WEBVTT

1
00:00:00.280 --> 00:00:08.440
[upbeat music] Welcome to the Rebooting Show. I'm Brian Morrissey. This has been a terrible start to the year for the media industry.

2
00:00:09.020 --> 00:00:16.660
Media layoffs have topped three thousand in the past year, and just since the start of the year, we've seen layoffs at a broad array of publishers.

3
00:00:16.720 --> 00:00:36.020
We've seen pullbacks, we've seen publications like Pitchfork get folded into GQ, and I would say general disarray and chaos in the industry, perhaps to a degree I have not witnessed in my twenty-odd years, both covering this as a journalist and analyst and also living in it, since this is my career.

4
00:00:36.660 --> 00:00:43.100
So I will totally cop to being biased to wanting the creation of a sustainable media ecosystem.

5
00:00:43.449 --> 00:00:51.140
And I think this past week and just the start of the year, and really the past year has proven that we've got a long ways to go.

6
00:00:51.780 --> 00:01:04.840
And I wanted to get at the root causes of what's happening right now, because this tummel is coming during what's been, by all measures that I can see, a fairly resilient economy.

7
00:01:05.000 --> 00:01:14.510
Now, anytime I say that or write that, I get emails telling me that there are people struggling out there and losing jobs, inequality is a problem, and I totally understand that.

8
00:01:15.240 --> 00:01:25.440
But that aside, there is a disconnect without a doubt right now between what is happening in the broader advertising market and what is happening in the media sector.

9
00:01:25.520 --> 00:01:35.460
You see this if you look at the advertising holding company results. You see this if you look at the platform's revenue and profits. Okay?

10
00:01:36.020 --> 00:01:47.090
And so that tells me that there are structural shifts happening to this market rather than the typical cyclical challenges that are, uh, just the nature of markets. They go up, they go down.

11
00:01:47.780 --> 00:01:59.340
And that's why I wanted to have on Scott Messer to talk about these changes. Scott is a longtime digital media executive who understands the mechanics of the digital ad industry as well as anyone I know.

12
00:01:59.840 --> 00:02:25.120
I always appreciated Scott personally when I was at Digiday because he not only showed up at the content sessions at our events, uh, instead of hanging out at the pool bar or hitting the slopes like some might have done, but he was an active participant who, for lack of a better term, loved geeking out on programmatic and the complicated but critical infrastructure that underpins modern advertising and media.

13
00:02:25.400 --> 00:02:37.220
So very happy that Scott would take the time to have this discussion with me. If you like this show, uh, please support it by giving it a rating, a review, or even recommending it to a colleague.

14
00:02:37.280 --> 00:02:42.680
I would appreciate that personally. I also wanna thank, uh, Outbrain for being a sponsor of the Rebooting Show.

15
00:02:43.140 --> 00:02:50.300
Outbrain is working to help publishers improve their ability to monetize their audiences based on the attention their content earns.

16
00:02:50.860 --> 00:03:01.220
Onyx by Outbrain combines unique prediction technology and engaging high-impact formats carved out within exclusive in-article placements across premium publisher environments.

17
00:03:01.600 --> 00:03:07.540
Now, this is critical because we are arguably moving from an era of viewability to one built around attention.

18
00:03:08.020 --> 00:03:19.780
Viewability has long been one of those important, but in my view, and this is my view, flawed metrics that quickly became game to the point where its value as a proxy for effectiveness deteriorated.

19
00:03:19.820 --> 00:03:31.900
And that's why attention is now gaining momentum as the metric of choice that delivers a stronger view of the return on ad spend that campaigns deliver than viewability alone does.

20
00:03:31.920 --> 00:03:43.299
Now, Outbrain has over fifteen years of experience in building technology to predict user engagement, and it's created Onyx in order to apply that same AI-driven approach to predict moments of attention.

21
00:03:43.740 --> 00:03:54.280
This will help advertisers get better ROI and publishers monetize their content more effectively, and that would be nice because we need a sustainable media ecosystem to emerge from this turmoil.

22
00:03:54.880 --> 00:04:10.460
Check out Onyx's Attention Playbook. I will leave the link in the show notes, and it'll also be in the newsletter, which I hope you all read. And to find out more about Onyx, visit outbrain.com/onyx, that is O-N-Y-X.

23
00:04:11.040 --> 00:04:17.760
Thank you so much to, uh, my friends at Outbrain. They've been, uh, great partners over the last couple years, really throughout my career.

24
00:04:17.860 --> 00:04:29.300
And, uh, we actually just recently collaborated on nice dinner in, in Las Vegas during CES, everyone's favorite event, a-along with our friends at, at Adelaide, and that was great. Hope to do more of those.

25
00:04:29.980 --> 00:04:40.160
So here now is my conversation with Scott. [upbeat music] Welcome to the podcast. I'm glad we're finally doing this. Thank you. Glad to be here, Brian. Yeah.

26
00:04:40.440 --> 00:04:50.740
We should call-- This should-- This will be like a pilot episode of Ad Tech Therapy. So, you know, I came across an article you wrote in AdExchanger. I'm, I'm a sucker for a good cheery headline.

27
00:04:51.140 --> 00:05:00.520
And I don't know, maybe it was the editors at, at AdExchanger put it. You put, "The Precipice of Peril: Publishers and Ad Tech Face a Harsh Two Thousand Twenty-Four."

28
00:05:00.640 --> 00:05:10.740
In here I had a resolution to be resol positive in the new year, and then you just dragged me down. Definitely had some help from the editors at AdExchanger. Thank you very much, Anthony.

29
00:05:10.960 --> 00:05:22.640
But I would be lying if I said that I didn't think about you- I know... when this headline hit, and that this is the- Yeah. This is a Brian title right here. I'm sorry to knock you off your glowing sword.

30
00:05:22.720 --> 00:05:30.660
No, I joke that there's two types of publishers I would talk to about twenty twenty-three, the kind that told me it was a shitty year, and the kind that lied to me. [laughs] That's great.

31
00:05:30.690 --> 00:05:42.680
'Cause look, it was a difficult year for most publishers, right? And no matter what they might say in public, like it just was. And I think it's worrying because that's why I thought y- your piece was really good.

32
00:05:42.690 --> 00:05:48.140
It's 'cause there's a lot of structural challenges that, uh, is facing the publishing industry.

33
00:05:48.180 --> 00:05:58.060
It's one thing, the cyclical, the economy, you know, there's lots of stuff go-- it's boom and bust, particularly if you're reliant on advertising. Let's start with traffic, okay?

34
00:05:58.160 --> 00:06:06.080
And traffic has been at the front end of this industry for a generation. Business models were built on traffic, monetizing traffic.

35
00:06:06.140 --> 00:06:16.660
There are playbooks out there for getting traffic, for making money off traffic, et cetera. I don't know, I mean, the traffic is down quite a bit already, right? [laughs] Yeah. That's the biggest challenge.

36
00:06:17.060 --> 00:06:22.420
You know, to those people, I would say like, I would always say in monetization, if you give me a page view, I will give you revenue.

37
00:06:22.430 --> 00:06:31.360
The single hardest thing to do in all of digital media and really in media, if you think films are as well, is making an organic page view.

38
00:06:31.420 --> 00:06:39.284
It's super hard.You are getting somebody to do something that they weren't tricked into doing, that they're doing out of their own volition.

39
00:06:39.364 --> 00:06:46.564
And when traffic dips, it usually hits the bottom line really hard, especially for the SEO publishers.

40
00:06:46.604 --> 00:06:54.324
The search traffic is off a cliff right now, and people are extra scared because of what generative AI has in store for it.

41
00:06:54.984 --> 00:07:03.054
Social publishers have abandoned the newsfeed and turned them into just wastelands of traffic-driving places.

42
00:07:03.504 --> 00:07:14.564
And so you're really left with newsletter, direct traffic, referral, which is like texting people links, and then you start running out of, like, how you get traffic real quickly.

43
00:07:14.964 --> 00:07:26.584
And without traffic, you can't make- Yeah... CPMs. Yeah. And so I don't know if you saw, like, Red Ventures is looking to sell CNET, which is not a shock. But, um- Half of what they bought it for, right?

44
00:07:26.924 --> 00:07:35.754
The graph [laughs] yeah, the half in a couple of years. The graph that accompanied the story that Sarah Fisher wrote was chilling. Mm-hmm.

45
00:07:35.784 --> 00:07:45.224
You know, and it's from SimilarWeb, not from Comscore, but it just showed the global uniques just a couple of years ago were at, like, 130 million, now they're down at, like- Yeah... 30 million.

46
00:07:45.284 --> 00:07:53.384
You kinda wonder, like- Um-... where did the traffic go, right? If it's not going to- You tell me... these publishers- Where did it go? Then where... I think a lot of it stays with Google.

47
00:07:53.564 --> 00:07:58.584
Google still seems to be making money. Jason Kint used to share- [laughs] Well... a chart a long time ago.

48
00:07:58.644 --> 00:08:10.634
It either isn't generated, it's staying inside the social platforms, TikTok, Instagram, Snap, or it's staying within Google. That's the thing is it's not, no one's cross-pollinating.

49
00:08:10.664 --> 00:08:21.104
Like, you're not moving users from one place to another. Before social and before Taboola, getting traffic was also really hard. People forget that.

50
00:08:21.224 --> 00:08:30.824
In 2010, the only way you could get traffic was, like, buying links from people, doing pop-unders. It was really hard and really shady.

51
00:08:30.984 --> 00:08:38.854
And then content circulation, like Taboola and Out gave publishers a method to do it. The earliest ones were about content circulation.

52
00:08:38.884 --> 00:08:56.944
And then the social channels sprung a well of traffic to everybody, and it not only created growth for publishers, but it gave rise to, like, the entire D2C industry because now there were these centralized pools, ponds where you could fish out of, and these closed loop ecosystems.

53
00:08:57.404 --> 00:09:01.174
But now they're shutting off, and, like, you're seeing DTC- Yeah... die as well.

54
00:09:01.284 --> 00:09:12.044
It's not because people don't like Allbirds as much, which is true, but they can't advertise as much, and the CAC values are upside down and, you know, venture money's hard to come by.

55
00:09:12.104 --> 00:09:18.054
So traffic just doesn't hurt, does, it doesn't- Yeah... just hurt publishers. Right. No, 100%.

56
00:09:18.104 --> 00:09:30.794
And I can remember, it's, it's funny you mention that about the content rec ad networks 'cause I can remember talking with someone in ad ops many years ago before it became, like, easy, like, if you're coming up short on a campaign, you can get traffic, right?

57
00:09:31.024 --> 00:09:36.314
But you had to do, you had to do all, all sorts of crazy stuff that got dumped onto ad ops, [laughs] right?

58
00:09:36.344 --> 00:09:54.584
I remember telling one story about how a sales team had double sold a campaign to this, to, to whatever vertical, and so they just sort of negative geo-targeted the client's geography, so they were seeing the ads that they were expecting to run, and then they were running a, a totally different part of the publication.

59
00:09:54.684 --> 00:10:04.244
But there's the kind of things you had to do in ad ops back then. You don't have to do that, you didn't have to do that then anymore. So with the lower traffic, you also have, you know, other lever is...

60
00:10:04.284 --> 00:10:18.404
And that's being hit, right? And I think the big shoe to drop for monetizing page views is the, the cookie deprecation, right? Explain where, well, where we are with that.

61
00:10:18.844 --> 00:10:27.764
I mean, obviously it's happening over incrementally. Google i- as this has been, like, a long striptease for Google with eliminating this in Chrome. Mm-hmm.

62
00:10:27.804 --> 00:10:37.504
But explain, like, where we are with that, and I've seen some reports of the impact it's had so far. I don't know how directionally accurate they are. Yeah.

63
00:10:37.564 --> 00:10:49.584
You made a jump though from, like, CPMs are down to it's the cookie's fault, and it's not just the cookie's fault. There are things coming for that, and there's macroeconomic factors- Okay, let's separate it out then.

64
00:10:49.614 --> 00:11:02.134
Okay. 'Cause th- that is fair there, 'cause I think the economy broadly held up in 2023, but, like- Yeah... when the economy, when people start reading about a downturn, it doesn't matter, advertising gets cut.

65
00:11:02.224 --> 00:11:11.864
I don't care how many times they run that op-ed in Ad Age. And that happened, and then on top of that, there's just... There's been a shift a- about where budgets are going.

66
00:11:11.924 --> 00:11:15.284
It is not going to content publishers as far as I can see. [laughs] Correct.

67
00:11:15.624 --> 00:11:31.784
So I think 2023, the forecast was, like, it was either '22 or '23, it was, like, minus seven or minus 11% investment in digital web, but plus 12 into search and plus something double digit into social channels.

68
00:11:32.404 --> 00:11:40.964
So the total volume of marketing as, like, a GDP spend or whatever is going up a little bit, and that's encouraging to the holding companies and to economists.

69
00:11:41.444 --> 00:11:52.344
But when you start parsing out where it goes, it's not going into the digital web. That's because walled gardens are getting better, and they're good, they are good places for marketers to spend your money.

70
00:11:52.414 --> 00:12:03.014
So more is going into Google and going into search. More is going into Meta and Instagram, and more is going into TikTok, and which are the next publisher.

71
00:12:03.064 --> 00:12:09.804
So the cookie, like, has a- Explain, explain for those who are not hip deep in, in this world why clean rooms are important.

72
00:12:09.824 --> 00:12:14.784
Well, clean rooms are really important because they are helping brands transact on a data level.

73
00:12:15.384 --> 00:12:25.604
The, the, from the retail media network side, they are becoming publishers themselves because they're representing inventory. Yeah. They're selling ads. They have a data product.

74
00:12:26.044 --> 00:12:39.136
But what happens when, you know, a grocery store retailer runs out of space on their website?Are they gonna start creating content and trying to attract more users to read recipes on their site?

75
00:12:39.536 --> 00:12:46.576
There will become a limit for retail media networks where they reach the end of their pond. So that's one area where they're gonna have a challenge.

76
00:12:46.616 --> 00:13:08.336
The second one is they're still under the radar of the big Google and the big Amazon, and once they start trying to steal the real search budgets or really impeding on their normal business, you will see those two large monopolies move in on them and take their businesses back from them, or give them little nicks in the back of the Achilles heel.

77
00:13:08.816 --> 00:13:13.776
But they're under those radars right now, and they're still in their honeymoon phase because they're only a year or two old.

78
00:13:13.816 --> 00:13:23.726
They are driving revenue, but nobody's really done the ROI math or the longevity math or done the calculations of, well, what happens when we hit our first big challenge?

79
00:13:24.276 --> 00:13:36.276
Traffic scale, advertising, people connect the dots of are we really driving results? Retail media network will-- hasn't been challenged yet, and they will be. Maybe not this year, probably next year.

80
00:13:36.676 --> 00:13:50.106
But is it safe to say that has incrementally taken some share that would normally go to content publishers? I'd say that it did. It does. You know, there's a lot of the budget that never would have gone to them.

81
00:13:50.136 --> 00:14:02.826
It might be a shifting of search and, you know, some of the performance budgets. But there's money that would, whether it's test budgets or, you know, affiliate revenue or even branding and awareness plays.

82
00:14:03.356 --> 00:14:18.036
You're giving them somewhere else to spend that comes with these really fancy and productive, in my mind, data systems that are backed by clean rooms and great technologies like LiveRamp and Habu, which they just acquired, and the DMPs that support them.

83
00:14:18.056 --> 00:14:30.516
But they're taking, they're definitely taking money from- Right... publishers. Like, they're growing. By the way, Amazon is the number one retail media network, and then it gets very small very fast. Right. Yeah.

84
00:14:30.596 --> 00:14:39.536
So what- So that leaves us with the cookie What would then the other shoe to drop, the cookie. Yes. C- I just wanted to get to the cookie, Scott. Yeah, I know, man. You know I feel about it. Everybody wants to get there.

85
00:14:39.556 --> 00:14:48.776
You gotta have little vegetables on the way, Brian. The first and the foremost of the cookie is only forty percent of the web is currently addressable when we talk about the open web.

86
00:14:49.136 --> 00:15:01.396
Between Safari browsers taking roughly forty to fifty percent share of the market, that's, that has no cookies in it. Incognito browsers can't use the data, that's restricted.

87
00:15:02.056 --> 00:15:16.056
And then you also have, um, Chrome on iOS devices. You're like, "It's Chrome, that's gotta be great." Actually, you're running Chrome in an Apple WebKit, and so you have to abide by the Apple rules.

88
00:15:16.096 --> 00:15:19.186
So Chrome on iOS doesn't have cookies either.

89
00:15:19.236 --> 00:15:32.616
And so all that's left, I've looked at a bunch of different publishers, and we run the analysis, only about forty to fifty percent, usually forty to forty-five percent of your audience is addressable on a cookie basis today, and it's been that way for a couple years.

90
00:15:33.116 --> 00:15:44.906
So- Mm-hmm... most of the damage has already been done when you talk about it from a majority perspective. We do see CPM differences of about sixty percent. So if it's a dollar on Chrome- Oof...

91
00:15:44.936 --> 00:15:55.336
it's forty cents on Safari or in incognito mode or in these browser modes because a lot of the internet's driven by retargeting and by data.

92
00:15:55.396 --> 00:16:03.536
Ps have been built to find data, to either use it in for a triangulation basis or a pure targeting basis.

93
00:16:03.916 --> 00:16:13.496
But if you don't have data, metadata about you, advertisers don't know who you are, even though you might be a very highly valuable user or in market for something.

94
00:16:13.776 --> 00:16:22.896
And for marketers who only have been targeting on cookies for the last three years, they ignored every Safari dev- or every iOS device out there.

95
00:16:22.956 --> 00:16:30.536
I've heard a marketer tell me that she didn't mind using cookies to target tech enthusiasts because her budget was scaling just fine on Chrome.

96
00:16:30.736 --> 00:16:38.636
And I was like, "Well, what about all the tech enthusiasts that own iPhones?" And then, you know, they show you the door and, 'cause they don't wanna ask their client about it, and they don't wanna push on it.

97
00:16:39.076 --> 00:16:53.196
So what's coming up is not the big part, but it's the remaining part. And we know how big the drop can be. And there's math that publishers can do, I'd love to help you out on that, of forecasting how much risk you have.

98
00:16:53.756 --> 00:17:05.426
Some publishers have more or less risk based on a variety of attributes. But when the cookie goes away, it's not going down to zero like Safari did or to sixty percent down.

99
00:17:05.876 --> 00:17:18.876
It's gonna come back in some ways because Google, Chrome specifically, actually cares about trying to make an internet that works for everybody, unlike Apple or Safari specifically. Yeah.

100
00:17:19.256 --> 00:17:29.536
So they're replacing it with privacy sand- Well, Google says Google makes a lot of money. Google, look- Yeah... their capitalism is good for all of us. I mean, Google has an interest in this. In a few ways. Yeah.

101
00:17:34.116 --> 00:17:53.816
[gentle music] But, you know, Google is facing its biggest challenge with, you know, AI search really.

102
00:17:53.836 --> 00:17:57.046
And I think nobody knows how that is gonna pan out.

103
00:17:57.056 --> 00:18:11.336
But when you add up all these things together, it's hard for me to see, like, an optimistic case, but I would love for you to make one, for the commercially viable open internet that you and I have known for a long time.

104
00:18:11.836 --> 00:18:21.496
I mean, is, is, is this just ending? Uh, I feel like we're between eras, but I, I can't say for s-sure. It sure does look that way to me. Between eras is a good way to put it.

105
00:18:21.896 --> 00:18:30.716
I don't think the internet dies, despite the, the drama that we may have. I don't think monetization dies. I am afraid that it may be drastically different.

106
00:18:31.116 --> 00:18:42.928
We're hurtling towards this, like, world of authentication right now, where a user with that data has no valueSimilar to ad block users four or five years ago.

107
00:18:42.988 --> 00:18:53.668
If an ad block user doesn't see ads, what's the value on your website? If you have a Safari user that sees very cheap CPM ads, what's the value on your website?

108
00:18:54.228 --> 00:19:08.888
You'd rather hold them hostage for an authentication, get them to sign up for a newsletter, buy something from you, rather than wanting to show them ads. So there's a world where we're gonna just turn away invaluable...

109
00:19:08.948 --> 00:19:11.167
Not invaluable, less valuable users.

110
00:19:11.748 --> 00:19:28.128
And you're already seeing that happen between SSPs and DSPs with something called traffic shaping, where the SSPs and DSPs are either limiting what they're sending or limiting what they're listening to or receiving from the, from each side.

111
00:19:28.148 --> 00:19:41.748
And that is having a major consequence on how publishers are able to monetize. And it's destroying the industry from the inside. It's a, it's an escalating world that is leading us to implosion.

112
00:19:42.108 --> 00:19:51.228
Was that the optimistic take? [chuckles] This is not an optimistic take, Scott. No, I don't know. [laughs] This sounds terrible. Okay, let's go back and be optimistic about this. Is there a but? [chuckles] Yeah.

113
00:19:51.268 --> 00:20:05.968
But ad tech survives, and martech survives. As long as- But ad tech is here to help. Thank God. [chuckles] Thank God. You know, marketers want to reach eyeballs. They want, and they want to measure it, right?

114
00:20:06.008 --> 00:20:11.648
They wanna target and measure. And if you can show the ROI, you will get budget, right?

115
00:20:11.728 --> 00:20:28.648
Tap ad, bless their hearts, a long time ago, created multi-touch attribution, or they didn't create it, but they proliferated it from saying that if somebody sees an ad on their mobile device, they'll make a purchase on their desktop because at that time, nobody was buying anything on mobile.

116
00:20:28.688 --> 00:20:40.948
And so as soon as that attribution was able to be shown that advertising on mobile was important to a desktop conversion, the budgets started flowing harder into mobile web environments.

117
00:20:41.628 --> 00:20:49.428
Eventually, the devices caught up and Shopify really pushed this forward, so now we can do transactions on mobile, so it's a little more one-to-one.

118
00:20:49.448 --> 00:20:59.708
But what we need to really think about preserving is measurement because if marketers can't measure it, they won't spend it, and cookies played a very large role- Uh... in measurement. Yeah.

119
00:20:59.728 --> 00:21:10.788
No, I think a lot of times the cookie discussion goes to targeting, but the measurement is just as important, if not more important, right? Absolutely. 'Cause what's the- Um... related- Only for D2C...

120
00:21:10.928 --> 00:21:22.118
a related issue, yeah. But a related issue is around attention metrics. I think, you know, I hear lots more about attention metrics and like this industry loves to go from, like, thing to thing.

121
00:21:22.728 --> 00:21:26.788
You know, just like a few years ago, it was all viewability, viewability, and then guess what?

122
00:21:26.808 --> 00:21:34.028
Viewability apparently beget more problems [chuckles] that so you create another solution to the problems that were created by the last solution.

123
00:21:34.048 --> 00:21:44.528
First of all, is, are attention metrics truly taking off in the marketplace? I'd say, I wouldn't say taking off, but they're definitely taking hold, and they're starting to weave their way in.

124
00:21:44.908 --> 00:21:57.348
I don't doubt that attention is correlated with performance, right? If you're not paying attention to the ad, or if it's loading out of view, or it's unviewable, it probably performs worse.

125
00:21:57.768 --> 00:22:08.888
But the devil's in the details of where do you set a threshold? How do you measure this multi-dimensional metric like attention? Viewability was kind of easy.

126
00:22:09.148 --> 00:22:15.828
Is the ad in view or out of view of a certain number of pixels for a certain amount of time? The industry developed a standard. It became standard.

127
00:22:15.908 --> 00:22:27.028
Group M had a different interpretation of the standard, which was challenging. But most everybody coalesced around standards of viewability. Will there be a standard of attention?

128
00:22:27.588 --> 00:22:34.908
Like this ad, you know, will 70% be the threshold of attention? That means the vendors have to start agreeing on methods.

129
00:22:35.208 --> 00:22:44.468
So before attention with a capital A takes off, you need to have multiple vendors providing comparable solutions.

130
00:22:44.948 --> 00:22:55.048
So I think you're gonna see this fragmentation of solutions and then a consolidation of solutions coalescing around a single standard, and then it'll look more, it'll follow the pattern of viewability.

131
00:22:55.327 --> 00:23:01.368
But attention's really an unfair metric for publishers because we don't know how it gets measured.

132
00:23:01.408 --> 00:23:09.448
There's no standard for it, and the only way I know to get an attention metric is to license it from a vendor who's selling it to you.

133
00:23:10.128 --> 00:23:20.378
So there's another, you know, 100 grand out of your pocket to license one of these platforms so that you can know how you're being evaluated and maybe optimize a campaign. That's an ad tech tax.

134
00:23:20.388 --> 00:23:32.368
Yeah, there's, I mean, like all these, yeah, all these metrics are black boxes at the end of the day. It's unclear how attention is truly measured, right?

135
00:23:32.428 --> 00:23:43.068
And look, I mean, with viewability, I think we saw how it was gamed at the end. Now it's about- Yeah... MFAs because the MFAs,

136
00:23:43.988 --> 00:23:53.088
it seems to me a lot of them sprung up because they scored really well in viewability metrics. That's for sure. That was like their leading, that was their leading punchline.

137
00:23:53.688 --> 00:24:01.328
Like high viewability, high completion rates, brand safe, all the things they wanted, but they didn't drive business outcomes.

138
00:24:01.337 --> 00:24:05.118
[chuckles] I do take challenge, and Chris Kane has heard me say this- Yeah, it's just reverse engineering...

139
00:24:05.118 --> 00:24:15.468
that like they may have driven awareness, but then if you look at their attention, it was probably low, and they probably didn't drive awareness. They only measure their transactional value of those ads in that study.

140
00:24:15.888 --> 00:24:30.008
I don't refute the study. I'm just saying they, there was a part of advertising that wasn't considered. Right. Yeah. So you mentioned the ad tech tax. In your piece, you talk about a loss of, of media productivity.

141
00:24:30.148 --> 00:24:43.892
I think that's a, a bit of a twist. I mean, there's a lot of focus goes into obviously this is a very complicated supply chain and-A lot of it is driven by data, as you said, and nothing is free in life.

142
00:24:43.952 --> 00:24:52.112
And so publishers a lot of times complain about all the intermediaries, but in some ways they're adding a lot of value.

143
00:24:52.152 --> 00:25:04.772
A lot of the intermediaries are adding a lot of the value in the transaction, and, and that's why their take is, is high. Explain the, explain the k- what goes into like a loss of media productivity.

144
00:25:04.782 --> 00:25:14.212
So the loss of media productivity, it actually came from the ANA's programmatic media supply chain transparency study, and that was a term that they- Yeah... came up with.

145
00:25:14.252 --> 00:25:25.172
But if you look at the waterfall of costs as it goes through the chain, on the beginning side, there's the transactional cost of SSPs and DSP fees and data targeting fees.

146
00:25:25.192 --> 00:25:35.892
But then once a, an impression gets delivered onto a publisher page, there's a bunch of other things that would devalue that impression for a marketer.

147
00:25:36.362 --> 00:25:47.532
And so specifically from the ANA study, first there's viewability, and then there's IVT, there's non-measurable impressions, there's impressions that are just purely wasted on MFA sites.

148
00:25:48.192 --> 00:25:58.162
And then you get to the final number, which is the true ad spend. By the way, IVT is, is, is invalid traffic, right? Invalid traffic. It could be sophisticated invalid traffic or- Yeah...

149
00:25:58.162 --> 00:26:09.112
general invalid traffic, which is- I'm not sure why V is in that acronym, but- You're correct, because IT would've gotten too much. I use TRV. It's okay. I'm not... Very good.

150
00:26:09.672 --> 00:26:22.992
So from a, from, for those listening at home and not looking at the chart, when you go from $1,000 of total ad spend in the ANA report, sellers get $710. So that's 29% makes it through to the publisher.

151
00:26:23.492 --> 00:26:38.712
But then you go from $710 down to $360 of true ad spend, as they call it. So really they're, the ANA is saying that only 36% of your ads are effective or like, or no, that's not it.

152
00:26:39.072 --> 00:26:45.432
Like you spent, you should have only spent 360 bucks. Like that's all you really bought of working media. Yeah.

153
00:26:45.632 --> 00:26:57.422
So I think you're gonna see DSPs and mid- middlemen and ad tech vendors like really approach this part of the conversation of like, "We're here to solve IVT," or, "We're here to solve MFA. We're going to do avoidance."

154
00:26:58.112 --> 00:27:06.032
Because the, the left side of the graph where it's transactional costs are kind of fixed. DSPs take their fixed percentage, SSPs take that.

155
00:27:06.112 --> 00:27:19.692
So when you're talking about helping marketers, you're solving for the, the soft side of the equation and recovering- Yeah... you're trying to reduce the, the lump, if you will, loss of media productivity. Yeah.

156
00:27:20.272 --> 00:27:24.072
So there's obviously a ton of different types of publishers, right?

157
00:27:24.292 --> 00:27:35.272
But if you, if you could avoid it, like wha-- I, I often hear from publishers, obviously most publishers claim they're premium, but like premium publishers, they're running away from this.

158
00:27:35.352 --> 00:27:45.352
Like if they could be less exposed to this part of the media world, they would be. Like when I go around the table, I'm like, "What are you optimistic about?" More publishers are saying events.

159
00:27:45.672 --> 00:27:54.372
Like nobody is saying programmatic advertising. Like I, I just don't hear it. Maybe I'm talking, I'm not talking to arbit publishers and stuff like this. I just don't hear it.

160
00:27:54.412 --> 00:28:09.452
And this is a business that, look, a lot of the industry was built around digital advertising. I just don't see that as the foundation going forward for a lot of publishers that are starting out now. I just...

161
00:28:09.492 --> 00:28:20.112
Yeah, you look at what's go- if you compare, so let's say the messenger to more successful launches that took place, whether it was a Punchbowl or even a Puck and Semaphore, totally different models.

162
00:28:20.152 --> 00:28:26.852
And like, to me, the Semaphore model was very much a tra- or Semaphore, the, the messenger model, very much a traffic model.

163
00:28:27.472 --> 00:28:36.282
Very much about like h- we're gonna use a lot of different ways to get people to a website, and then we're gonna be really good at monetizing through programmatic and stuff like this.

164
00:28:36.752 --> 00:28:49.352
The results speak for themselves, at least from what I've read. Won't most publishers just try to get a- as far away from this as possible? Yes, but not everyone is able to. If they can. Yeah. Not everyone's able to.

165
00:28:49.492 --> 00:29:00.332
Okay. The, the niche verticals and like the niche areas of the internet will do better because you have passionate fans, and you need a smaller base to sustain yourself.

166
00:29:00.732 --> 00:29:08.552
The, the general interest areas are already commoditized. They have a number of competitors.

167
00:29:08.592 --> 00:29:17.452
It's really hard for them to differentiate themselves, not only from brands who are buying them, but also from consumers who are reading them.

168
00:29:17.852 --> 00:29:24.472
And you don't have social traffic anymore to help you out, so that's like one less performance area for you.

169
00:29:24.492 --> 00:29:34.262
And if you wanna do it in the new social platforms, TikTok, Snap, Instagram, you can't drive traffic from them, and they don't really have like ad networks where you can monetize from them.

170
00:29:34.852 --> 00:29:48.292
So you're doing it for, at worst, posterity stake, and at best, for building credibility or brand building and trying to get that audience, and then you have to pull the cash lever on them somehow.

171
00:29:48.672 --> 00:30:00.552
It's exit through the gift shop, and you're just building paths to the gift shop, and if somebody leaves your site or your universe without going through the gift shop, it's not worth keeping that path open.

172
00:30:00.632 --> 00:30:10.932
So yeah, do events, do newsletters, do anything you can to be less reliant on traffic because it's super unstable right now.

173
00:30:11.412 --> 00:30:25.312
Be less reliant on the open auction because it's super uncertain which way that crumbles, and do more direct sales and take control of your future. But those are like the hardest things to do. Mm. Yeah.

174
00:30:25.352 --> 00:30:33.132
It's not an easy business. So if you add it all up, it just seems like there's gonna be fewer publishers that are able to...

175
00:30:33.352 --> 00:30:45.104
I don't mean to be like a bummer, but I just don't understand how when you add all of these factors up, how you don't come to the conclusion that there are going to be fewer publishers that are commercially viable.Out there.

176
00:30:45.184 --> 00:30:55.664
I don't know if it's a good or a bad thing, but I just don't th- I don't see- It's a bad thing... I don't see necessarily how all of this adds up to, to, to a path for growth. It's a bad thing. Right.

177
00:30:55.904 --> 00:31:06.204
The loss of publishers, 'cause it's not just gonna be- It is a bad thing... it's not just gonna be entertainment celebrity sites that die, it's gonna be journalism as well. It's gonna be your favorite sports site.

178
00:31:06.264 --> 00:31:13.224
Yeah, well we're already seeing this. All the things that we... We're, we're seeing this, uh, there's, I don't know, there was 20,000 some, like, layoffs last year.

179
00:31:13.364 --> 00:31:22.804
I mean, the layoffs keep coming, and unfortunately I think they're gonna continue. And- Be optimistic about it, Brian... I don't know, it's gonna continue to be a hard business. We need to find solutions. Yes.

180
00:31:22.844 --> 00:31:27.824
And we need to find ways to help things, which is why having these dialogues are really important.

181
00:31:27.864 --> 00:31:40.264
There was a great article yesterday from Sara Slus in AdExchanger about traffic shaping and the impacts of it from the buy side, the sell side, from minority-owned publishers and how that's impacting the industry.

182
00:31:40.284 --> 00:31:50.904
And everyone needs to take a good, long, hard look at that- Mm-hmm... and see what this arms race is doing to us. There's other areas like curation that's gonna become really important.

183
00:31:51.484 --> 00:31:56.064
Contextual is a really big battleground for how publishers can survive.

184
00:31:56.124 --> 00:32:08.644
There's, there's still, like, many wars and battles to be fought where publishers can reclaim some power in the, the struggle or the balance of leverage with the buy side.

185
00:32:08.824 --> 00:32:33.784
[gentle music] But I mean, like you said, like, you know, data is in the driver's seat, right?

186
00:32:33.864 --> 00:32:51.724
And to me, there are a few candidates for the original sin of digital publishing, but my top one is the separation of the audience data from the media impression because it automatic- that commoditized context at the end of the day.

187
00:32:51.824 --> 00:33:00.744
And I remember it was put to me, it was that the industry was basically, you know, find a cookie, hit a cookie at the end, and that's what it was.

188
00:33:00.884 --> 00:33:13.944
Now, I think there's hope for a return to context as a valuable signal, but [laughs] I don't know if it's a v- I don't know if in many cases it is the most valuable signal.

189
00:33:13.984 --> 00:33:25.244
To me it's all of these things happening and I'm like, I don't see how content publishers can possibly compete with platforms as far as finding a specific person for a specific product.

190
00:33:25.344 --> 00:33:38.004
I just don't see that as a good place to be competitive. I think they can be competitive, but buy side has a lot of these tools already. They don't necessarily need or want the publisher's version of it.

191
00:33:38.344 --> 00:33:50.424
Context is not doomed. I was just speaking with my friend Lila about that today. It's not doomed. Context is very important, and the right ad in the right place is extremely important.

192
00:33:50.804 --> 00:34:05.604
Context runs through all of these platforms, and it goes through services like IAS, DoubleVerify, GrapeShot, GumGum, Iris.tv. They're all providing taxonomies to buyers. They have scraped the web.

193
00:34:05.664 --> 00:34:18.044
They have taken this information from publishers. It's legal, I get it, but they have ingested this information, and context runs through everything. It is one of the biggest triangulation points of success for a DSP.

194
00:34:18.544 --> 00:34:31.504
So when we do talk about the cookie going away, it's not everything. Context will still be there, but publishers make a lot less money from context when they don't get to use it in their side of the equation.

195
00:34:31.884 --> 00:34:41.124
It just gets tacked on to, like, a, you know, a SaaS fee inside a DSP, or maybe it's included for free. But when a publisher can do it- Mm...

196
00:34:41.144 --> 00:34:47.364
they can drive private marketplace deals, they can do better insights, they can run better direct deals, they can improve performance.

197
00:34:47.604 --> 00:35:00.733
So some of it's on the publisher for not doing a good enough job organizing their own contextual data, and some of it is on publishers as a whole for not creating a better standard system that buyers can buy into wholesale.

198
00:35:01.173 --> 00:35:10.804
And then some of it's on buyers for not really caring where or how they got their context, as long as they could just push a button and it worked. Yeah.

199
00:35:10.904 --> 00:35:25.424
I kinda think there's not enough focus placed on the buy side to some degree about a lot of the issues that we're talking about, it seems like the only, the big, the constituency that can truly drive change is the buy side at the end of the day.

200
00:35:25.734 --> 00:35:26.184
Absolutely.

201
00:35:26.224 --> 00:35:40.524
You know, and publishers can complain constantly at a digiday town hall or whatever, but at the end of the day, it's the buy side, and the brands are being inducted into the halls of fame, and they're not really addressing these issues.

202
00:35:40.604 --> 00:35:50.104
Yeah. Like, the problem, like, header bidding was created to thwart the buy side of Google, and what they were doing, you know, preventing us from other demand.

203
00:35:50.144 --> 00:36:05.144
And then bid duplication was created because of the downward pressure on pricing from DSPs, and their inability to be able to differentiate from two bids, and that started pushing prices up.

204
00:36:05.204 --> 00:36:13.944
A publisher's job is to drive yield and get more value for their, a single impression, and we have very few tools to do it.

205
00:36:13.984 --> 00:36:23.154
We're like little tiny villagers with sticks and stones, and the DSPs are like the T-1000 from Terminator 2. They introduce, like- Yeah...

206
00:36:23.244 --> 00:36:40.584
dynamic, dynamic flooring strategies which we're, which means that you're sending out different floor prices for the same impression to different SSPs based on algorithmic modeling on where the algorithm thinks you can extract more value from a DSP.

207
00:36:40.724 --> 00:36:47.984
Because DSPs and their bid shading algorithms are looking at where you can get the cheapest impressionAnd then buying that.

208
00:36:48.064 --> 00:36:57.364
So when dynamic bid sheet, when dynamic flooring was introduced, that's when The Trade Desk came out and said, "We're gonna start passing bid data below the floors."

209
00:36:57.844 --> 00:37:01.964
And they-- That was like their shot back across the bow, and you don't hear about it as much.

210
00:37:02.024 --> 00:37:15.924
But as long as buyers keep pushing down on pricing and keep slimming down what they're buying, publishers will continue to find ways to sell them stuff. And now bid enrichment is really, like, the front of that.

211
00:37:15.934 --> 00:37:25.164
It's a new acronym for you every month. Wait, explain bid enrichment [chuckles]. I don't even know bid enrichment. Yeah. Bid enrichment- I could let it go, but I was like, eh, I got a thought. Oh no, this is a great one.

212
00:37:25.204 --> 00:37:34.914
So bid enrichment is something that's been happening for a long time, and usually it was at the SSP level or potentially the DSP level where before you passed- I can tell by the term, let me just make a prediction. Yeah.

213
00:37:34.914 --> 00:37:36.844
Because the term is very, like, nice.

214
00:37:36.904 --> 00:37:46.844
It's like, oh, this sounds great, it's like it's definitely gotta be something kinda sketchy because it's [chuckles] like, it sounds like the kind of, you know, sort of nice wrapper someone would put on something that's kinda sketchy.

215
00:37:46.864 --> 00:37:54.534
But go on. I could be wrong. Yeah. Well, there are also words called like bid washing and bid laundering that are kinda, like, gonna come up, so- That sounds- You're, you're heading in the right direction...

216
00:37:54.534 --> 00:38:05.844
that probably sounds the worst [chuckles]. Bid enrichment is not bad in itself, and it's something that happens in a lot of places for a lot of good reasons, and a lot of expected and anticipated reasons.

217
00:38:05.944 --> 00:38:17.144
In, in, in short, it's anything where an intermediary or the publisher themselves are adding information to the bid request to help the buyer or to make it more attractive.

218
00:38:17.564 --> 00:38:27.524
In this particular instance, bid enrichment is moving from the SSPs or from the intermediary vendors down to the publishers, where it's happening on the publishers' pages.

219
00:38:27.624 --> 00:38:43.004
And before the, the ads are called, there's a probabilistic match that is made to the user that calls to the identity graph in the cloud and brings back all known identities related to that user.

220
00:38:43.064 --> 00:38:51.104
Those could be deterministic IDs that are in the cloud and in the graph, but it was a probabilistic match that brought them down to the publisher page.

221
00:38:51.574 --> 00:38:57.864
And then those IDs make their way into the bid request, and there's some more technical parts we can get into, but I'll save you that for now.

222
00:38:58.344 --> 00:39:10.924
And they get into the bid request, and they go to the buyers, and the buyers don't know if it was something that came through a probabilistic match or if it was a naturally browser-derived ID, which is what they were expecting.

223
00:39:11.274 --> 00:39:18.283
And nobody's really said whether it's good or bad for campaigns, but it's not what they're expecting, and that's a challenge.

224
00:39:18.824 --> 00:39:31.064
So as this enrichment is happening, buyers are seeing, or sellers are seeing real lift in Safari auctions because buyers are now all of a sudden getting IDs where they didn't.

225
00:39:31.664 --> 00:39:37.884
And they're like, "We like users with IDs," and they start buying them. So you see a lift, and that's good stuff. Yeah. Okay. Yeah.

226
00:39:38.404 --> 00:39:46.164
So final thing is I wanna learn more about, like, you talk about curation houses, like, having a moment. Explain what you mean by them. Sure.

227
00:39:46.224 --> 00:39:58.404
So in my mind, a curation house is a service that is representing publisher inventory and adding value to it, and helping package and sell that inventory to the buy side.

228
00:39:58.844 --> 00:40:09.304
Some of the natural characteristics are having a data offering, a rich media offering, and having a sales team in market to get insertion orders.

229
00:40:09.744 --> 00:40:24.464
So it's not just programmatic reselling of ads, and it's not just vanilla flavor bid enrichment, but somebody has, like, a really nice way of packaging and presenting this inventory to the buyers, and they're in market marketing that offering.

230
00:40:25.024 --> 00:40:35.253
So my favorite example, and I think the most classic, is Cargo. They have been great for publishers. Mm-hmm. They built a, a wonderful ad network of publishers. They have creative rich media services.

231
00:40:35.324 --> 00:40:45.224
They've built incredible data offerings, and they have a tremendous sales team in market winning premium priced insertion orders that they then run on their publisher sites.

232
00:40:45.674 --> 00:40:49.664
And that s- company has grown to do this in many other areas.

233
00:40:50.054 --> 00:41:00.484
And now we're seeing more companies look like curation, where you're helping publishers sell, and you're not devaluing their inventory, you're not doing it on a tonnage basis.

234
00:41:01.024 --> 00:41:08.104
It's fairly transparent, and that's what publishers need because individual scale doesn't get it done anymore. Yeah.

235
00:41:08.604 --> 00:41:18.604
I'm, like, laughing 'cause we came, like, all this way, and, like, in your last answer you were talking about, so I'm like, all this stuff just to put a display ad, like, on, like, The Washington Post, like, website?

236
00:41:18.844 --> 00:41:29.524
Amazing. And then you sort of come back to it, it's like, well, the future is like, looks like the ad networks of the past a little bit [chuckles]. They got better tools but, like, you know.

237
00:41:30.284 --> 00:41:33.954
I think a lot of times we sort of get ahead of ourselves with like, "Oh, this is like the NASDAQ.

238
00:41:33.954 --> 00:41:43.344
It's super sophisticated" and stuff, but when you come down to a publisher, it's like, you know, it's kind of basic black, blocking and tackling can get a long w- can go a long way.

239
00:41:43.644 --> 00:41:49.934
And I think those businesses probably a bunch of years ago people would say, "Oh, that's kind of the past," but not really. Yeah.

240
00:41:49.934 --> 00:42:00.804
So ask your-- Like, when I ask, like, "Is that a ad tech tax or is it a cost of doing business?" Most people will say it's a cost of doing business because you don't have to do it or you can get rid of it.

241
00:42:01.384 --> 00:42:11.344
But the truth is, we've engineered ourself into this complicated world where you have to do it. You have to pay for fraud verification. You have to pay for targeting. You have to do all these things.

242
00:42:11.364 --> 00:42:25.204
But it's because we've complicated everything, and you nece- don't necessarily know how it's helping your campaign or exactly how this extra money you're spending and the loss of media productivity on the transaction side is improving your campaign, and that's a hallmark of a tax.

243
00:42:25.684 --> 00:42:33.354
If you don't know, if you're paying it and you don't know why, and you don't know what it's going to, that's a tax. And so- Yeah... we need to get rid of everything. Okay.

244
00:42:33.384 --> 00:42:42.954
So, so yeah, so final thing is, like, if we're doing this at the end of the year, what is something that we will be looking back and say, "Hey, we should've been optimistic about this"?

245
00:42:43.164 --> 00:42:55.840
Like, what is gonna be something that from a publisher point of view is going to happen this year that we can be optimistic about?The things we're optimistic about is that publishers are gonna get lucky.

246
00:42:56.340 --> 00:43:06.140
We're gonna get lucky in some ways, and so there's gonna be some breakthroughs, and that hopefully there's dialogue into the marketplace. I think publishers need Privacy Sandbox to be successful.

247
00:43:06.560 --> 00:43:21.500
Now, that comes with a large bag of conundrums that I don't know if those are good or bad or not, but we need more help solving things. You know, there's a, a new study out about the rebranding of news.

248
00:43:21.560 --> 00:43:31.850
Like, we gotta stop hurting ourselves. I forget what the stat is, but it's like seventy percent of buyers say that they support journalism, and they want it to live. Yeah.

249
00:43:31.850 --> 00:43:42.740
But also seventy percent of buyers have, like, over-restrictive keyword blocking. So what they're saying and what they're doing don't match yet. I'm optimistic that we can work through a bunch of that.

250
00:43:42.800 --> 00:43:54.060
Like, that's our eternal optimism, is that maybe somebody will listen to us, and we can all put down- Yeah... our guns and n- our programmatic guns and knives and get to working together. Yeah.

251
00:43:54.160 --> 00:44:09.600
I mean, the block lists are insane. Like, I mean, just the fact that when we're talking about the sophistication of a lot [chuckles] of this ecosystem, when buyers are just saying, "Keep me away from this word."

252
00:44:10.240 --> 00:44:13.260
Like, it's just, it's insane that's happening in twenty twenty-four.

253
00:44:13.740 --> 00:44:26.640
It just-- I mean, it's not like i- if they were doing something a little bit more sophisticated to block me, I would sort of be a little bit more understanding of it, but this just seems like lazy CYA behavior to me.

254
00:44:26.670 --> 00:44:40.180
It doesn't even seem effective. Big CYA behavior. All right. All right, so some things won't change in twenty twenty-four, one of which is big CYA energy is gonna still be [chuckles] alive and well. I like that one.

255
00:44:40.920 --> 00:44:50.680
Yeah, you can bet on that. All right. Okay, Scott, this was fun. Appreciate you doing this. It was a great conversation. Thank you for having me here, Brian. Always wanted to show up. Always love working with you.

256
00:44:50.870 --> 00:45:03.240
Appreciate it. Thank you. Thanks for listening. Thank you to Jay Sparks for producing the Rebooting show. If you have a podcast that you're considering making, you should check out Podhelp us and what Jay can do for you.

257
00:45:03.680 --> 00:45:05.260
Go to podhelp.us.

258
00:45:06.540 --> 00:45:25.549
[outro jingle]