.280
Of course, I wouldn't use my name, but, you know.

36
00:05:19.780 --> 00:05:29.200
Uh, and [laughs] so, uh, they started writing articles that, like, name checked me in weird ways like, "This is a low level guy at Google you should know," and stuff like that. And- What a compliment.

37
00:05:29.230 --> 00:05:37.250
[laughs] You should put that on your LinkedIn. And then one day they posted... Mind you, this was back when Business Insider was really popular in 2000 and beyond. This sounds like when Nick was still writing.

38
00:05:37.300 --> 00:05:39.120
Was this Nick Carlson? Nick. It was Nick. Uh, yeah, it was Nick.

39
00:05:39.320 --> 00:05:47.350
And, uh, [laughs] and like one day they, they were like, "The 100 Twitter accounts you must follow if you're in advertising," and I was, like, ahead of Rupert Murdoch.

40
00:05:47.500 --> 00:05:52.540
I was, like, number 18 on their list of the 100 Twitter fol- accounts you must follow.

41
00:05:52.620 --> 00:06:02.690
And I went from being, like, having 50 Twitter followers and tweeting about my lunch to suddenly being an influencer, uh, and the rest is history. That is back when, like, Twitter was, like, the...

42
00:06:02.740 --> 00:06:12.052
Original Twitter is the one that I miss when it was about, like-Tweeting about your lunch and just doing little one-liners and, like, I've never evolved o- on Twitter. Um- Yeah.

43
00:06:12.612 --> 00:06:20.972
And so that's why- I remember te- texting my tweets on my BlackBerry- Yeah... to like the, the phone number. Yeah. That was how I started using Twitter. So I haven't evolved at all.

44
00:06:21.052 --> 00:06:30.332
I've, I'm tweeting the same thing I did, like, back then, and, uh, what I found is it's, it's, uh, it's very niche. It's, it's [laughs] there's not a big market. I've capped out at the market.

45
00:06:30.372 --> 00:06:41.622
The number of people who will follow me on Twitter is, is, is finite, I found out. How many followers do you have? Like, I don't know, like 40, 41, 42,000 or something like this.

46
00:06:41.692 --> 00:06:45.721
Are you more of a- I don't know how many are real. Are you more of a TikTok guy? No, I don't do TikTok. Yeah.

47
00:06:45.732 --> 00:06:54.052
Twitter's the only social network I, um, act- well, I use Instagram a little bit, but just- But aren't you like a media celebrity? I mean, I think this should be part of your portfolio.

48
00:06:54.492 --> 00:07:04.332
You're underperforming if you don't have a Twitter presence. Well, let's talk about that because, like, I don't, I don't really... I think Twitter's kinda silly, to be honest with you. And- Sure...

49
00:07:04.372 --> 00:07:18.092
the, the fact that it is, it is getting as much, like, attention a- as it is sort of amuses me to, to a degree, and... 'Cause nobody, at the end of the week, Ari, nobody, nobody says, "You know what? It was a good week.

50
00:07:18.352 --> 00:07:27.452
I got a lot done," if Elon Musk asks, "But I really wish I, I spent more time on Twitter." [laughs] I agree with you. Nobody. No.

51
00:07:27.702 --> 00:07:35.172
It, it's a, it's a- It's like someone saying like, "You know, it was a good night last night, but I wish I had more menthol cigarettes." [laughs] That's exactly what it's like.

52
00:07:35.212 --> 00:07:47.692
It's like the, you enjoy that sweet, minty pull of nicotine- [laughs]... and you go one after another, and you really do regret it the next day. But then the next evening they're out again.

53
00:07:47.872 --> 00:07:56.512
So, uh, I don't know what to tell you. I'm totally addicted to Twitter, and I hope Elon doesn't destroy my only- Okay. But you, you-... joy in life. You, you had a, you had...

54
00:07:56.532 --> 00:08:05.232
Th- and this is the Twitter segment, so you had, like, a little, uh, tweet thread today. See, you've evolved your Twitter use. There, there wasn't... I refuse to acknowledge tweet threads exist.

55
00:08:05.812 --> 00:08:13.352
And you're trying to unpack, like, you know, the, I think the big question is, what the hell is this guy doing?

56
00:08:13.432 --> 00:08:25.852
I mean, he's, like, the most successful entrepreneur of a generation, multiple generations, I don't even know. And, like, at least from the outside, from mere mortal, I'm not a billionaire, I don't know.

57
00:08:25.932 --> 00:08:32.732
I don't put rockets into space. Doesn't seem like he knows what the hell he's doing. Yeah. It's interesting. I love situations like this.

58
00:08:32.812 --> 00:08:41.542
I love trying to understand people's psychology and maybe try to empathize a little bit with where they're coming from. And this is absolutely fascinating because- Yeah...

59
00:08:41.612 --> 00:08:50.772
it's the richest person on earth, like, quite literally, someone who has made multiple companies hugely successful, and he's not just, you know, tripping a little bit.

60
00:08:51.132 --> 00:09:04.012
He's systematically mismanaging a situation to a degree that is, uh, it seems like it must be on purpose almost. It's like performance artist Andy Kaufman as CEO. It's just extraordinary.

61
00:09:04.452 --> 00:09:14.392
And, and the, the really ha- you have to ask the question why. You can't just sit there and just say, "Oh, he's just a bad manager," 'cause that doesn't line up with the other pieces of information we have.

62
00:09:14.452 --> 00:09:26.252
Maybe he's a very aggressive manager. Okay, fine, but that doesn't explain, uh, why he tweets antagonistic things towards his advertisers. Maybe he's a right-wing, you know, incel, I don't know. Maybe.

63
00:09:26.532 --> 00:09:32.892
But, like, [laughs] uh, the entire pack- He has like 11 kids, Ari. I don't think he's an incel. [laughs] He does. Yeah. Incel is not the word I was looking for.

64
00:09:32.952 --> 00:09:40.592
What's the, what's the right word for, like, the right-wing simp? I don't, I don't know what, what term would be appropriate. A red pillar. He's a red pillar. Oh, okay. That's what I was looking for, not an incel.

65
00:09:40.672 --> 00:09:48.951
So, uh, yeah, there's, like, a lot of crosscurrents here that, you know, getting to the bottom of it is just absolutely fascinating, 'cause there's obviously something we don't know.

66
00:09:49.222 --> 00:10:01.652
There's some, like, you know, abandoned, you know, uh, sled in his attic that, you know, says tweet on it or something. Like, there's something going on here that i- g- is beyond the surface.

67
00:10:02.072 --> 00:10:07.372
You know, uh, if you ask some people, and I put this in my tweet storm, like, he's doing this for Saudi Arabia. That's the crazy people.

68
00:10:07.452 --> 00:10:17.472
They're like, you know, Saudi Arabia gave him the money so he could destroy Twitter so that he could bring, you know, right-wing petrol fascism to the West. So wait, I vote- And you can unpack that for me.

69
00:10:17.492 --> 00:10:25.301
[laughs] I vo- I voted in this, I voted in this, in this poll, and I didn't, I didn't get whatever the most popular answer was. But, like- Okay... give us the options that you came down to- Okay...

70
00:10:25.301 --> 00:10:34.592
about what the hell is going on. Yeah. So these are not really mutually exclusive, but let's just go through them fast. So option one is sort of Occam's razor, which is he's just bad at managing stuff.

71
00:10:34.632 --> 00:10:38.452
That is the easiest explanation for why he's screwed this up so bad. Okay.

72
00:10:38.552 --> 00:10:47.392
Option two is what I call the producers, and that is this idea that he's intentionally running it into the ground, just like in the play, in the movie The Producers.

73
00:10:47.792 --> 00:11:03.192
But in this case, the reason he's doing it is because he wants to effectively force the debt holders to capitulate and to realize their debt's worthless, and then he buys it back for pennies on the dollar, and thereby recapitalizes the company and gets out of the terrible situation.

74
00:11:03.312 --> 00:11:07.432
Option three is what I affectionately call Leroy Jenkins.

75
00:11:07.552 --> 00:11:16.712
Leroy Jenkins, of those you, who you don't know, is a meme where, you know, people just throw all caution in the wind and jump into a fray against best, better wishes.

76
00:11:17.352 --> 00:11:27.992
And so this is the idea that maybe he is so, like, bitter about having to buy Twitter. You know, he lost in court. He's forced to buy it. He knows it's a bad idea, that he's like, "You know, F it.

77
00:11:28.072 --> 00:11:36.811
I'm just gonna do whatever the hell I want and have some fun." The next one is what I call the conspiracy theory. So that's the one that either China or Saudi Arabia is forcing him to do this for some reason.

78
00:11:36.892 --> 00:11:45.012
I mean, this is like crazy pants tinfoil, but, like, maybe this- Ooh, I like this one. It's a pretty good... Like, you know, he took money from Binance. He took money from the Saudi prince.

79
00:11:45.612 --> 00:11:56.072
He's doing this for no good reason, and he keeps talking about the elites and the leftists and that sort of thing that makes no sense. So maybe there is a political aspect to this somehow.

80
00:11:56.152 --> 00:12:03.272
I mean, I, I don't have any evidence of that, but, you know, we don't have any evidence for him doing anything in a rational way, so who knows?

81
00:12:03.372 --> 00:12:18.244
Um, and then the last one, which I think is, like, the apologist one is called, what I call breaking eggs, which is-That his approach to product management is to just go as fast as possible and just to absolutely not care about collateral damage.

82
00:12:18.344 --> 00:12:35.144
And he believes that by writing code and launching things, he'll get to where he wants to go, and it's, looks much worse in this environment than it might at Tesla, for example, because this is an existing social network where breaking eggs means, you know, literal people and companies get damaged.

83
00:12:35.204 --> 00:12:42.504
And so somewhere in between those kind of five explanations for what's going on is, uh, is the truth, I think.

84
00:12:42.644 --> 00:13:00.044
[gentle music] Or there might just be- What's your, what's your vote? What is your... First of all, what won, and then, but what did, what did you vote for? Well, the poll is still going on. Um- Oh my God.

85
00:13:00.184 --> 00:13:06.684
Ari, this is going up on Tuesday. It's fine. [laughs] But I don't, I don't, I don't wanna, like, muddy the waters in your Twitter poll. Okay.

86
00:13:06.784 --> 00:13:19.804
Well, well basically, the, the consensus is sort of the last one, breaking eggs, which is... Well, and combined with sort of incompetence. Like, he's just a, he's a bad people manager. He's a bad manager in general.

87
00:13:19.984 --> 00:13:24.444
He's a brilliant person who's a bad manager, and his idea is just to...

88
00:13:24.584 --> 00:13:31.184
He has a vision for this thing, which is much less advertising and much more payments, much more verification, and he's gonna get there the hard way.

89
00:13:31.594 --> 00:13:43.704
And the hard way is going to hurt a lot of people and a lot of things, and it may also hurt him because he may lose all of his advertising dollars and not be able to complete his mission with the capital he has. Yeah.

90
00:13:43.744 --> 00:13:53.264
Did you... I, what I'm, like, interested in, 'cause you have experienced this of, like, being sort of between the two worlds, is being between the tech world and the advertising industry, and these are...

91
00:13:53.384 --> 00:14:05.544
Like, that, that's not advertising, it's advertising industry. And they, they've always been sort of, to me, like, very strange codependent relationship to some degree because both sides, like, don't get each other.

92
00:14:05.644 --> 00:14:14.384
And that's why Silicon Valley, like, I always felt, like, the tech industry, like, didn't take ad tech seriously. Is this true? [laughs] This is always my- [laughs]

93
00:14:14.444 --> 00:14:24.264
I think, I think that- Silicon Valley people only care about app installs and stuff. Like, that was the only thing that got them, like, they're like, yeah, dealing with, like, a lunch and learn, like, no.

94
00:14:25.504 --> 00:14:33.984
Silicon Valley people, uh, uh, by that I mean kinda the engineering mindset- I, I use it as shorthand. It's shorthand. It, but it's accurate shorthand.

95
00:14:34.224 --> 00:14:39.624
They believe in sort of science and measurement and direct paths from A to B.

96
00:14:39.704 --> 00:14:47.824
And so when they think about, how do I change an algorithm to increase the click-through rate of an ad, that makes a lot of sense to them, and they're pretty natural at that.

97
00:14:47.844 --> 00:14:59.294
And they've, and if you think about the success of Google and Facebook, incredible, like, incredible domination over the advertising business. If you think about a brand, about what does Coca-Cola mean to you, right?

98
00:14:59.324 --> 00:15:10.804
The red color, the script, the bottle. That's in your head, man. [laughs] Yeah. That's, you know, advertising works between your ears, uh, and that's hard. They do not get that at all.

99
00:15:10.824 --> 00:15:18.704
And then the way that plays out, uh, this is the example I always like to give when I'm in the context of product management, uh, 'cause I've been in ad tech and product management.

100
00:15:18.764 --> 00:15:29.774
I'll, I'll ask an engineer, like, "If, if I had a, a million dollars to spend to launch a new shoe for Nike, and you could launch it for, you know, 900,000 instead of a million, is that a good thing or a bad thing?"

101
00:15:29.904 --> 00:15:42.294
And every engineer, of course, will say, "That's a good thing. I saved you $100,000. That's fantastic." And the answer is, "No, I was paid by Nike to spend a million dollars. You under-delivered by 100,000."

102
00:15:43.984 --> 00:15:51.764
Uh, and that, that in a nutshell is, is the problem. Okay. The advertisers are not coming back to Twitter, right? No. No.

103
00:15:52.084 --> 00:16:03.344
Because I think this is, to me, this is a thing, like, where I sort of go maybe conspiracy, like, is the answer because he's obviously a brilliant person, right? Obviously.

104
00:16:04.024 --> 00:16:16.204
And he's mishandled this, the, the, like, advertising to such a degree that it, it has to almost be, like, willfu- willful. [laughs] Like, I, I went through that entire whatever.

105
00:16:16.364 --> 00:16:20.804
I, I joked that it was, like, the biggest lunch and learn ever that he did with advertisers on Twitter spaces.

106
00:16:20.964 --> 00:16:34.704
It was 111,000 people at the end, and, like, silent advertisers because of, like, who is going to, like, press this guy and get attacked by, like, every troll, like, known to humanity? Like, no CMO is gonna do that.

107
00:16:35.124 --> 00:16:43.584
Right. Not, not happening. But, like, the stuff he was talking about, he was like, "Yeah, I just, I just go there and, like, shit post all the time. You should, you should just try it. Just get in and, like, do it."

108
00:16:44.144 --> 00:16:55.794
Yeah. [laughs] I'm like, "Dude, have you ever talked to, like, a CMO?" Like, no, Jim Stengel's not gonna get on there and, like, start, like, shit posting, like, with meme lords and stuff.

109
00:16:56.364 --> 00:17:07.344
They're all, like, terrified of, like, getting fired because they're always fired. [laughs] Well, yeah, CMOs are always fired. Like, that's, that's a nat- law of nature. That has nothing to do with Elon or not.

110
00:17:07.524 --> 00:17:26.864
I think just- No, but it just, it, it spoke to this, like, misunderstanding of, like, if he didn't know that, that, that people were going to, that, that advertisers, the advertising industry, it's not advertisers, the advertising industry was going to flee, that, that means that he didn't know what they truly value.

111
00:17:27.904 --> 00:17:38.164
Two, like, he didn't understand at all the nature of, of the business and how risk-averse these companies are. Yeah.

112
00:17:38.354 --> 00:17:45.364
I, I think from what I've heard or seen, Elon doesn't understand advertising at all, even in a Silicon Valley frame.

113
00:17:45.444 --> 00:17:56.264
Like, I don't believe he understands how Google and Facebook make money either, which is quite a, quite a, uh, quite a fact if that's true, 'cause he's, he hasn't s- It's all very reassuring to me. Yeah.

114
00:17:56.324 --> 00:18:05.704
I mean, I would think, I would think this would've played out with, like, "Hey, we're gonna create the next AdWords, but for Twitter," and that would've been a naive thing to say, but, you know, reasonable.

115
00:18:05.784 --> 00:18:15.632
But in fact, uh, he doesn't seem to understand the direct response market or the brand market.He doesn't seem to understand that something like 80% of Twitter revenue is brands, not direct response.

116
00:18:16.152 --> 00:18:32.992
And then to get, to reinforce what your points were, was that he doesn't understand how risk-averse brands are, how much the environment matters to them, um, how Twitter was always a marginal buy, uh, because it's smaller and more risky to begin with.

117
00:18:33.592 --> 00:18:44.902
And his top priority of effectively making it riskier by allowing more, quote... I'm putting it in air quotes, "free speech," which, you know, bullshit phrase. But also- [laughs]...

118
00:18:44.912 --> 00:18:56.452
you know, allowing people to be verified for $8, uh, plus just mix it up and keep it exciting, those are, those are all good reasons to not advertise on Twitter. Yeah.

119
00:18:56.472 --> 00:19:06.922
And then you throw this into the environment of we're just entering a recession where people will be cutting back their ad budgets, and you could just imagine the conversation. Like, "Hey, we need to save 10%.

120
00:19:06.972 --> 00:19:12.542
What should we do?" "Hmm. I don't know. Why don't we just turn off Twitter until we figure that thing out?" Yeah.

121
00:19:12.542 --> 00:19:23.532
I mean, I have, I have non-people who don't work in the advertising industry as friends, and I just got, like, text message in one of my groups, like, with, like, the, the brand impersonations and stuff out there. Yeah.

122
00:19:23.592 --> 00:19:29.432
And, like, every, every brand is getting them, and, like, if he doesn't get that, like...

123
00:19:29.452 --> 00:19:36.582
Like, he probably doesn't even know about the screenshot industrial complex and how, like, much of the ad industry [laughs] has been ruled by it and stuff.

124
00:19:37.062 --> 00:19:46.182
[laughs] 'Cause, like, that's the part that I find interesting in that, like, maybe he didn't surround himself with, with the people who, who tell him like, "No, Elon, this isn't how this works."

125
00:19:46.472 --> 00:19:56.452
Well- Well, he seems like he fires the people who, who tell him that. [laughs] Yeah. I mean, okay, let's, let's just go through the last... Let's go through, uh, six months prior to the acquisition.

126
00:19:56.492 --> 00:20:04.812
We have Twitter f- uh, Twitter after several, several earnings releases in a row, talks about how the revenue's delayed because they're building a new ad server.

127
00:20:05.012 --> 00:20:10.652
That, that's kind of a red flag to anyone in ad tech, like building a new ad server. That sounds hard and complicated.

128
00:20:10.772 --> 00:20:21.071
Uh, then they sell MoPub, which was a division of Twitter that was really important in the mobile, uh, direct response world. They sold it for a billion dollars.

129
00:20:21.112 --> 00:20:30.992
That's a nice amount of change, but if you expect Twitter to be a leader in direct response advertising, selling this really valuable, irreplaceable asset doesn't seem like the best move.

130
00:20:31.052 --> 00:20:40.952
Then the new CEO comes and fires most of the ad team, including the leader, Bruce Falck. This all happened before Elon entered the building. Now let's go over the last week of activities. Two weeks.

131
00:20:41.012 --> 00:20:52.032
Elon shows up, day one, posts fake news about Paul Pelosi's attempted assassination. Good move. Day three, I think, he threatens to sue advertisers. [laughs] He didn't threaten to sue them.

132
00:20:52.072 --> 00:21:00.921
He threatened to, like, incite a mob against them. No, he d- That's the accurate- Yeah, you're right. I think day three he threatened a mob against people who refused to advertise. Oh.

133
00:21:00.932 --> 00:21:09.062
And then, like, day five, he agreed with someone who threatened to sue them a- with torturous interference. Oh, okay. Yeah, that's fair. Right?

134
00:21:09.092 --> 00:21:24.832
And then by day eight, we have a deluge of fake verified people making fun of the very brands that would be advertisers in really painful ways, like, uh, accusing Chiquita of fomenting co- coups in Latin America and, and- Not wrong.

135
00:21:25.052 --> 00:21:37.092
Not wrong. I didn't say it was wrong. Uh, and [laughs] they were funny. Uh, Eli Lilly, uh, giving insulin away for free, uh, which apparently hit their stock price today- Yeah... by, by a large amount of money.

136
00:21:37.172 --> 00:21:44.512
And, uh, and, and this is also, uh, just as a footnote, all the auto manufacturers just pulled all their ads right away because they don't wanna support Tesla.

137
00:21:44.532 --> 00:21:54.032
So one category, one vertical category that represents, you know, a pretty big piece of brand advertising is just off the table. So I must have missed a couple things, but that was, that's a quick little history lesson.

138
00:21:54.812 --> 00:22:05.092
Yeah. It's pretty insane. I mean... But it is entertaining. I mean, when he said like, "Are you not entertained?" Like, I think we, uh, are all entertained. Entertained. Maybe, maybe too much.

139
00:22:05.592 --> 00:22:14.812
Let, let's pivot to talk about where ad tech is now, because you've, you've, you've been in the industry for a while, but then you s- you, you're sort of out of it. But you'll get, you'll get, you'll get pulled back in.

140
00:22:15.872 --> 00:22:23.322
Uh, yeah. So I sold my company, um, Desk- Once you're in ad tech, you can't, you can't, you can't get out of ad tech. No, some people have. I've seen some people get out of ad tech. Yeah.

141
00:22:23.352 --> 00:22:31.951
You usually have to flee the country as well. Keep, keep that blue- They usually end up in Mexico. Keep those blue blazers in your closet ready. Yeah. 'Cause you're gonna need them again. Yeah. You know, I...

142
00:22:32.092 --> 00:22:40.472
Every event comes up, like CAN and Demand Toco, and people are like, "Are you going?" And I'm, I'm think to myself, "Maybe I should go." But then I realize I have no reason.

143
00:22:40.612 --> 00:22:49.292
Not only would I have to pay for it myself, but also I would literally have nothing to do. It's really sad. Fun. So no, you, you sold, you sold Beeswax. Sold Beeswax. But, but yeah.

144
00:22:49.472 --> 00:23:01.072
E- explain to people what you, what you were doing with, with Beeswax and the exit. Yeah, sure. So we started Beeswax in 2014. It was a, uh, DSP more or less. We called it a bidder as a service, but really it was a DSP.

145
00:23:01.202 --> 00:23:10.582
All right. DSP, demand side platform. Demand side platform. And that's the- Yeah. No acronyms. Jay, you know- Real talk. No acronyms... you have no idea, you have no idea where this shit is gonna go next, okay? Okay?

146
00:23:10.612 --> 00:23:28.292
Yeah. [laughs] So- I'm gonna be jumping it a lot. [gentle music] [laughs] All right. So you wanna buy ads in a programmatic way, meaning using servers, not people. You're not buying martinis.

147
00:23:28.372 --> 00:23:38.432
You're, you're, like, using computers. You have to use a product called a DSP or demand side platform. The two leading ones are Google and, uh, and a publicly traded company called The Trade Desk.

148
00:23:38.462 --> 00:23:39.972
I'm pretty sure everyone's heard of The Trade Desk.

149
00:23:40.112 --> 00:23:50.932
We were a small company that focused on the most technical users, and so we had APIs and a lot of customization, and we enabled, uh, folks who wanted to buy ads in very sort of advanced ways.

150
00:23:50.992 --> 00:23:57.562
L- kind of like the way a hedge fund buys stocks. They don't just log into Charles Schwab and buy some stocks. Okay. They have direct connections to the exchanges.

151
00:23:57.592 --> 00:24:01.864
This gave them, like, an advantage, 'cause it's a bidding environment. Basically. It's like-That's right.

152
00:24:02.324 --> 00:24:12.924
So we, we grew it very rapidly and had a very nice exit where we sold it to Comcast's FreeWheel division in two thou- uh, 2021, early '21. Uh, the company's doing really well inside of Comcast.

153
00:24:12.984 --> 00:24:33.603
I exited a year later in early '22, and now I've started a couple of things, most notably a company called Marchitecture or marchitecture.tv, which is, uh, in-depth interviews with executives and CEOs of ad tech and martech companies, sort of no-nonsense, get to the facts, help you understand if they're vendors you wanna work with or what they do and how it works.

154
00:24:34.284 --> 00:24:48.244
And, and this-- I, I love the, the Marchitecture i- idea, concept because, like, I've, I've been confused by ad tech for over a decade. Yes. [laughs] Yes. That's the-- that's what it's for. I just keep asking.

155
00:24:48.504 --> 00:24:58.064
They-- You know, I get these executives on, and they give me their usual bullshit spiel about, like, how they bring buyers and sellers together. Now you know- And I'm sort of like- Now you, now you know how I feel.

156
00:24:58.084 --> 00:25:04.344
And I'm like, "This is my show. You tell me what you really do. No." [laughs] You know, "I do not understand what you just said. Break it down."

157
00:25:04.544 --> 00:25:12.904
But, like, this has been like a co- So the-- And I think the need for it has always been, like, in some ways the Achilles heel of, of the ad tech industry, right?

158
00:25:12.944 --> 00:25:23.124
Like, we've been subjected to Terry Kawaja's, like, LumaScape slide for... Uh, when did he start it? It's probably been, like, a decade. Yeah, probably a decade. Like, at this point. Yeah.

159
00:25:23.134 --> 00:25:31.184
And anyone who hasn't seen it, it's this eye chart of-- I don't know, there's-- He's got, like, multiple, like... But there's, there's hundreds of logos on there.

160
00:25:31.204 --> 00:25:38.304
And the, the knock on ad tech was always it was too complex, it was too confusing, there was too many middlemen.

161
00:25:38.324 --> 00:25:46.944
And then, like, Google, like, basically just, like, scooped up each and, a bunch of logos and just built their own thing, and then people were like, "There's not enough competition. There's no innovation."

162
00:25:46.954 --> 00:25:53.104
[laughs] It's like, well, which do you want? Right. Uh, I think innovation requires some level of fragmentation.

163
00:25:53.284 --> 00:26:13.704
If you think about all the different ways that advertisers and their agencies can buy digital media across medium, like, you know, TV banner ads, uh, mobile ads, et cetera, across creative types, video, text, imagery, across different optimization types using different kinds of data, different geographies, it's a very complicated business.

164
00:26:13.804 --> 00:26:24.794
Much more complicated and, and hopefully more efficient than, uh, traditional media like TV and, uh, broadcast and magazines. So, so I'm not scared of the complexity.

165
00:26:24.884 --> 00:26:36.004
I think it's a good talking point for those who want to, you know, sell their lack of complexity. But the [laughs] the, uh, the complexity, it's not a problem in and of itself. Yeah.

166
00:26:36.044 --> 00:26:47.404
But, like, so with, like, Marchitecture, the, the idea is, like, anyone who is on the buy or the sell s- side, you know, as an advertiser or a publisher, um, has, like, an extensive tech stack.

167
00:26:47.444 --> 00:27:00.024
Like, and just being able to, like, run ads, count the ads, put the ads in front of theoretically the right person at the right time and stuff like this requires an inordinate number of partners.

168
00:27:00.084 --> 00:27:11.824
I mean, I know this having done many, many, many events in this industry. [laughs] Yeah. Uh, it does, uh, to do it right. I mean, there-- The market actually segments pretty clearly into large and small.

169
00:27:12.284 --> 00:27:23.044
So if you're a small publisher, you may get, get by entirely with the Google stack on the sell side. You just put some tags up and make money. If you're a small advertiser, you might get away with Google plus Facebook.

170
00:27:23.204 --> 00:27:33.584
But as soon as you become large and you become Estee Lauder or, or General Motors or even, like, large, like, you know, Peloton. I know they're not doing that well, but they do a lot of advertising.

171
00:27:33.864 --> 00:27:45.704
You suddenly have a lot of complex requirements around data collection, privacy, uh, optimization, creative, geographies, multiple agencies, uh, the mix of brand and direct response.

172
00:27:45.764 --> 00:27:52.724
All these things create complexity and, and whatever you can think of, there's probably a vendor for you, right? Right.

173
00:27:52.764 --> 00:28:04.764
And so the idea behind Marchitecture is that when you have to figure out the vendors that you're gonna use, first of all, I-- this is my own, like, outside opinion. Like, everyone sounds the same.

174
00:28:04.774 --> 00:28:11.214
Like they just- They do... sound the same. They do. Like, they're, they're just-- The words are just mixed up. It reminds me of, like, a Mexican restaurant where you're like- Yeah...

175
00:28:11.214 --> 00:28:19.784
"Hmm, the chimichanga is actually just the enchilada, [laughs] and the enchilada is really just a burrito." And like- Yeah. And so you really have to dig in to, like, understand the differentiation.

176
00:28:20.004 --> 00:28:30.544
And, and tell me if I'm wrong on this one, a lot of the people who are digging in to understand the differentiation are, are what is euphemistically called non-technical. Yeah, I think that's a big part of it.

177
00:28:30.604 --> 00:28:37.804
Uh, the idea behind Marchitecture is to replace your first meeting. Uh, so if you, you're interested in, uh, say a creative solution.

178
00:28:37.884 --> 00:28:44.024
So you want a software product that lets your designers create a lot of banner ads really quickly. And that's very valuable, right?

179
00:28:44.184 --> 00:28:49.324
You go to the websites, you find six or seven vendors, and they all sound exactly the same, and you really just cannot tell.

180
00:28:49.624 --> 00:29:00.304
So you have to have a meeting with all of them, and then you, your meeting has five people in it times, you know, six vendors, and suddenly you've spent $10,000 before you even really figured out what you're even a- trying to do.

181
00:29:00.504 --> 00:29:09.104
Having myself, who's an, you know, admittedly an expert, you know, grill the CEO for half an hour on video about what they actually do, you do not need that first meeting.

182
00:29:09.164 --> 00:29:16.484
You could immediately eliminate that vendor and say, "No, no, no, they're, they're not right for us," or, "They sound great," like, "Let's get them in right away." Um, that's the idea.

183
00:29:16.544 --> 00:29:23.094
I'm not gonna-- You're not gonna be able to watch my video and immediately choose a vendor or, or solve your business problem, but we're gonna save you a lot of time and money.

184
00:29:23.964 --> 00:29:30.704
So give me the case for why ad tech is interesting right now. I haven't been paying att- I've been living in Miami for a couple years. I don't know what the hell's going on.

185
00:29:30.844 --> 00:29:37.134
Well, I'll tell you what's interesting in ad tech. How about that? So there, there's really three things that are interesting in it. There's more than three, but three, uh, top things.

186
00:29:37.144 --> 00:29:45.864
Well, you can just say no, and then we'll just end the podcast. [laughs] I think for ordinary people, ad tech should be a little interesting, and they should pay a little more attention than they do.

187
00:29:45.884 --> 00:30:01.464
For the average consumer who is not in our business, it's pretty interesting that Apple, just three years ago, uh, went up on stage, they showed a picture of a bicycle, and they said, or Tim Cook said, "Doesn't it bother you when that bicycle is following around the web?"

188
00:30:01.494 --> 00:30:11.164
Yes. "And it's so creepy there's this bicycle, and wouldn't you rather it didn't happen? Well, we're introducing privacy." And that sounded great. We're just-- We came up with a new solution.

189
00:30:11.174 --> 00:30:18.944
This is like a Saturday Night Live, like, commercial. It, it pretty much- It's called privacy. [laughs] I'm, I'm not very far off from what actually happened. And everyone claps and goes, "That sounds great.

190
00:30:19.024 --> 00:30:30.194
Apple's our friend." And then, uh, you know, round numbers, I would say, uh, a trillion dollars of value disappeared from the public markets. Yeah. Uh, people went out of business.

191
00:30:30.284 --> 00:30:37.924
Peloton practically went out of business. People can't sell, uh, sheets and pillows online anymore. You know- So wait, explain- Facebook fired 10,000 people.

192
00:30:37.964 --> 00:30:45.524
Yeah, so explain, because this is what I was sort of getting at, why it, it is kind of an interesting time. That was what I was going to- Yeah... get to, otherwise we would've just finished this thing.

193
00:30:45.664 --> 00:30:58.600
IsWhat specifically did Apple do with both- Yeah... IDFA and then this thing called ATT? Yeah. So let's put it in perspective. So the iPhone was introduced, I guess now 11 years ago, 10, 11 years ago.

194
00:30:58.640 --> 00:31:06.700
A new medium, whenever there's a new medium, advertisers find that, and some mediums are less attractive to advertisers and some are more attractive to advertisers.

195
00:31:07.080 --> 00:31:16.740
So in the history of the mobile phones before the iPhone, it was a very unattractive business. There were very few ads on BlackBerrys and flip phones. Very few. There were some, but there were very...

196
00:31:16.780 --> 00:31:30.020
So the iPhone was introduced, it became incredibly popular, as we all know, and it had a feature in it which allowed app developers and marketers, and different people to anonymously, anonymously identify the users who were looking at the apps.

197
00:31:30.080 --> 00:31:37.430
It was called an IDFA. It was called something else at the time, they changed the name, but that's too much detail. So, so this anonymous identifier- You're at the right altitude. Yeah. Thank you.

198
00:31:37.430 --> 00:31:48.680
The anonymous identifier was amazing. Like, it was incredible for everyone in the ecosystem because it allowed you to optimize what ads were shown to that user based on their behavior, both on- But it was a hack.

199
00:31:48.720 --> 00:31:57.740
It was like the third party cookie was a hack. Is this another hack? It was not a hack, no. No, no. It was not a hack. No. It was made for this. It w- well, yes, it was the I- IDFA stands for Identifier for Advertising.

200
00:31:57.820 --> 00:32:06.360
Okay. Um, it was not a hack. It's like Beto O'Rourke. Remember when he, like, declared for president? He's like, "I'm made for this." And like, uh, like Beto, it had a limited shelf life.

201
00:32:06.430 --> 00:32:13.320
[laughs] Uh, so the IDFA was magical, and it actually solved the hack that was the case in the browser.

202
00:32:13.420 --> 00:32:22.600
In the, in the web browser, we had the hack, which was cookies, which were not meant for advertising, but were glommed on by everyone, and they, they disappear, and they don't follow you around, and they're not very reliable.

203
00:32:22.980 --> 00:32:32.550
The IDFA was magical and worked great, and as a result, a very robust advertising and marketing ecosystem, uh, developed in mobile w- uh, highlighted by Facebook and Instagram.

204
00:32:32.640 --> 00:32:41.640
I'd like to say that while Google Search is the world's greatest advertising product ever created, Instagram is the world's greatest brand advertising ever created.

205
00:32:41.700 --> 00:32:48.830
You know, uh, basically you could show handbags and shoes and it's really direct response-ish- Yeah... but still. Anyway, the point being, the...

206
00:32:48.860 --> 00:33:01.420
You could say if Apple had never put the IDFA there in the first place, it would've played out very differently and the whole ecosystem around advertising in apps and app economics would've been different, but it was there for many years.

207
00:33:02.000 --> 00:33:13.220
And, and Apple decided, for some good reasons, to make it so that that identifier was only opt-in. So it didn't ex- it was not shared with all these parties by default.

208
00:33:13.230 --> 00:33:21.080
The user had to be shown a prompt called the ATT, which is, I forget what it stands for. App Tracking Transparency, that's what it stands for. Yeah. ATT.

209
00:33:21.300 --> 00:33:33.600
And, and so when you'd open an app, and probably most of the listeners here are aware of this, you would sometimes see this interesting dialogue that said, "Would you like to allow this app to track you around your, around the web?"

210
00:33:33.640 --> 00:33:41.570
Or whatever the language is. Yes. Really scary language. Yes. Yes. Scary [laughs] yes. Right, right. Please. Really scary language. Uh, and, and if, and,

211
00:33:42.520 --> 00:33:47.029
and, you know, obviously a minority of people will say yes when, when prompted like that. Yeah.

212
00:33:47.040 --> 00:33:56.380
And Apple went even further by saying, "Not only do you have to use the scary prompt, you are not allowed to give people any incentive to say yes. Uh, you're...

213
00:33:56.460 --> 00:34:07.290
And if they say no, you can't track them by any other method." Yeah. "They have to be dead to you." And, uh- And not to, not to be a conspiracy theory here, theorist here. Yeah. But it, it does, it, it, it...

214
00:34:07.340 --> 00:34:15.840
I think I should point out that Apple also did not have an advertising business really, and they sold phones, so they didn't give a shit about, like, kneecapping anyone. Oh, absolutely.

215
00:34:15.940 --> 00:34:20.940
I don't think they're- Because they tried advertising once with iAd and it was a disaster. Okay.

216
00:34:21.000 --> 00:34:25.469
Well, there's some nuance to that, but yeah, I don't think you need to be a conspiracy theorist to say that there were- [laughs]...

217
00:34:25.520 --> 00:34:33.180
clearly they, they saw that their two rivals, their number one rival is Google, and Google obviously has enormous advertising exposure.

218
00:34:33.200 --> 00:34:44.180
And n- and their number one kinda pain in the ass, I wouldn't call them a rival, is Meta Facebook. I mean, I think there's no love lo- lost between those companies. So hitting them in the kneecaps was kinda fun for Tim.

219
00:34:44.220 --> 00:34:50.880
Well, I mean, T- Tim, Tim Cook was... I mean, uh, I think going back to Steve Jobs, it was just like your business model sucks. I mean, they were- Yes...

220
00:34:50.960 --> 00:35:00.800
very, very upfront with having zero respect for, for how, uh, Facebook decided to, uh- Absolutely. And as for Apple iAd, you know, Apple's been...

221
00:35:00.880 --> 00:35:07.580
Apple, the original launch of iAd under Steve Jobs, uh, was very video and TV oriented, and didn't do very well, as you said.

222
00:35:08.500 --> 00:35:17.410
But it didn't go away, and they've been slowly creeping back, and they've built actually a very substantial business of search ads on the App Store. Oh, yeah.

223
00:35:17.440 --> 00:35:20.980
Similar to Google Ads, where basically you search- That's the advertising, that's what I think.

224
00:35:21.140 --> 00:35:30.180
When Silicon Valley thinks about advertising, they think about like that, 'cause it's like a toll booth, and they're like, "I love this." Yeah. "This is a revenue spigot. I can turn on the revenue spigot." Yeah.

225
00:35:30.220 --> 00:35:35.640
I mean, go into the App Store and search for Uber, and I bet you'll see a Lyft ad, and search for Lyft, you'll see an Uber ad. It's great.

226
00:35:35.760 --> 00:35:43.480
Uh, and, uh- Do you remember when, like, when there was the controversy of, like, whether Google would, like, sell, like, keywords based on, like, trademarks and stuff? Yeah.

227
00:35:43.510 --> 00:35:53.260
And then, like, they ran the numbers and they were like, "Pfft. Sue us." [laughs] Yeah. Yeah. They got sued, they got sued by a lot of people. I think they won those lawsuits. I mean, clearly they're still doing it.

228
00:35:53.300 --> 00:36:01.980
You know, branded key- branded keywords was what the controversy- Yeah... was over. Yeah. Brooks Brothers was pissed about it. [laughs] It, it really is- Maybe that's what happened to Brooks Brothers.

229
00:36:02.060 --> 00:36:10.660
Maybe it wasn't the boxy suits going out of fashion. Are they out of business, Bro- Brooks Brothers? I don't know. They should be. All right. A little- Well, who's wearing a suit?

230
00:36:11.140 --> 00:36:13.660
I guess they're not gonna be sponsoring the reboot anytime soon.

231
00:36:13.700 --> 00:36:25.090
[laughs] I mean, people do need- Look, I'm not, I'm not, I'm not toadying up to Brooks Brothers to get a- I'm just saying, your audience probably has, like, 100% blue blazer penetration. [laughs] Okay. All right.

232
00:36:25.140 --> 00:36:29.080
Let's get back to Apple. Okay. And, and ad tech. Yeah. Yeah.

233
00:36:29.120 --> 00:36:39.560
So the other shoe that dropped, and this goes into the conspiracy theory bucket, is that very clearly Apple has been increasing their investment in advertising corresponding to the rollout of ATT.

234
00:36:39.660 --> 00:36:42.950
[laughs] So Apple collects a lot of information. Coincidence, I'm sure. Yes. [laughs] I know.

235
00:36:43.020 --> 00:36:49.360
Apple collects a lot of information about what you do on your phone, and it has a different preference that's not, doesn't have the scary language. It, it...

236
00:36:49.400 --> 00:36:54.440
Apple has this language that's like, "Would you like to ha- help Apple to, you know, make your phone better with ads?"

237
00:36:54.560 --> 00:37:04.742
And, and they, they b- claim that-It's all kosher because instead of passing your data around to these anonymous evil ad tech companies, they keep it- Yeah...

238
00:37:04.772 --> 00:37:07.722
and they keep it, and they keep it on your phone, so that's really the distinction.

239
00:37:07.932 --> 00:37:17.592
They say if they, if they profile you and show you those creepy ads, but the data's on your phone, then that's cool, nothing to be worried about, not a privacy issue.

240
00:37:17.632 --> 00:37:34.952
But if the creepy ads are coming from outside your phone, you know, you're basically one step away from 1984. [laughs] Okay.

241
00:37:35.462 --> 00:37:43.932
And this has had a tremendous impact on a particular sector of, really of the advertising industry. Mm-hmm. Because it's like

242
00:37:45.052 --> 00:37:53.072
y- you have this strange situation where the, the ad holding companies are reporting, um- Yeah... great results. They're raising, like, you know, their, their forecasts and stuff.

243
00:37:53.552 --> 00:38:03.312
And then you have Snap, which is, like [laughs] their stock is down, like, 85%. Right. Meta just, like, laid off 11,000 people. Like, it, it's really...

244
00:38:03.892 --> 00:38:12.612
It seems like it's really impacted a particular form of, like, of targeted advertising. Yeah. I mean, app advertising is really what's been hit the worst.

245
00:38:12.752 --> 00:38:18.272
You know, the b- the browser has had problems with cookies and identity for quite some time, but it's been kind of a slow burn.

246
00:38:18.412 --> 00:38:29.732
The Apple ATT impact has come on very strong over the past 12 months, hitting companies that advertise in their apps on a direct response basis. That combination is what got killed.

247
00:38:29.812 --> 00:38:39.632
What's fascinating about advertising, and I'd love to see a study on this, but I just know it anecdotally, is that there's al- there's almost two totally separate ecosystems that don't talk. It's like- Yeah...

248
00:38:39.652 --> 00:38:49.632
red state and blue state America. And that is, like, what I call ad tech, what you call ad tech- Yeah... is one ecosystem, and then mobile is totally different. They're different people- Oh, yeah... different companies.

249
00:38:49.672 --> 00:38:57.422
They don't talk to each other. Agencies- And they're massive. The o- and they're in Silicon Valley. They're massive. They're Silicon Valley. The, the app download people, like, are just- Yes...

250
00:38:57.472 --> 00:39:03.892
it seems like they're in a different, like, world. Yeah, they are. They're different advertisers, different technology companies, different measurement companies.

251
00:39:03.932 --> 00:39:14.732
And that's one of the things that, that is also interesting, which is if you look at a mainstream ad tech company, I, I'll just call a name, like The Trade Desk, one of the biggest, they really don't do much app installs.

252
00:39:14.952 --> 00:39:23.032
Like, it's just not part of- Yeah... what people use them for. If you go to, like, a big holding company like Publicis, probably not a lot of app installs going on there.

253
00:39:23.132 --> 00:39:31.652
Probably if you're a Publicis Fortune 500 customer and you need your app to get installed, you probably are calling up Applovin or someone like that to do it for you. Yeah.

254
00:39:31.692 --> 00:39:41.152
Maybe your agency's evolved sen- tangentially, but they don't have the skills to make it happen. All right. So we're entering in this new period where targeting is gonna be harder, right?

255
00:39:41.432 --> 00:39:51.252
Like, I mean, what, what's going on in the browser with, you know, Google is going to eventually deprecate this third-party cookie. It's going away at some point, right? The Apple stuff's happening over there.

256
00:39:51.292 --> 00:39:59.512
What seems very clear is that the era of, like, easy, cheap targeting is sort of coming to an end. So what...

257
00:39:59.632 --> 00:40:10.092
I, I, I talk about how, like, it seems like we're in this period where a lot of, like, eras are ending, but we're just unsure about what follows. Yeah. Just like democracy, right? Yeah. That's what I meant.

258
00:40:10.662 --> 00:40:14.192
The- [laughs] Right. Follow. Sorry to get all political on the ad tech talk. Uh, the...

259
00:40:14.232 --> 00:40:25.212
So, the, the, it's, uh, it's definitely true that consumers are going to be less identifiable on their digital devices because of legislative and privacy changes. Like, that, that's a fact, right?

260
00:40:25.572 --> 00:40:31.011
So then you, you have to ladder that up to, like, what is the impact on, you know, advertising and on business.

261
00:40:31.232 --> 00:40:45.731
And, and what is interesting is that while it may be impossible, let's say, for an advertiser to target the same consumers on a one-to-one basis across channels, that may just not be possible anymore, it is possible for publishers to do it.

262
00:40:45.852 --> 00:40:59.772
So a large-scale publisher who has a login wall or a reg wall knows who those consumers are at a one-to-one basis, and so does, uh, most CTV, connected TV apps have logins. They know who the users are.

263
00:40:59.892 --> 00:41:08.972
The cable companies and the phone companies know who the consumers are, right? So we haven't had a wholesale removal of identity. It's not an anonymous web anymore.

264
00:41:09.032 --> 00:41:19.332
It's just the power is shifting, uh, and the technology is shifting. So on the power, probably this should be good for scaled publishers in some way in that an advertiser in...

265
00:41:19.392 --> 00:41:28.982
Over the last 10 years, advertisers have sort of seen publisher relations as an optional part of the business, where they could just log in to their favorite DSP, the same acronym- Yeah...

266
00:41:28.992 --> 00:41:37.932
of choice, and get access to those consumers wherever they might be looking at content. Yes. Uh, which was bad for publishers. Describe to me as, as see a cookie, hit a cookie. Yeah.

267
00:41:38.012 --> 00:41:45.372
Or, um- By a publisher ad salesperson. Uh, I've also heard of it as, you know, digging for oil on other people's land. Yeah.

268
00:41:45.412 --> 00:41:49.752
Yeah, I mean, the, my mom always told me, like, "Why buy the cow when you can get the milk for free?" Sure.

269
00:41:49.992 --> 00:41:59.422
The, the difference is that the, in this case, just to use that metaphor, like the advertiser is getting the milk. The farmer doesn't even know where the cow is. Um- Okay... right?

270
00:41:59.492 --> 00:42:07.232
Um, so we're moving to a world where the farmer knows where the cow is and the advertiser doesn't know a- anything. There's a, there's a fence. The advertiser isn't even allowed in to look. Yeah.

271
00:42:07.292 --> 00:42:15.952
[laughs] So, so that's maybe good news for scaled publishers, that they can recapture some of that- Okay... uh, control. And bad news for the long tail, 'cause the long tail, it's always bad news for them.

272
00:42:16.022 --> 00:42:21.012
'Cause they don't have, they don't have Salesforces, they don't have a l- a, a lot of first party, what's known as first party data.

273
00:42:21.072 --> 00:42:27.832
Like, and they're, they're always, like, dependent on, on these kind of automated solutions for monetization. Yeah. It's bad news in two ways.

274
00:42:27.912 --> 00:42:38.032
First of all, it's bad news in that the results have gotten worse, and that's why we've seen the implosion of D2C companies. So the D2C companies can no longer attract customers at, at affordable ROI.

275
00:42:38.152 --> 00:42:45.352
So some of them have gone out of business or, or had to sell because they literally just can't find the customers. And secondly, it's bad- And that's why you look at Shopify, it's like stock price.

276
00:42:45.452 --> 00:42:56.682
Like, I made a mistake during the pandemic of buying Shopify stock. Yes. And like, it, the flaw of, of Shopify was the fact that they outsourced acquisition to cheap ads on Instagram. Right.

277
00:42:56.712 --> 00:43:09.732
And then, like, those cheap ads on Instagram became a lot more expensive and a lot less effective, and the DTC companiesCould only exist in a, in a world of cheap ads, it seems. Yes. Yeah, well, it was a magical time.

278
00:43:09.812 --> 00:43:11.462
I use the word magical a lot, but it was like- [laughs]...

279
00:43:11.492 --> 00:43:22.952
the Instagram ads just worked so well that if, if you imagine, the ads worked so well, they created thousands of businesses, uh, ba- based solely on attracting customers fr- from Instagram.

280
00:43:23.012 --> 00:43:29.382
They- Tell these stories to your kids at night. It was a magical time, kids. Uh, you'd be surprised- Cheap load CAC... the sort of conversation. Right.

281
00:43:29.392 --> 00:43:37.652
And the second way this is bad for long tail is that they become more dependent on Google and Facebook. Like, they're already very dependent on, on them, uh, but when they c- they can't...

282
00:43:37.712 --> 00:43:44.892
In a world with less data and less ability to differentiate, they become even more dependent on whoever can.

283
00:43:45.262 --> 00:43:51.132
And, you know, Google and Facebook are kind of the one-eyed men in this situation where, like, everyone's blind- Mm-hmm...

284
00:43:51.152 --> 00:43:57.212
but they have the scale and the ML models and everything to make sense out of the anonymous sort of world. Yeah.

285
00:43:57.272 --> 00:44:03.312
So final thing, and then we're gonna go, is, and, and this may be fraught because y- you, you were an employee there.

286
00:44:03.352 --> 00:44:14.452
Give me the, the, the case for, and then the case against forcing Google to relinquish its ad tech assets. Yeah, sure. The case for... And this, I'll actually...

287
00:44:14.472 --> 00:44:21.852
I don't think this is what you're asking, but I'll answer it this way. [laughs] Why should Google want to- I'm used to this, uh, also too, Ari [laughs]. Why should Google want to spin out?

288
00:44:22.212 --> 00:44:28.972
Google should want to spin out the ad tech assets. Their network bus- they divide their business into search, uh, YouTube, network, and everything else.

289
00:44:29.072 --> 00:44:35.082
Network is, uh, declining faster than the other groups, uh, so revenue was down over the last 12 months.

290
00:44:35.512 --> 00:44:41.312
It is lower margin, uh, than the other groups, and it's the source of virtually all their antitrust and privacy problems.

291
00:44:41.352 --> 00:44:50.032
They should happily spin it out and focus on the other parts of their business, which are amazing and are growing well. I mean, YouTube had a tough quarter, but could use a lot more attention.

292
00:44:50.052 --> 00:44:56.072
That's why Google should want to. Why the government should want them to and everyone else, well, it is very big, and it's very...

293
00:44:56.152 --> 00:45:08.632
And by combining it with their other businesses like Android and YouTube and Search, there are inefficiencies in the market and more money is being taken out of advertisers' hands in a way that theoretically shouldn't happen if it was an independent company.

294
00:45:08.792 --> 00:45:20.052
The, uh, against, the pr- the con is that it actually is more efficient to have it as part of YouTube than the rest of the search, and advertisers would have worse ROI if it was an independent company. Yeah.

295
00:45:20.252 --> 00:45:25.252
I mean, this is always, like, the push and pull. Like, people want the innovation, but they want the efficiency at the end of the day.

296
00:45:25.261 --> 00:45:33.821
And, like, they want, like, to have competition in the ecosystem, but they want it to be super easy. And it's like, which do you want? Yeah. Like, you don't get everything. That's right.

297
00:45:33.821 --> 00:45:42.752
So, uh, also the final thing on this is I find The Trade Desk a very interesting company. I'm not trying to sell ads to The Trade Desk, but call me if you- You should. I'm sure they would like them... is what...

298
00:45:42.761 --> 00:45:45.312
And, and only because, like, they've been able to compete.

299
00:45:45.372 --> 00:45:55.952
Like, I would always, like, have meetings over the years with, with different people that were doing different things in, in the digital ad ecosystem, and my first question was always like, "Wait, why won't Google crush?"

300
00:45:56.172 --> 00:46:07.012
[laughs] And it was just a stock question. But Trade Desk has, like, been able to, like, compete on, like, a, a big scale. What did they get right? It w- uh, kept it simple and executed, right?

301
00:46:07.072 --> 00:46:15.752
They just did what their customers wanted, produced plain vanilla innovation after plain vanilla innovation. And by the way- Bring the endorsement [laughs]. No, I mean, that's what people want.

302
00:46:15.872 --> 00:46:25.342
They want the products that work and get better every quarter, and that's what they gave them. Uh, w- and, and, uh, I use the word innovation. Like, they weren't building boring, bad stuff.

303
00:46:25.412 --> 00:46:30.432
They were building interesting stuff. It just wasn't mind-blowing. It's not like they... It's not a nuclear reactor.

304
00:46:30.652 --> 00:46:40.892
And, uh, you know, I, I was in a similar business at Beeswax when I was running it, and we also had a lot of success by just being very focused on product, making it better every quarter, treating our customers well.

305
00:46:41.212 --> 00:46:51.092
And, uh, and the result is, you know, a fast-growing, highly profitable business with very loyal customers. So goodbye. Yeah. All right, Ari. We're gonna leave it there 'cause my battery's running out.

306
00:46:51.332 --> 00:47:04.732
Thank you so much for this little walk t- uh, through the wilds of, of the ad tech world. And everyone, when you're getting Get Smart about ad tech, check out marketecture.tv and get smart. Thanks for having me, Brian.

307
00:47:05.352 --> 00:47:15.132
Yeah, thank you. Thank you for listening this week. We will be back next week with a new episode. The Rebooting show is produced by Podhelpus. Podcasts are a great way to expand your client base.

308
00:47:15.772 --> 00:47:27.992
Podhelpus lets you focus on having engaging conversations, giving your brand the full stack of services needed for a professional look and sound. Start your podcast today at podhelp.us.

309
00:47:28.892 --> 00:47:37.712
[outro music]