WEBVTT

1
00:00:00.100 --> 00:00:09.500
[upbeat music] Welcome to the Rebooting show. I am Brian Marcey. The following is a spotlight episode I created in Cannes with my partners at Xco.

2
00:00:09.640 --> 00:00:17.550
Xco's General Manager for the Americas, Johanna Berkfust joined, as well as Jason White, the Chief Product and Technology Officer at The Arena Group.

3
00:00:18.020 --> 00:00:26.350
In this discussion, we talk about the essential challenges of running a publishing business in a more with less era, and this was a big theme of Cannes.

4
00:00:26.820 --> 00:00:34.680
And in particular, Jason zeroes in on the hard choices this forces on publishers as they, by necessity, prioritize depth over breadth.

5
00:00:35.060 --> 00:00:51.820
Really appreciate Xco for their support, and Jason and Johanna for joining this conversation. Here it is. [upbeat music] Johanna, Jason, thank you for joining me here in Cannes. You're welcome.

6
00:00:52.200 --> 00:00:59.260
How's your week been? Anything weird happened? I mean, I think actually our dinner yesterday, the wind. The wind, yeah. The wind.

7
00:00:59.420 --> 00:01:09.120
I thought some rosé was about to take off a couple of times, which is probably a first- Yeah, some wine glasses-... a first for Cannes for me... were flying at some point. Exactly, and, uh, fedoras. Yeah, exactly.

8
00:01:09.160 --> 00:01:18.220
So, so that was a little weird. But yeah. Yeah, nothing weird on my end other than the torch experience was great. Oh, yeah? Right when I first got in, there was a- Did Michael Kassan carry it?

9
00:01:18.800 --> 00:01:27.660
[laughing] I didn't see him, not on my leg. [laughing] Yeah. All right, let's get into it. I don't wanna start bleak. We just, we just started, we, we talked beforehand.

10
00:01:27.740 --> 00:01:39.060
I, I was saying how I don't like to be called cynical. At the same time, I don't know, I feel like a weather, a weather forecaster in England, right? It is... It's gonna rain. Like, I don't know what to say. It's cloudy.

11
00:01:39.690 --> 00:01:44.540
[laughs] I don't know what... Yes. Okay. Well, maybe at best it could be, like, cloudy and it might rain, but, like, bring an umbrella.

12
00:01:44.780 --> 00:01:56.800
You know, I just saw, like, Comscore figures came out today, and, you know, it's Comscore, but everyone's down. Traffic is down. It's 7% at CNN. They made sure that both CNN and Fox are down equally.

13
00:01:57.360 --> 00:02:03.700
[laughing] Which is, like, that's suspicious. It's magic how that happens, isn't it? Right? Isn't that very suspicious? Yeah. They're like, "No, it's just algorithms." I'm calling something.

14
00:02:03.840 --> 00:02:11.730
It's just, it's just algorithms. Yeah. But, like, somehow we got it- Might have to do with that... exactly 6.6% down- That's right... for both of them. The Times is the only one up. Thank you, Wordle.

15
00:02:12.160 --> 00:02:30.100
NBC News down nine and a half, USA Today down nine, Washington Post down 21. Ouch. Business Insider, ouch, down 29%. Even the Journals and, and Politico are down 12 and, and 9%. LA Times 31%, Atlantic 26%. It's- Yeah...

16
00:02:30.140 --> 00:02:39.829
it's bleak stuff. It's a trend. It is a trend. It seems like we're entering almost, like, a more with less era, is what I call it, right? I mean, there's just... This is not a blip, right? Yeah.

17
00:02:39.860 --> 00:02:50.240
There's gonna be, there's gonna be less traffic. Yeah. And that's been the lifeblood. How do you adapt to that? I think from our perspective, it's investing in intelligence.

18
00:02:50.560 --> 00:02:58.780
I think the first thing that you have to understand from a digital media publisher side is the days of not knowing the value of your audience and your content are kind of gone.

19
00:02:59.140 --> 00:03:08.850
Forever, we've never known what I would say the digital media publisher's KPI should be, which is revenue per session, and that's embarrassing for an entity to kind of admit that they don't know.

20
00:03:08.890 --> 00:03:16.120
Marketers forever have had CRM experts. They know their audiences, they know value of their users, the value of their products, their margins, et cetera.

21
00:03:16.680 --> 00:03:25.970
We've played an eyeball game for the past, you know, 30 years in digital media, where it's, you know, slap an ad on it, slap an ad on it, slap an ad on it, and there's a, you know, linear line- Mm-hmm...

22
00:03:25.970 --> 00:03:38.440
to revenue, and revenue per thousand. But at the end of the day, we don't know or haven't traditionally known, but we have invested in knowing what, what is the value of this content? What is the value of that cohort?

23
00:03:38.560 --> 00:03:41.980
And how do we grow and scale those cohorts in interesting ways?

24
00:03:42.000 --> 00:03:50.400
Because we're entering an era of, from a publisher perspective, of audience, and we need to know the value of those audiences that advertisers see the most value in. Yeah.

25
00:03:50.440 --> 00:03:57.880
So break down revenue per session, 'cause, I mean, that... What you're saying is... I mean, it's, it's a nuance- Yeah... but I think it's important. I mean, 'cause it was always RPM. Yeah.

26
00:03:58.180 --> 00:04:08.850
It's a god, what I call, like, a god metric, right? It's one of those things... Or a k- power KPI. It is the KPI from a digital media perspective 'cause it contains both volume and revenue, right?

27
00:04:08.860 --> 00:04:17.750
And so both of those things, when you're looking at them, it aligns the edit side and the product side along with the revenue side and monetization side of the house.

28
00:04:18.230 --> 00:04:25.960
And so when you look at both of those things, the interesting thing is we've always thought every page view was created equal in the era of programmatic, where- Yeah...

29
00:04:26.000 --> 00:04:34.280
advertisers are buying, picking off users, and we all say these things on all of these panels, but the reality is they're not, and we know that they're not.

30
00:04:34.480 --> 00:04:41.220
'Cause advertisers know the value of those audiences, and for the longest time, digital media publishers have felt like canvases, blank canvases- Yeah...

31
00:04:41.260 --> 00:04:49.610
that advertisers just paint on and, you know, "Give us a little more of this, give us a little more of that," and we're like, "Well, I don't know the value of that." I know content context.

32
00:04:49.700 --> 00:04:55.280
I know scarcity of The Big Bang Theory when I was at CBS, so we knew how to price that using tools like Yieldex and the such.

33
00:04:55.840 --> 00:05:05.600
But when it comes to audience and understanding, like, what should this cohort be valued at that this advertiser wants, GM, and we look at that and go, "Well, Ford is now interested in that."

34
00:05:05.760 --> 00:05:16.580
So we know that that cohort is probably an auto intender. We don't need to know their data, but we know that those are concentric circles that overlap, and there's value there for an auto intender, as an example.

35
00:05:16.740 --> 00:05:27.440
I think also, though, pricing those cohorts differently. Oh, 100%. So knowing your power, sort of power users or your- Right... highly engaged users. Right. It's kind of what you talked about. So- Yeah...

36
00:05:27.480 --> 00:05:34.560
in the past, it was all about scale. Right. So the more traffic you have- Well, a cookie is a cookie is a cookie. Exactly. That's right. Yeah. And, and the thing is, you don't know it. No. Like, when...

37
00:05:34.770 --> 00:05:41.110
I, I remember I used to, at Digiday, did this confession series that was delicious because- Yeah. [laughs]... people- Heard that one...

38
00:05:41.170 --> 00:05:46.360
people, people frequently don't tell the truth when they have their names attached to it, but they do if they are given ano- anonymity.

39
00:05:46.460 --> 00:05:55.140
But I remember one of the ones I did, it, it, it resonated with me, the person that said, "It's... We're just see a cookie, hit a cookie." Yeah. "That's, that's what this industry is." Eat a cookie.

40
00:05:55.150 --> 00:06:03.920
And the reality is the advertiser side knew- Yeah. Yeah... the, what w- or, or, you know, or the demand side knew who was on the other side of that cookie. Yeah.

41
00:06:03.940 --> 00:06:11.200
But for a publisher, right, like, it was just, like-Just traffic numbers. Yeah. It's just- Or impressions. Well, it was a unique- Yeah.

42
00:06:11.420 --> 00:06:20.000
[laughs] Well, when you start to align on the right thing, the right metric, like, you start coming up with great ideas. Like, we s- we started ideating on...

43
00:06:20.010 --> 00:06:24.390
'Cause like, let's talk about, go back to revenue per session. You're talking about session, how do you extend the session?

44
00:06:24.760 --> 00:06:33.100
When you start asking questions like that, you start thinking about, like, product experiences for the user. Video, vertical video, i.e. social video.

45
00:06:33.240 --> 00:06:38.400
We create it and we put it on the social platforms, but we have, like, small followers.

46
00:06:38.420 --> 00:06:46.300
So, you know, most digital media publishers have small followers on the social platforms, and we're making a pittance for that from the social media companies.

47
00:06:46.340 --> 00:06:52.180
Why not take that unit and the content that you invested in and move that onto your site?

48
00:06:52.220 --> 00:07:00.860
You have tens of millions of people on your site, and if you can have them consuming more of the content you're creating- In the way that's natural to them right now. That's right. Especially from the video perspective.

49
00:07:00.960 --> 00:07:10.720
Mm-hmm. Just 'cause I would say from a social aspect, that's, at least when you look at younger audiences, that's how they are entering the world of- Correct. Yeah... content consumption.

50
00:07:10.740 --> 00:07:21.650
It's through a vertical video format- Yes... through a social platform. Correct. It's not, frankly, unfortunately, reading, you know, a, a, a piece of written content in a- What? Johanna, don't tell me this...

51
00:07:21.660 --> 00:07:28.080
in an ar- in an article format anymore. And there will be people- Oh. [laughs] My audience is female, though. Yes. [laughs] I admit that my audience is not Gen Z. That's right. Okay.

52
00:07:28.260 --> 00:07:37.780
[laughs] So we have to please everyone. That's why I'm doing this podcast- Yeah... for my Gen Z- But, but-... listeners... I would, I would also say that podcasts and audio as well as video- Yeah... are, are on the rise.

53
00:07:37.980 --> 00:07:40.790
Yeah. I definitely think COVID was a boon. No, I have different segments. My little... Like- Yeah. [laughs]...

54
00:07:40.800 --> 00:07:48.200
you know, there's people who never read the newsletter, and there's people who never listen to podcasts, and they're just... You know, I think you just have to create different products for people. Absolutely.

55
00:07:48.240 --> 00:07:58.480
That's correct, and it's value, but, but ultimately user pathing. Yeah. Right? So co-horting is what we're... You know, when you talk about audience, basically co-horting. I like that. Cohorting. Is it- Yeah.

56
00:07:58.540 --> 00:08:06.450
Is it- Is that a w- is that a word? Is this a, is this a Gen- It's a word now, yeah. [laughs] Is this a Gen X per- [laughs] Yeah. Is this a Gen X person that wants to read and consume articles? Yeah. Yeah.

57
00:08:06.460 --> 00:08:15.280
For example, let's take them down an article path. Yeah. Is this somebody that wants, like- Personalization... premium articles? Yeah. Let's take them down a subs path. Is this somebody that's shopping? Yeah.

58
00:08:15.379 --> 00:08:22.320
Let's take them down an e-commerce path. Is this a younger audience- Yeah... that wants to consume video? Let's take them down a more of a video path. Yeah.

59
00:08:22.400 --> 00:08:31.650
But, and this is why I wanna get into the product question, look, everyone knows there's way more demand for video. The, the rates are better. Yeah. I don't think anyone's like- Amen... "You know what?

60
00:08:31.780 --> 00:08:39.540
I wanna serve more, you know, static display. That's where we're going." [laughs] Punch the monkey ads. Yeah. [laughs] That's what I want. Let's go back. What was it, like 480 by 60? Whatever. Let's bring them back.

61
00:08:39.580 --> 00:08:47.960
Yeah. But at the same time, how do you build the, the products when you have a page-based model? I keep asking people...

62
00:08:48.260 --> 00:08:57.389
We did a, a, a dinner last night where we did one, like, on Monday, and I was just asking them, like, "Do you really think that people are coming to, to webpages in five years?

63
00:08:57.440 --> 00:09:04.260
That they're, like, visiting a webpage, hitting a back button, and, like, that's gonna be the main way you're engaging with, with the audience?"

64
00:09:04.290 --> 00:09:11.370
'Cause I think a lot of times what I see is the TikTok experience is completely seamless, right? Yeah.

65
00:09:11.440 --> 00:09:22.480
And I think sometimes with publishers, it's how they're trying to chase, "Okay, look, this works," and everything like this, but we have a totally different container. Yeah. So how do you sort of bridge that?

66
00:09:22.800 --> 00:09:34.180
I think that you're, you've always challenged people to see ahead, and I think that you're spot on. We look at it as building the brand. First thing is understanding value, like we talked about.

67
00:09:34.240 --> 00:09:38.480
Once you understand your audience more and that value, then you can start doing things like...

68
00:09:38.800 --> 00:09:45.620
We, we saw, for example, that we were taken over by private equity, and we've turned the company around, and we're profitable now.

69
00:09:45.640 --> 00:09:50.389
And we're also taking a look at a lot of the brands in the portfolio, and we look at our venture network.

70
00:09:50.400 --> 00:09:57.640
There's great, you know, brands there like Powder Magazine and Surfer Magazine and Skateboarder that have loyal followings.

71
00:09:57.680 --> 00:10:10.920
And when you start to think about, you know, let's take that content and put it everywhere, social platforms, et cetera, 'cause you're right, the, you know, people potentially won't be coming to web, traditional websites in the future, but then also experiential type stuff.

72
00:10:11.000 --> 00:10:17.480
Once you become a brand, then you can start to ideate on that and go, "Let's do events. Let's do surfer events." You know?

73
00:10:17.720 --> 00:10:29.330
Surfers for the longest time have been reading Surfer Magazine, and back in the day, they did events. Why did that stop? Yeah. Let's revisit that. Athlon is known for a periodical that did the NFL Draft. Mm.

74
00:10:29.580 --> 00:10:45.860
Let's do draft day and, you know, make that be a brand experience for users and invite advertisers to it, have celebrities to it, you know, old players and the such, and really design that where you're providing the content for it, but that's content that's a module that can go anywhere.

75
00:10:46.360 --> 00:10:56.890
Yeah. Johanna, how do you think about that? I mean, with if... For a lot of publishers, I feel like the page is not gonna be the main monetization driver. I think it's becoming way more sophisticated. I do disagree.

76
00:10:56.900 --> 00:11:04.670
Yeah. Yeah, I would disagree with that, actually. Okay. And it, it was quite interesting listening to Jen and Lisa, and we kind of talked about this in our prep call as well. Yeah.

77
00:11:04.680 --> 00:11:17.370
But I think publishers fall sort of two sides of the coin. Either they're investing in building out their own technology products- Yeah... versus relying on, you know, like a DoubleVerify or IAS. Yeah.

78
00:11:17.400 --> 00:11:32.440
Obviously, brand safety is becoming such a huge topic, but when advertisers are relying on working with companies that basically do a really good job protecting brands against brand safety, but the publisher knows their content way better than anybody else.

79
00:11:32.580 --> 00:11:44.660
Yeah. So why not build your own technology that's gonna do a better job of doing that versus relying on some of the bigger tech platforms? So when it comes to video, I think it's the same thing.

80
00:11:45.200 --> 00:11:59.170
If you're able to have a seamless video experience that can kind of, I guess, hit the bar, meet the bar of what TikTok is doing, then why not leverage that and, and build a really good experience for, for the user?

81
00:11:59.680 --> 00:12:08.700
I think with video in particular, it was always a very difficult format because you have things such as streaming and bandwidth and, you know, loading issues.

82
00:12:09.240 --> 00:12:12.710
So I guess, you know, at, at XCO, that's what we doubled down on. Mm-hmm.

83
00:12:12.730 --> 00:12:25.272
And how, how do we give publishers the best tech?That's gonna give the user a very seamless video experience where there's no buffering, the right ad, the right user, the right content experience- Yeah... basically.

84
00:12:25.852 --> 00:12:37.932
I, I'll be interested in, in this m- i-if we assume that we are entering this more with less era, right? And, and obviously AI- Yeah... is, is here and, and it, it's gonna... It's not going [chuckles] away.

85
00:12:38.092 --> 00:12:46.272
Quotes by the way. Is, is where [chuckles]... Well, you know, it's used- Sure. It's used- It's thrown around, so [chuckles] A little bit- It's the new year of mobile. Broadly. Yeah. Exactly.

86
00:12:46.282 --> 00:12:55.462
[chuckles] But it's only Wednesday. Yeah. Yeah. [laughing] But what I, what I end up, I end up, like, wondering is, you know, it's, it's always build, buy, partner, right? Yeah.

87
00:12:55.472 --> 00:13:05.622
When it comes to, like, tech, and how that equation changes for publishers because, you know, we talked about this. Yeah. You gotta make tough choices, right? So- Yeah... I think a lot of publishers...

88
00:13:05.752 --> 00:13:16.532
I think some publishers will become B2B in some ways. Yeah. Yeah. Not a B2B in audience, but be suppliers, and just th- just be like, "You know what? We're gonna take... We're not gonna do any of it." Yeah.

89
00:13:16.542 --> 00:13:25.562
"And we're just gonna supply to the people who..." I mean, you see with, like, Perplexus, like, we're just gonna be... We can be a supplier, and I think that's a path, right? I think- Yeah...

90
00:13:25.592 --> 00:13:32.112
you know, you're- We're really good at creating content, and that's- Yeah... what we're gonna do, and you get content- Just stick to what you're- Yeah... and you get content, and you... Yeah. Yeah. Yeah.

91
00:13:32.152 --> 00:13:42.162
I mean, I think that's- Yeah... you know, it's a valid strategy. Yeah. I don't know if it's... It's obviously not right for, for all brands, but how do you see that, that, you know, people making, like, those decisions?

92
00:13:42.202 --> 00:13:44.352
'Cause every, every path- I would-... has trade-offs. Yeah.

93
00:13:44.392 --> 00:14:00.192
I would be sort of a bit cynical in that sense of, I think it depends on what stage or, you know, the business itself, do they have capital to be able to invest in their own technology? Yeah.

94
00:14:00.232 --> 00:14:06.892
You know, the publishers- Nobody's talking to me about their fancy CMSs anymore. No, no. I, I- I, I just [chuckles] I mean, publishers are definitely under duress. They just did WordPress.

95
00:14:07.052 --> 00:14:14.752
Either, you know, you're very strapped and you're outsourcing- Yeah... a lot. Which is mostly the case. Or, you know, or you're owned- Yeah...

96
00:14:14.772 --> 00:14:23.052
by a, you know, a bigger, a bigger company that has deep pockets, and I think you're gonna be more on the sort of build side. For the things that are endemic and natural to your- Yeah... cause.

97
00:14:23.112 --> 00:14:34.382
Like w- this is one thing that we s- uh, did at CBS and we're replicating at The Arena Group. When we came in at CBS, we were making s- you know, $4 in revenue and $6 in costs. Like, it was all upside-downville. Yeah.

98
00:14:34.432 --> 00:14:40.472
And, and we were building our own tech. Yeah. Built our own ad server, order management system- Yeah... forecasting engine. But why? It's like building your own serum.

99
00:14:40.492 --> 00:14:46.972
[laughs] You're not an ad tech company, and you're not even a public media company. You're nestled up underneath, you know, the bosom of, of CBS.

100
00:14:47.392 --> 00:14:54.522
So once we realized, you know, who we were, we- we're a premium content company, and when you start to identify... You start to understand- We're not a CRM...

101
00:14:54.612 --> 00:15:00.272
your identity, exactly, then you can start to ideate on things like product that we should invest in, like CTV.

102
00:15:00.632 --> 00:15:11.812
You know, and that's how we came to be the first 24/7/364 news app with CBS News, CBS All Access, creating a subs product, because that's where we knew the future of consumption was gonna be.

103
00:15:12.072 --> 00:15:22.052
I, I think you have to stop making it complicated. You have to boil it down and go, "What am I?" And then you ideate on that, and when you ideate on that, you come up with ideas like you're talking about.

104
00:15:22.092 --> 00:15:30.852
Take that premium content and distribute it in all the places that people are consuming it. Mm-hmm. But what do you end up not, not investing in? 'Cause I mean, you have to... I mean, we talked about hard choices. Yeah.

105
00:15:30.892 --> 00:15:41.612
But give me a couple examples of l- of a hard choice that you had to make. I, I think that you don't invest in certain things like certain technology components. Like, you mentioned CMS. CMS is a great example.

106
00:15:42.172 --> 00:15:51.172
It's very easy for a company to go, "Oh, well, we're publishing content." So why not build my own? "We should have the thing that controls the thing-" But also, why reinvent the wheel? Right. You know? But it...

107
00:15:51.292 --> 00:16:00.892
And where does that end? Yeah. Yeah. Right? And it doesn't. A- and then the next thing you know, you kinda look at it and, and go, "Oh, well, that was ridiculous," and we spent a ton of money on that.

108
00:16:00.952 --> 00:16:10.152
Print, you know, we kind of took a look at some of the traditional print stuff. We had Sports Illustrated, which was one of the problems with that brand. When we looked at it from a, you know, business perspective, i.e.

109
00:16:10.272 --> 00:16:18.272
making money, that was hemorrhaging money, right? And that's something- Yeah... that we shouldn't support that anymore. That's silly. Let's not do that.

110
00:16:18.752 --> 00:16:28.132
So I think, you know, stuff like that is kind of anecdotal examples. So what's core then in this world? I mean, I would guess, like, things that are closest to the, the, the audience are core.

111
00:16:28.152 --> 00:16:37.852
Like, how you- Core to the user experience. Yeah, understanding- Yep... the audience. Like how... I, I, I could see a case where, like, if you're gonna, like... You've got a finite amount to, to invest in- Yeah...

112
00:16:37.872 --> 00:16:45.142
that that would be kind of core. It's stu- it's stuff around CDPs. It, when you start- Yeah... to, we're say audience, audience, audience, audience, so let's unpack some of the things that you invest in.

113
00:16:45.292 --> 00:16:53.892
CDPs is an area that we're investing in. I think those things are very relevant for publishers now. They've been relevant for marketers for years. Yeah. Yeah. Now publishers need to lean into those.

114
00:16:54.332 --> 00:17:05.132
We're gonna build LLMs. If you think about it, publishers come and do click actions on sites, whether they're reading 1,000, 5,000 articles. Like, to your point, that's gonna go away slowly, but watching videos- Yeah...

115
00:17:05.152 --> 00:17:20.002
understanding metadata, e-commerce, understanding those click actions, a- all of that information, storing that, and then, you know, doing LLM models on top of it to do smart user pathing to go, oh, this per- like we were talking about earlier.

116
00:17:20.132 --> 00:17:29.412
They need to go here. They need to go there. They need... That's something that a publisher should invest in. Now, there's licensing that comes along with that. We're not gonna build our own CDP.

117
00:17:29.472 --> 00:17:35.462
We're gonna license that. Mm-hmm. But there's certain things like, you know, CRM, data scientists, and the such- [chuckles]...

118
00:17:35.482 --> 00:17:43.482
that you should invest in, and that's more human powered capital than it is, like, building technology. It's the people that know how to use the technology- Or how to analyze...

119
00:17:43.482 --> 00:17:49.832
architect the databases- Yeah, or how to analyze it... schemas- Yeah... building those, message buses back and forth, you know, in those databases.

120
00:17:49.912 --> 00:17:58.402
But I think people that understand data and know how to use it and present that for a more elegant user experience, to your point, it is the winning recipe, I think.

121
00:17:58.412 --> 00:18:07.752
So do you end up becoming almost like a general contractor- Yeah... when it comes to technology? Basically. Yep. You knowing how the things, like- Mm-hmm... link together. So how does... I, I have to ask you a question.

122
00:18:07.792 --> 00:18:15.292
Sorry, Joanna. It's okay. Go for it. It's, like, legal. [laughs] It's legal. I, I'm legally obligated to [chuckles] to ask. Because I wonder, 'cause I get...

123
00:18:15.352 --> 00:18:25.532
I think in a lot of conversations I have, things are changing so much, right? Yeah. And that can be, that, that's scary and maybe exciting, but it could also lead to some, like, paralysis, right?

124
00:18:25.632 --> 00:18:35.352
Because you, you don't even know what to build, right? Like, if you see the advances that are happening seemingly weekly, if not daily, and it could be a little bit paralyzing. What do...

125
00:18:35.752 --> 00:18:44.832
How do you think whatever is to come with this AI era, how does that impact how publishers think about- Everything? [laughs]... their strategies? Well, I mean, there's...

126
00:18:44.992 --> 00:18:52.102
L- look, we, we focus a lot on the, the, the search side, the distribution side- Yeah... et cetera, and, like, a lot of that stuff's out of your control at the end of the day.

127
00:18:52.452 --> 00:19:04.192
But you can control, you know, uh, your own strategy and, and how, and how you react to that. So how do you see AI changing the sort of decision-making that publishers are making with what to build? With what to build?

128
00:19:04.552 --> 00:19:15.000
Yeah, we talked a little bit about this in terms of does it make it easier for publishers to build their own tech to a degree?But I can see that being like, you know- Potentially... potentially- Of course...

129
00:19:15.400 --> 00:19:30.120
attractive But I think with AI it can be used for so many different things, right? It can be used to automate, it can be used to scale, it can be used to, I guess, become more efficient- Yeah... as a business.

130
00:19:30.480 --> 00:19:35.560
But is that where you see, both you see the, the biggest impact, the positive impact for publishers, right?

131
00:19:35.680 --> 00:19:45.560
I think a lot of times it's viewed as a threat in this, 'cause there's, it's reality that, that, there are, you know, a lot of negative implications. I think you have to jump on the opportunity. Yeah, yeah. I, I agree.

132
00:19:45.570 --> 00:19:55.580
A- as with, as with- Yeah... you know, the way this... So I used to sort of be more on the influencer marketing side- Yeah... back in 2010 when YouTube was just becoming a, a format.

133
00:19:55.940 --> 00:20:05.250
Brands were not working with, you know, influencers. We had to do a huge education piece. Mm-hmm. Now that's arguably- Yeah... all that you see on the La Croisette is- Yeah, but influencers-...

134
00:20:05.250 --> 00:20:12.670
you know, influencer activation... this place is crawling with- Exactly... influencers and athletes. And I think, I think- A lot of tall people here. That are making good money. [laughs] Tall people.

135
00:20:12.720 --> 00:20:26.899
[laughs] Should've worn my heels. I think with AI it's gonna be one more of these sort of fundamental shifts in the way that the internet [laughs] Yeah... works, if that makes sense, and I think we have to jump on it.

136
00:20:27.080 --> 00:20:33.620
Yeah. So at least on the XCO side, we've doubled down on the optimization or yield side.

137
00:20:33.940 --> 00:20:42.740
We've always been very supply side focused, so how do we give tools to publishers to help them leverage their cohorts and, and make them more valuable?

138
00:20:42.820 --> 00:20:52.950
There's a lot of technology on the advertiser side, to what you said earlier, of- Yeah... them having all this data and understanding the users, but the publishers weren't looking at it in that- Well, that's always-...

139
00:20:52.950 --> 00:20:53.920
similar way... been the challenge, right?

140
00:20:53.980 --> 00:21:04.860
I mean, it feels like particularly with audience data, like h- And I think with AI you can do a lot of clever things where in the past you'd probably have to hire a significant amount- Mm... of data scientists- Yeah.

141
00:21:05.040 --> 00:21:11.650
And o- That's right.... people in order to- And that doesn't, that do-... to do that... and that doesn't scale. It doesn't scale. No. It doesn't scale basically. That got, that got us in the problem that we're in.

142
00:21:12.260 --> 00:21:14.400
We've leaned into... Sorry, you- No, no, no.

143
00:21:14.460 --> 00:21:23.260
Go ahead, because I wanna get into the internal question, 'cause I think one of the questions with the more with less a- and when it meets with AI is how do you use it to be more efficient?

144
00:21:23.280 --> 00:21:30.680
I don't mean to try to, like- No, the, but the analysis paralysis-... cut a bunch of ad ops people... that you brought up is important. [laughs] No, no, but, but I think that it... So a couple of different things.

145
00:21:31.200 --> 00:21:40.349
The oth- yeah. Analysis par- paralysis is- Y-... definitely something we- And stop doing stupid things, right? Uh, but- Yeah. [laughs] Usually a good strategic framework. Yeah. Yeah, yeah.

146
00:21:40.349 --> 00:21:46.080
Make decisions, but don't, don't make stupid decisions basically. That's right. Yeah. [laughs] That's right. Sounds easier said than done. [laughs] I mean, we, we lean into...

147
00:21:46.640 --> 00:21:57.220
So let's go back to understanding your KPIs and analysis paralysis. Once you understand your KPIs, you can... And your operating levers. B- th- there shouldn't be more than, like, 20 or so.

148
00:21:57.500 --> 00:22:10.580
But there's one golden KPI, but then the l- levers that move that KPI, and you start ideating around that, and that should take you down a path of, you know, having a list of things that you wanna work on, and then you cut that list down, no more than five things.

149
00:22:10.600 --> 00:22:23.740
Like, all of those institutional things to streamline stuff. But once it starts getting complicated, uh, you g- you have to cut that out. Y- you have to laser focus everybody on, we need to extend sessions on our site.

150
00:22:23.750 --> 00:22:29.800
Mm-hmm. Okay? How do we do that? What are the product ideas around that? Mm. Vertical videos. Mm. And, okay, let's write that one down.

151
00:22:29.860 --> 00:22:38.300
And you start writing those ideas down, and then you know the things that you can get done in a quarter- Mm-hmm... or six months, and then you invest in those things. So that's how we look at it.

152
00:22:38.400 --> 00:22:46.580
AI from an efficiency perspective, we look at the ad operations side. Yeah. And, you know, uh, uh, I'm sorry, but, like, people that fear certain things that...

153
00:22:46.620 --> 00:22:54.550
That's like the accountant fearing Excel back in the '80s and saying- Or like redline technology-... that's gonna take my job... for lawyers. No, it's a tool. Yeah. [laughs] It's a tool. Yeah.

154
00:22:54.550 --> 00:23:03.890
And the smart accountants that knew how to do, you know, the pivot tables, and the joins, and all of that were ones that were more- Only had to work 12 hours a day versus 24 basically... precisely. Yeah.

155
00:23:03.890 --> 00:23:10.340
And they had a better work experience- Yeah... and their clients had a better experience as a byproduct of that. It's the same thing with, you know, ad operations.

156
00:23:10.400 --> 00:23:22.380
We do order-to-cash automations, and we lean into those technologies, built a company around that, Jiffy, that ultimately takes the order-to-cash paradigm, that last mile for digital media publishers, especially in the direct side.

157
00:23:22.680 --> 00:23:29.050
You have an insertion order that comes in on a PDF. You have to take that information, you have to put that in the order management system, the CRM, and the ad server. Yeah.

158
00:23:29.080 --> 00:23:37.490
The same information, and we expect human beings to do this job? Yeah. That's a terrible job. Yeah. Terrible. That's absolutely something that should be automated. Yeah. Yeah.

159
00:23:37.520 --> 00:23:45.530
Doing creative checking, [laughs] you know, on a site, and screenshotting, and sending that to a client. A human being still does that job. That's something that can be fully automated. Yeah.

160
00:23:45.540 --> 00:23:51.780
So you have to look at the positive applications in the use case- Absolutely... that can make your business, A, more efficient, and your workers happier. Yeah.

161
00:23:52.360 --> 00:23:59.280
The, it's the, the cognitive ma- I love the term co- cognitive manual labor. Yeah. Like, because- Yeah... that's reality, and I think a lot of times

162
00:24:00.220 --> 00:24:10.410
those of us, like, on the, you know, the, the laptop class, we love automation when it, when it, it- Yeah... it hits, you know, more analog world- Yeah. Yeah... of, you know, [laughs] it's like factories and- [laughs]...

163
00:24:10.420 --> 00:24:17.910
you know, with globalization we're like, "Yeah, well, that's just how the economy works." Yeah. It evolves. And then all of a sudden, all of a sudden the technology that comes after, like, the laptop- Yeah...

164
00:24:17.910 --> 00:24:25.060
class, we're like, "Let's stop this." But, but let's slow it down a second. Let's slow this... [laughs] Yeah, wait a second. I don't like this. [laughs] But yeah, I mean, it's inevitable. I think- Not in my backyard.

165
00:24:25.140 --> 00:24:35.570
Yeah. [laughs] [laughs] On all sides of the media industry, there's just... I would love to know a global figure of, of how much money actually goes towards transferring data from one system to another system. Mm.

166
00:24:35.620 --> 00:24:41.240
So I- [laughs] A lot... I mean, I won't, I won't name the event and who, who the person was, but I heard...

167
00:24:41.280 --> 00:24:51.650
So obviously on, on sort of, like, the news side, the brand safety, and allow lists, and block lists is a huge topic, and nobody knows how to- It is... get on or off one of these- It's set it and forget... but, so- Yeah.

168
00:24:51.660 --> 00:24:57.240
A 25-year-old sets it- But, but what's fas-... and then they forget it, and then you wonder why your campaign doesn't work... but you know what's fascinating? Yeah.

169
00:24:57.250 --> 00:25:09.040
Is basically, uh, an senior agency exec said one of her biggest pain points is the fact that for each campaign and brand- Yeah... they have a different obviously setup in six different DSPs. Yeah.

170
00:25:09.080 --> 00:25:27.578
Each DSP and each brand has a different allow list and block list that somebody manually has to go and update in basically six different systems per client.And that's why you never get off a block list because nobody even knows in what system or what list- Yeah...

171
00:25:27.928 --> 00:25:33.528
where it actually lives. So that's, that's- And a ban played on... that's an area where automation, yeah. That is what they mean by that statement.

172
00:25:33.548 --> 00:25:40.007
And they probably, they probably get- It's like, oh, don't worry about that... you know, 100- That is the problem... 100 emails per day. Yeah. Like, "Hey," Yeah... "I think I may be on a block list."

173
00:25:40.028 --> 00:25:49.328
And nobody- Houston, we have a problem. We have a problem, yeah. I will, I will- So-... obscure the publisher name, but a very prominent business publisher was handed a 4,000-term block list- Block list...

174
00:25:49.868 --> 00:25:56.878
in which, in which they- their name was on it. You should ask AI to try to write an article with that without using one of those words. [laughs] The literal name- And see what it came up with...

175
00:25:56.878 --> 00:26:03.288
of the publication is on that. Yeah. Well, I, we've had a client in the past group black. I mean, think about the, this- But these lists, these lists-... the DEI category.

176
00:26:03.348 --> 00:26:11.508
These lists- Black Lives Matter was blocked- Yeah... on a Black-owned media sites. Domestic violence, yeah. Well, yeah, they're covering that. Because nobody's- But they're not covering that in a hate speech way.

177
00:26:11.528 --> 00:26:14.708
Nobody's, nobody's taking words off. Right. They're only adding. That's correct.

178
00:26:15.008 --> 00:26:26.728
And, and that, but that pro- prohibits, you know, somebody wants to invest in a DEI campaign to run on a, a, you know, a Black-owned site, like, that prohibits them from running on there because they had a block on that the entire time.

179
00:26:26.748 --> 00:26:37.628
There does need to be a better process, and automation can be applied to that as well, for sure. But I think that that's an easy problem to fix that we should- I- Yeah... I think AI could do a lot with that problem.

180
00:26:37.658 --> 00:26:44.618
100%. Try to automate that. 100%. So what are the hardest problems to fix? Stuff like that. Yeah. [laughs] I think. Which sounds really simple, actually.

181
00:26:44.808 --> 00:26:52.218
[laughs] Well, but, but they're, they're throwing a ton of human capital against it- Yeah, 100%... to, to do it- Yeah... to start it in the first place, but then they forget it. Yeah.

182
00:26:52.568 --> 00:27:05.528
So you need to apply technology to then fix that. I think some of the harder problems are... So from a digital media publisher perspective, Google Chrome's... Sorry, Google core update. Ground kind of shifted.

183
00:27:05.588 --> 00:27:13.828
You were just quoting numbers at the top of that. Mm. So to bring it alpha to omega, we're investing a ton in our audience teams to understand that- Yeah... fully. Mm.

184
00:27:13.928 --> 00:27:19.388
How should we re-architect all of our sites, which we are doing. But again, it's all about brand.

185
00:27:19.688 --> 00:27:30.397
If you look at that core update, you know, they even take it down to the level of understanding pronoun usage at a hyperlocal level, and so I'll use this as an example. Wait, explain that. Sports site.

186
00:27:30.397 --> 00:27:40.058
So a sports site, so we have Athlon. It's the same model as, like, SI. So we have national desks, and then we have local desks. And on the local side, 'cause we can only scale so many- Mm-hmm...

187
00:27:40.058 --> 00:27:47.928
writers so much, we strike partnerships with people that are in those local markets, dallascowboys.com, denverbroncos.com- Mm...

188
00:27:47.948 --> 00:27:57.328
atlantabraves.com, and put those on the subdomain, where we tap into those guys for what their expertise are. They're, they're editors. They're there. They're on the scene. I was here.

189
00:27:57.428 --> 00:28:09.888
I was hanging out in the Dodger locker room. We went over to the Dodger game and covered this. Like, the nuanced level, the new Chr- the new core updates love that because that's something that's relevant.

190
00:28:09.928 --> 00:28:19.068
It shows investment. It shows relevancy for the users, which is gonna equal a better user experience. But those are things that you have to invest in- Yeah... in order to win.

191
00:28:19.088 --> 00:28:30.348
Well, that's why I, like, sort of, I don't know, I go back and forth over, like, things like CouponScape because on the one hand, look, it's a di- I mean, like, I understand it. You know, you set up a subdomain.

192
00:28:30.528 --> 00:28:41.048
They supply. Yeah. Like, it goes straight to the bottom line. It left... I was told, like, at one dinner, like, a publisher lost a $7 million hole in their balance sheet. Mm. Yeah. I mean- Yeah... that's tough. Mm-hmm.

193
00:28:41.068 --> 00:28:48.978
Yeah. And then they're going to appeal to, to, [laughs] to Google, which is also- Yeah... I think the, the sort of guise- [laughs]...

194
00:28:49.008 --> 00:28:55.938
of just hiding behind the algorithm, the fact that, like, you have to go in, in person to appeal, and then the meeting got canceled. Yeah. Yeah. [laughs] Shocker.

195
00:28:56.288 --> 00:29:05.098
But at the same time, like, I understand, like, publishers have resp- responded to incentives in the market. That's right. And those incentives have been set up- Yeah... by, by the people on the beaches- Yeah...

196
00:29:05.128 --> 00:29:11.918
at the end of the day. Correct. And then they're blaming the, the publishers responding to the incentives that they set up. Yeah, that you set up. Yeah. Why are you doing that? Yes.

197
00:29:11.918 --> 00:29:20.388
[laughs] It's the game that you- That you've... you set the terms of the game. [laughs] Yeah. So I mean, that's just, uh- That's like passing Go in Monopoly and going, "You are mad at me for passing Go?

198
00:29:20.468 --> 00:29:24.908
I thought we were supposed to pass Go." That was, that was, that was the goal. Like $200, right? Yeah. [laughs] Like... [laughs] Yeah.

199
00:29:25.208 --> 00:29:31.128
But I guess that c- comes into one of those things is, like, you control what you can control. Yeah. Yeah. And- That's right... you know, I don't know.

200
00:29:31.188 --> 00:29:40.388
There's a lot of intergalactic battles going on between technology giants and, and governments and, and there are things that- Yeah... publishers really can't impact that much.

201
00:29:40.468 --> 00:29:48.608
They can't, and you shouldn't get distracted by that. Like, you should understand the outputs of those things because those are tectonic, like, shocks to the...

202
00:29:48.648 --> 00:29:54.568
And it can impact your business, but just understand those things once they, once they settle, how can it impact your business?

203
00:29:54.908 --> 00:30:06.188
Develop your strategy, and then laser focus on executing on that strategy like we talked about. Yeah. I think agility, so learning to become more agile businesses. Yeah.

204
00:30:06.408 --> 00:30:16.468
I think historically, media companies are quite set in their ways- Yeah... in terms of processes or sort of just quarterly planning or annual planning. I think you have- And overcomplicating it...

205
00:30:16.868 --> 00:30:23.098
I think you have to be more- Yeah... agile and how a little- I think it's underrated how conservative- A little bit more op-... at small C- You know, things can, you know-... this industry has been...

206
00:30:23.138 --> 00:30:34.308
we onboard large- Very much... we onboard large- Yeah... premium publishers, and that process, you know, sometimes it's one developer that needs to put one piece of code. It can take months.

207
00:30:34.408 --> 00:30:45.508
So I think agility and- Yeah... being a little bit more opportunistic rather than sort of, I guess, fear, fear-mongering around- Yeah... the changes that, you know, we've talked about the cookie going away- Yeah...

208
00:30:45.518 --> 00:30:49.848
for the last- I'll, I'll handle the fear-mongering. Yeah. Don't worry. Yeah. [laughs] No, he's got us covered. Yeah.

209
00:30:49.958 --> 00:30:59.628
[laughs] But no, but gone are the days of, of fear-driven organizations being overly conservative- Yeah... and building, you know, fiefdoms and the such and creating bureaucracies.

210
00:31:00.228 --> 00:31:05.988
This era that we're entering in is the- Have to move fast... anti. Yeah. Exactly. Have to move fast. Correct. Absolutely. And that's gonna- Create bureaucracy at your peril.

211
00:31:06.268 --> 00:31:15.528
Yeah, it's gonna break a lot of that, I feel like. Yeah. Because, I mean, there's just some crazy stuff that still exists. Yeah. You know? Like a, a big newspaper, you know, they, you know, commerce.

212
00:31:15.588 --> 00:31:22.528
You know, they set up a commerce operation, and they, and then they're fighting. Like, the newsroom says, "Ah, you can't do the- Yeah... you can't do the credit card reviews."

213
00:31:22.628 --> 00:31:31.648
Do you know what kind of bounties, like, a credit card? Yeah, exactly. They're like, "But our personal finance team, you know, does that as part..." And it's like, you, you- Yeah... can't afford that stuff.

214
00:31:31.868 --> 00:31:42.178
By the way, if they knew more about value- Yeah... if everybody rallied inside that publisher around revenue per session, and that's the golden god- Yeah...

215
00:31:42.178 --> 00:31:50.578
it doesn't matter if you're an editor and your words are beautiful, and everything that you write is amazing. Don't say that, Jason. Well, we know that that's how they have viewed it. I know. I know.

216
00:31:50.578 --> 00:31:56.958
But you have to understand value, and that get rid of the conch shell. Look at data, and let that be your guide. Yeah.

217
00:31:56.978 --> 00:32:06.328
And get rid of all of that bureaucracy 'cause when you do that, a lot of that silly behavior goes away. Yeah. Okay, cool. Well, thank you. You use your words beautifully, by the way. [laughs] Yeah, she does.

218
00:32:06.368 --> 00:32:22.728
Thank you very much. She's absolutely does. Yeah. Okay. Well, thank you, Johanna. Yeah. Thank you, Jason. Thanks, guys. It was great. Appreciate it. Thank you. Bye. Thank you. [outro music]