Public Workshop on Competition in Television and Digital Advertising (Panel 2)

Public Workshop on Competition in Television and Digital Advertising (Panel 2)


MS. LUBELL: So, welcome to the
second panel of the afternoon, “Online Advertising – The Nuts
and Bolts of Digital, Mobile and Online Television.” We
have a lot to cover during this panel. And so, we’re going to try
to end, let’s see, probably shortly after 5:00. So I’m going to only briefly
introduce the panelists. You should all have their full
bios in a printed — with the agenda. So I’ll go through it
alphabetically, starting to my far right. Kevin Arrix is senior vice
president of DISH Media sales, where he’s responsible for
DISH TV’s and Sling TV’s advertising sales,
analytics and operations. He leads the team in charge
of the company’s advanced advertising initiatives,
which includes cross-platform addressable, programmatic
sales and dynamic ad insertion. And hopefully he’ll tell us
a little bit more about the meaning of those terms. Next to him, we have Joshua
Lowcock, who is the executive vice president and chief legal
and brand safety officer at UM, where he leads digital
strategy and innovation across all U.S. accounts, including Coca-Cola,
BMW, CVS Health, Sony and USPS. He’s also the architect of
the 4As advertiser protection bureau. Immediately to my right is
Preston McAfee, who is an economist who’s worked
extensively in pricing, auctions, antitrust, business
strategy, market design, computational advertising and
machine learning applied to exchanges. After 28 years as a university
professor, Preston joined Yahoo in 2007 as chief
economist, moved to Google in 2012 and joined Microsoft as
chief economist in 2014, from which he retired last year. Then to my left, Wendy Moe is
the associate dean of master’s programs, dean’s professor of
marketing and co-director of the Smith Analytics Consortium
at the University of Maryland’s Robert H. Smith School of Business. She’s an expert in online and
social media marketing, with a focus on analytics and has
consulted on these topics for fortune 500 corporations
and government agencies. Next to Wendy is Larry Solov,
who is president, CEO and general counsel of Breitbart
News Network, a news media company he cofounded with the
late Andrew Breitbart in 2007. The company has offices in Los
Angeles and London and bureaus in Washington,
Jerusalem and Rome. And finally, at my far left is
Greg Stuart, who is the CEO of the Mobile Marketing
Association, a global media trade group focused
on mobile marketing. MMA has more than 800 member
companies globally and employees in 15 countries. And the board includes chief
marketing officers from Marriott, Uber, Walmart and
Samsung, among others, plus senior executives from
Facebook, Google, Twitter and Snap. So now, I’m going to give
each of the panelists an opportunity to tell us a
little bit more about their views, their work and the
role of their organizations in digital advertising. So let’s go down the table,
and we’ll start with you, Kevin. MR. ARRIX: Thank you. So DISH Media, I think the
important things are that the assets that we have, which is
dish satellite television and Sling TV. When you look at the topic
of digital, Sling TV is right there. It is live television
served over the top. And it is digital because it
is IP-delivered and it gives you all the functionality
that you get from digital advertising and digital media. MR. LOWCOCK: So I’m
from Universal McCann. We’re a media agency. What that means is we plan,
strategize and define where media’s going to be bought
on behalf of our clients. The key point there is we’re
an agent for our clients. So we influence decisions. Our clients make the ultimate
decision where advertising is going to be placed. I also have the role
of brand safety. So I’m responsible for
ensuring our clients’ ads run in brand-safe environments,
but also helping keep the industry accountable and
keeping the platforms accountable for getting rid of
inappropriate content that’s online. MS. LUBELL: So are the — and
we should have some slides available for Preston. DR. MCAFEE: So I was
responsible — involved in digital advertising at Yahoo
and Google, also at Microsoft. And I wanted to start out by
saying why judging advertising effectiveness is hard. You heard in the previous
panel that it’s hard. You didn’t hear
very much about why. One thing that’s hard to
do with advertising is the counterfactual. So that is to say when —
actually, let me just give an example that was in the
Harvard Business Review
. People who saw brokerage ads
were 1,600 percent more likely to open a brokerage account
than people who didn’t see brokerage ads. And the article didn’t notice
that, huh, people who saw brokerage ads searched
for the word brokerage. They are different people than
the people who didn’t see the brokerage ads, a
different kind of people. Let me give an
example of that. Can we go to the
next slide, please? So this is real data from
a — oh, I have control. This is real data. A company — a Fortune 500
company tells me we’ve managed our advertising effectiveness. On the horizontal
axis is our ad spend. On the vertical axis is
our revenue, our in-store, same-store revenue. And the line through
them tells us our — the effectiveness of our
ad, our ad spend. And then, when asked why they
sometimes spend more than others, oh, we spend
more at Christmas. So the theory under which
that line measures the effectiveness of ads is the
theory which is if you don’t advertise, Christmas
doesn’t happen, right? That’s the theory. Okay, and so besides the
counterfactuals, there’s been a history of poor measurement. You hear a lot about TV. But what you don’t hear is
has anyone ever measured the effectiveness of TV. And the answer to that is,
approximately speaking, no. And so, it’s just
been believed. Why? Because our grandfathers
believed it, not because it’s been effectively measured. And then, the final thing I’ll
say is that people who sell ads are marketing
professionals. They are the best in the world
at convincing people to buy stuff. So it’s not surprising that
they convince people to buy advertising. And with that,
I’ll turn it over. DR. MOE: Thank you. That sets me up very well for
the marketer’s perspective. So I am a professor
of marketing. And the best way to give
you a sense as to what my perspectives are is to
give you — so introduce my history. So I started in the field
about 20 years ago in terms of research. And at that point, Internet
was just getting started. And prior to that, I had
worked at AC Nielsen, basically measuring
advertising effectiveness on sales. And over my career, as the
data changed, my research focus changed, moving from
understanding online consumer behavior to measuring the
effectiveness of online advertising in terms of last
click and attribution models and then moving on to
more recently social media analytics and how that
can help measure brand. One of the things that I loved
hearing about in the earlier talk was the issue of the
marketing or purchasing funnel. And the way I think about it
is, you know, in each stage of my career, I’ve focused
on a different channel of advertising. And when the Internet first
started, advertising on the Internet was just like any
other form of advertising, just in a different channel. It was just like TV broadcast. It was just banner
ads on websites. Over time, as the data got
caught up and the tracking caught up and the analytics
caught up with the ability to individually track, monitor
and measure, that’s when the Internet became this different
vehicle and we started looking at how do we reach — how do
we measure individuals, how do we predict individuals, how
do we figure out who their friends are so that we can
reach their friends, et cetera. And so, over time, the
landscape has evolved. And from my perspective now,
we still have these different channels and different types
of ads hitting different parts of the marketing funnel. So you have some of the larger
banner ads and the direct ads and the television ads going
after top of funnel consumers and the targeted Internet
direct ads — sorry, the targeted ads on Facebook,
throughout the Internet, those more capture behavior at
the bottom of the funnel. And so, there is an argument
to be made that the two are complementary in some way. MS. LUBELL: Thank you. Larry? MR. SOLOV: Sure. Sure. So I’m Larry Solov. Again, I run Breitbart News. Andrew Breitbart and I were
best friends and next door neighbors growing up. And since 2012, when Andrew
unfortunately passed too early, I have dedicated my
life to promoting his mission and legacy, which is, you
know, among other things, to democratize the news coverage
and to allow more voices, not fewer at the table. It is only by the free flow
of information and ideas and a robust discussion and
iteration on those ideas that our democracy, you know, is
preserved and in fact grows and thrives. So as a publisher with that
point of view, I think please understand that that’s my
approach to these issues when I speak about them. MS. LUBELL: Okay, and Greg? MR. STUART: Hi. My name is Greg Stuart. I’m the global CMO/COO of the
Mobile Marketing Association. I am a nonprofit
trade association. Unlike the previous panel,
I’m not here with any agenda. I asked my attorneys if
DOJ calls, do I go down. They said, yes, you
should go there. So that’s the only reason I’m
here, because they advised it. Listen, fundamentally you can
understand a trade group by who its board is. And as mentioned earlier, my
board is controlled and run, the majority, by CMOs, major
CMOs of large corporations. So that’s where we sort of
fixate, whatever we can do to help them architect the
future of marketing. That’s fundamentally
what the group does. And we do that in initiatives
in measurement, organizational design, brand safety, mobile
fraud, all the issues that sort of can either support or
plague an industry, we try to address collectively. I also do have Google,
Facebook, Snap and a number of the other digital
media companies. So we’re a holistic group
trying to solve these problems collectively is really what
we’re about as a group. DISCUSSION/QUESTION-AND-ANSWER
MS. LUBELL: Thank you, Greg. So before we get into some of
the complexities of digital advertising, I just want us
to start with the basics. And Wendy, you touched on
this in some of your opening remarks. But I was hoping
you could elaborate. You know, what are the
different forms of digital advertising and what can you
tell us about their respective purposes? DR. MOE: So the way I think
about it is we have the more traditional banner ads that
have been around for a very long time who have — which
probably mirrors the offline advertising on TV and
print media most similarly. And those are targeted in
much the same way that print advertising, magazine ads and
TV ads are targeted in that you look for the venue that
most — with the audience that most matches the audience
you’re trying to reach. And so, it’s a much
broader audience. You know, earlier, one of the
panelists referred to it as waste. I tend to refer to
it as opportunities. And I’ll elaborate more on
that as I start going — after I go through the other types
of ads, the way I see them. The other type are the more
targeted, whether it’s search, through an ad network or
through your social media feed. Those targeted ads
basically are micro level. They’re looking at
individuals, tracing their behavior across various parts
of the Internet to try and assess what it is that those
individuals are most likely to buy. And that relates to what
Preston just earlier — just commented on that those
targeting algorithms aren’t necessarily looking for
opportunities or new customer opportunities. But they’re looking to predict
who’s likely going to buy. So personally I think that’s
where the waste is, that you’re offering ads and
charging brands for ads to reach customers who had a high
probability of buying your product anyway without the ad,
versus if you were to go to purchase a banner ad
or a television ad. You reach people who maybe
you wouldn’t have considered before, but may still be
interested in your product because they’re similar in
some way to other people who are interested. And so, from that perspective,
individual brands, by over-targeting, are missing
opportunities to reach new customer groups. And from a consumer’s
perspective, when there’s too much heavy targeting, they
don’t get to see the other options out there that may
be a good fit for their own tastes and preferences. They don’t see new ideas. MS. LUBELL: Thank you. No, that’s really helpful. So as you all know, one of the
core questions that we hope to answer, or at least ask during
this workshop is the extent to which different advertising
markets are substitutable. So now that Wendy has told us
a little bit about some of the different forms of digital
advertising and their purposes, Preston, I was
hoping that maybe you could talk a little bit
about substitutability. DR. MCAFEE: Sure. Let me follow up on what Wendy
said as a preface to answering that, which is that
historically there was this quite sharp division between
search and display, where search ads were text ads. They were targeted almost
primarily on keywords, whereas display ads would be richer
media, graphic ads or even video and would be on webpages
other than search pages. And that’s been blurred a lot. Shopping ads have blurred
the distinction on search ads because now, on search pages,
you see graphic ads that are known as shopping ads. They also do a fair
bit of targeting. Besides geo, you can target
personal characteristics on search pages, time of day as
well as customer age and so on. And I actually like to
describe advertisers on a continuum. And I’m fortunate from the
previous panel because on one end of the continuum
is Procter & Gamble. And they are almost entirely
about the brand; that is, very few of us click on an
ad and buy shampoo. A rare event. So that’s not their market. Their market is getting me
— putting it in my head that when I’m in the store, I
should buy their shampoo and not some other shampoo; that
is, at a subsequent time. So they’re all about the
brand and not about the click. And then meanwhile, you
have on the far end of the spectrum, the other end
of the spectrum, malware distributors. Malware distributors get
nothing unless you click on their ad, in which case they
install a rootkit in your computer. And so, and in between, you
have performance ads and other kinds of ads where they value
a click, but they also value their brand and they’re trying
to do a different message. So when you think about
substitutability, you really want to cross that with what
kind of advertiser are we talking about. And things like brand safety
matter tremendously at the P&G side, not so much if
you’re a Russian troll. And so, and so, I don’t tend
to think of there being an answer that these two
markets are substitutable. It’s rather that these markets
are substitutable for this kind of — for this
kind of company. And, you know, one of the
reasons — or let me add one more thing and then I’ll
try to actually stop. The other thing that’s worth
knowing is there’s been this question about why did — you
know, who is it that watches five hours of TV. Is there anybody in the room
that watches five hours of TV every day? This is like the
national media, right? That’s declined somewhat. MR. ARRIX: They don’t
want to raise their hand. DR. MCAFEE: Yeah. So, and who is it that
didn’t adopt computers? About a third of America
never adopted broadband. Who are those people? The answer is the desktop
computer world is very unfriendly if you’re
not a good speller. It just is. So there’s a sociologist who
said you can account for most of the people who didn’t adopt
broadband are the same people who didn’t read a novel
in the last 12 months. This is — the desktop world
was unfriendly to people who are marginal spellers. Let me put it that way. And the thing that’s really
interesting is that the mobile world is not. The mobile world is
friendly to all of us. And as a consequence, it’s got
way — you know, one of the highest adoption rates of
any device in America or any product in America. And so, when I think about
digital, I actually split the mobile world, which is a
really good world except for the 1.7 second attention span. It’s a really good world for
reach, better than TV, but not a — but not so good at depth. And so, that’s really the
thrust of my answer, which is they’re substitutes,
but they’re not perfect substitutes. MS. LUBELL: Thank you. That’s helpful. And I think we’re going to
— we’ll go back to mobile. We’ll talk to Greg about that. But first, maybe if Joshua —
you know, it’d be helpful to hear about some of these
issues from an ad agency perspective. You know, how to advertisers
or how do you advise advertiser to allocate ad
spend across first digital and non-digital media and then
what about within digital. For example, how do you decide
whether to allocate ad dollars to search versus display
versus video advertising? And then also if you could
just talk to us about who makes the decisions
where to advertise. Is it the advertisers,
the ad agency, some other intermediary? MR. LOWCOCK: So it’s a lot
of things there to unbundle. I’ll start by we make
recommendations on how to allocate sort of
independent of the channel. It all starts with the
campaign objective. So we service our clients. Our clients give us a brief. That brief has objectives
for a campaign. And that could be a
brand-building campaign or a short-term
performance campaign. I think one of the things that
hasn’t come up today which is important to factor in is time
and the length of time that we have to operate
that campaign in. And to Preston’s slide about
Christmas, like if it’s holiday or Black Friday,
there’s a lot more pressure on shifting spend to traditional
media because you need to hit that reach of an audience
quickly, whereas if it’s outside of those campaign
constraints, you’ve got more independence. We plan media against data. So where it all starts
with a customer. Like who the customer the
client’s trying to reach. That customer at the end
is going to be a consumer. So we want to measure
effectiveness against that. And then, we want to be able
to find them and identify them on channels. We don’t plan television,
radio, print, out-of-home in siloes. We plan them holistically. So there’s not a TV planner. There’s an integrated
video planner. There’s not a radio planner. There’s an integrated
audio planner. From the — like from the
sell side, it might seem independent of that because if
you’re a seller of media, you might actually only
negotiate with one person. But from an agency strategy
and planning perspective and from the data we use, we look
at consumers holistically because consumers
are complex animals. If I asked anyone to raise
their hand, nobody would raise their hand and go I only watch
TV, I only watch radio, I only look at out-of-home. Like you can shoot
multiple channels. So what we look at is the
consumer behavior across multiple channels. You heard earlier from Susan
and others about multi-touch attribution. So we’re identifying the
audience, the consumer. We’re enriching as much data
as we can all the way down that path. And then, we’re looking at
the impact and constraints of media across all
of those channels. So TV drives search. Search can drive TV behavior. It can drive
consumption of content. Everything interrelates. It’s not straightforward. On the search side of things,
it’s getting increasingly more complex, because I’m trying to
remember everything you asked. On the search side, it gets
increasingly more complex because search can serve
multiple — you can have deliberate — and I used to
work on the P&G business in China of all places. People can search for
individual products. So they can search for Tide. So you might buy
the keyword on Tide. Or they might search for
something like stain removal. So then you look back
at campaign objectives. Is it higher funnel, like I’m
trying to understand how to remove stains. I’m trying to get my brand
out there, so using search for branding. Or it can be very product,
purchase-focused, trying to get people to your
individual site. Where search gets really
interesting today is it’s very easy from a market definition
perspective to talk about Google and Bing
as search engines. But Amazon is increasingly
considered a search destination. Pinterest is increasingly
considered a search destination. Facebook is a
search destination. And all of those environments
are where you’re buying search keywords. So, you know, to sort of
challenge Preston’s point of no one clicks in shampoo and
clicks on an ad and then buys shampoo, I am buying search
ads on YouTube that might result in people then going
to Walmart and then buying a product or I’m buying search
ads on Walmart or Amazon that result in people
buying a product. And I’ll have brand advertising
associated with that. So you will buy — and I
know I’m not making it easy. But in the complexity of
definition of a market, like I will buy brand advertising
on Amazon because if people search for shampoo, I might
want to push Head & Shoulders for P&G. I might also want to buy the
keyword that gets people to buy a particular product. I might push a brand messaging
to show the benefits and efficacy of the product or I
might just get people to add it to their cart. It’s very complex, and I get
back to it all depends on what the client briefs us on
as the campaign objective. And then, we make
recommendations based on what we think is the best fit
for finding that audience to deliver against the
campaign objective. We make recommendations
to our client. The client ultimately signs
off on that, except when we get to programmatic. But that’s like another
15-minute speech. MS. LUBELL: Thank you. MR. STUART: Hey, Karina? Karina? If I could comment a little
bit — MS. LUBELL: Yeah. MR. STUART: — because
I’ve probably done — have personally done more
cross-media research than probably most people, upwards
of now three dozen studies publicly, both when I ran the
IAB when we were trying to develop Internet advertising
as a business and now for mobile. And I’ll tell you what’s
interesting about the whole cross-media thing is that if
you do measure media and you look to see what’s really
going on there — and you’re trying to understand a
number of different factors. Reach, as was
mentioned earlier. Cost certainly factors in. Messaging is a
component of that. There’s so many things
that come into it. And it varies dramatically
by campaign objective, by category, by audiences going
after because fundamentally what you’re trying to do is
match your brand’s marketing efforts and media allocations
to the consumer behavior patterns and then influence by
individual creative messages. So I’ve seen campaigns even of
late, in the last five years, where print was the number
one performing channel. Now, I’m the head of the MMA. So I typically wouldn’t say
something like that, right? It wouldn’t be sort of the
role of the organization. But that’s fine. We’re all about helping
marketers do it more effectively. So you look at a pure sort of
return on investment of media and you can measure — well, I
do think you can do that with, as mentioned,
multi-touch attribution. It’s all over. TV still would be anywhere
between — well, we’ve seen the studies in the last five
years — 38 percent to upwards of probably 65 percent
of the total mix. TV still has that
kind of relevance. Untargeted, linear, dare I say
boring old TV in some regards still really works. So it’s just — it is a
very complicated question. And I believe that there is a
lot of movement between those channels. Where it’s gotten extra hard,
and Joshua and his company would certainly enforce —
there are so many options today. There is so many choices that
it’s almost incomprehensively overwhelming. We did a little thing for
another project the other day. We’ve heard that just within
video and audio, that there was probably 1,055
combinations that you could pull together of different
formats, targeting capabilities,
sound-on, sound-off. I mean, it’s just — it’s
infinite now and that’s part of what’s made this world so
complicated, at least from marketers and the role that
they’re trying to do on behalf of their — you know,
ultimately their shareholders. MS. LUBELL: And now actually
if you can tell us a little bit about mobile specifically? I think Preston mentioned
— MR. STUART: Mobile rocks. MS. LUBELL: Preston mentioned
that, you know, digital really took off with mobile. MR. STUART: Yeah. MS. LUBELL: I think I’d
read that more than half of marketers’ digital ad spend
last year was on mobile. So how do advertisers
approach mobile? Is it independent from
the rest of digital? Do they think about different
forms of mobile differently for say video,
in-app, display? Can you tell us a
little about that? MR. STUART: Yeah. So I think I’ll sort of carry
on one step to what I was saying. So I think what’s happening
in mobile — I had a friend of mine, a guy who
was at Colgate. And he said — he used to
talk about you Internet guys. That’s how he
would talk to me. He said, you Internet guys,
he goes, you came along. You did like Internet video
and Internet search and you did Internet display and that
was kind of like — you know, like there were a
couple of choices. He goes, then he said, now,
you mobile guys, he says, you’ve given me
so many options. Like it’s almost overwhelming
the choices that I have, that I can go in and
try to figure out. And so, I think that, again,
we’re really struggling with not having the best of — I
think measurement does exist. I disagree with some of
what’s been said earlier. I have measured media. It is very possible
to measure media. I have high confidence in
using experimental design to do that. So I don’t know if there are
smarter researchers here that take that to task. So be it. I think you can. And I think marketers have an
obligation to figure that out. I don’t buy this, like,
you know, you can’t do it. So, so that’s part
of the challenge. But I think the issue with —
I think that what’s happening though, to kind of get back to
your question, is there’s just so much experimentation as we
try to slot these things in, to try to bring comprehension
to an incredibly complicated space. And that’s what’s making it
I think so hard for marketers because they just — they had
— you know, it was just a simpler world before where you
could slot stuff in and you knew what its purpose was
and you could, you know, fool yourself to think that
that was all there was. I’m not sure that was
really the case at the time. But now, it’s just — I just
don’t think that they know. And by the way, in my previous
studies, I mean, we’ve seen mobile — I don’t think mobile
is 50 percent of total spend. I don’t think it’s anywhere
near that today for most marketers. And I wouldn’t advocate that. Our research would suggest
that mobile — generally we’ve seen 15 to 20 percent of a
total, the full stack, TV, everything, outdoor, print,
anything else that’s in the mix, digital, desktop. And, but we have seen studies
that say it could be as high as 35 percent on
an optimized basis. And so, I don’t know. So it gets to the core
of yours, but yeah. MS. LUBELL: Okay. MR. ARRIX: I think what
contributes to that though is when you look at Facebook and
Google’s spend, which has — you know, is high and the
majority of the consumption for Facebook and Google is
done on a mobile device, I think that sort of checks
the box for mobile. MR. STUART: Yeah. MR. ARRIX: So I think you
have to take that into consideration when those
stats are that high. MS. LUBELL: Makes sense. So Kevin, I want to get back
to you and talk a little bit about online TV. But first, just because
we’ve been hearing about the marketers’ perspective on
things, Larry, you know, it’d be great to hear how
publishers are thinking about digital advertising. For example, you know, what’s
the most important factor when publishers are looking
to move inventory? MR. SOLOV: So there’s a lot to
unpack here and there’s no way to completely get into all of
it in the amount of time that we have. So I’m going to paint with
some broad brush and I hope you guys will all
understand that. Plus I never get you all
together in one place. So I want to take
advantage of it. Digital online advertising is
not sold like television or print advertising. Instead it is sold almost
entirely on a programmatic basis. And what that means is it’s
sold on an auction with real-time bidding. And there’s a lot of
implications of the switch to programmatic that has happened
over the last several years, one of which is that when I
started, an advertiser would say, you know, I look at your
demographics on your site and you’ve got 20 million
unique visitors a month. And I want to hit
them generally. So I’m going to do a $150,000
ad buy at a certain CPM on your website. Great. Because of programmatic
advertising, direct advertising has — display
advertising online has almost disappeared, for all
intents and purposes. And advertisers do not buy
websites and they do not buy publishers. They buy people. You guys all know this, right? Last week, you thought, hey,
I’ve got a vacation later this year. I want to go take
the family to Hawaii. Maybe do a Google search for,
you know, cheapest hotels in Hawaii or you’re a baller and
you just look for like most expensive hotels in
Hawaii or something. And maybe you look at airfare. And then, for the next three
weeks, wherever you go, you get ads from American
Airlines, from the Outrigger Hotel, from maybe a hotel in
the Caribbean that thinks, on, this person lives in D.C.
Hawaii’s pretty far away. If I serve them an ad,
maybe they’ll think I’ll go somewhere shorter. You get the picture. And we see that all the time. And it’s very much connected
very often to what you have searched or what other
websites you’ve gone to. So that’s, you know,
how things are done now. So the second key point
is that digital online programmatic buying is bought
and sold on an exchange, just like commodities and
just like equities. But there is a big,
big difference. This exchange is completely
and totally unregulated. And practically everything
that the SEC would not allow you to do on a commodities
or equities exchange happens every day, all the
time in this space. Also unlike a commodities or
equity exchange, very often there’s one company — happens
to be Google — that for all intents and purposes
controls everything. Google controls the
entire pipeline, okay? So they control — and they
have multiple products. They do it through
multiple means. They control the demand
side, the sell side. They control the exchange. They control DFP, that
platform that I’m — I don’t know a single publisher that
does not use the DFP platform. Like I can’t think of one. And the reason is it’d
be so cost-effective and inconvenient to get off it. And they’ve baked
anticompetitive actions into DFP to discourage you
from going anywhere else. Again, I’m giving you a lot. It can be unpacked,
you know, over time. On top of that, Google
actually owns many of the ad networks that are buying. Google actually buys off of
this and in this system that it controls. And historically that auction
has taken place as a two-tier or two-stage auction with
a first price and a second price. Historically Google has
reserved the last look in that process, which one could
argue pretty convincingly is a competitive advantage. They also have information
that the buyers and sellers don’t have. You know, someone that wants
to buy ads has to put in a bid estimating what they think
other people are going to buy at. Google puts in a bid. They know what
everyone’s already bid. So again, a lot. The bottom line is the system
is rigged and rife with conflicts of interest. Google is trying to
get ahead of this. They’ve just announced that
they’re switching to a single stage auction process. I would submit that that
will not solve the problem. They will still be in a
position to control who wins, where the advertising goes,
how much and which publishers get paid and frankly they’ll
still carve out a margin by the use of algorithms
and other things. So why should you care? Well, if you’re an advertiser,
you should care because you’re probably paying more for
ads than you should online. And if you’re a publisher, you
should care before — because you are probably getting less
money for your ad inventory than you should. It’s a pretty opaque process. And I would submit — I will
give you my personal opinion — everyone should care
because no one tech and advertising company,
especially one with a proven viewpoint bias, should have
control over picking and choosing winners
in publishing. My takeaways would be if you
don’t want heavy regulation, the only way to avoid it will
be for the DoD to — the DOJ to start looking into
antitrust action. Otherwise you’re going
to have to regulate. And the other thing I would
submit to you is if you fix the problem today, Google is
now so dominant in the space that any fix that is not
remedial will not make a difference. Subject one. Do I have time for one more or
do you want to — MS. LUBELL: Well, so I think you’ve given
us a lot to unpack in that. And I want to — I want to
get to subject two as well. But I also — I know, Wendy,
that ad networks and ad tech in general is something
that you have also studied. And so, I just wanted to give
you an opportunity to talk about that as well, the roles
of the different ad tech companies in facilitating
these transactions. DR. MOE: So this is a tough
spot for me to follow. So I think — you know, I’ll
probably leave the details of how the bidding and all work
to some of the gentlemen to my right who have more
direct experience with it. From my perspective, I look at
it more theoretically from the position of the role of
targeted versus untargeted advertising and how
broad-based is that targeting. My sense, and again, I’m sure
people on this panel will feel free to correct me if they
disagree, is that the various ad outlets have, you know,
allowed for various degrees of targeting. My sense is that — so
Larry focused on Google. I would actually throw in
Google, Facebook and Amazon into this targeting discussion
in that you could work — brands can work with Google
and Facebook for varying degrees of targeting. So if a brand wants to target
top of the funnel and not go micro target, that
is a possibility. And the issue is are there
outlets, are there appropriate pages, are there content
providers that provide the set of audience members that the
brand is looking to target. I see — I vaguely remember —
I’m hesitant to reveal this — to say this just because I
can’t specifically remember the source. But I remember yesterday some
— there was a news article that revealed that Amazon for
the first time has surpassed Google in the number
of product searches. And so, from that perspective,
you know, if we’re thinking about where — the power
of Google to direct product searches and consumers to
appropriate products, that is weakening and
yielding to Amazon. MS. LUBELL: Thank you, and I
don’t want to completely shift gears. But I think that it’s
important to understand all of the different facets
of digital advertising. And one important piece that
we haven’t spoken about yet is video advertising in
particular and how that exists obviously across a number of
different media, including online television. And so, Kevin, you know, I’d
love to hear from you about how online TV providers
are thinking about digital advertising. You know, is it more similar
to other forms that the panelists have been describing
or is it more like ads on traditional TV? So maybe if you could tell
us a little bit about that? MR. ARRIX: Yeah. I’ll say that I think
everyone’s getting a lot of digital loaded on them in
a short period of time. We went from, you know,
defining digital and what it means to a quick crash course
in programmatic, which has taken some a decade
to figure out. You know, just going back, I’d
be remiss not to say that I think digital all started with
Internet and it started with consumer optionality. You know, consumers had the
chance to get content on a computer. And while desktop is not as
prevalent as mobile today, you know, I think everything
about advertising is where the consumer is spending time. I work at DISH now. I’ve spent the last
20 years in digital. And we used to talk about
how much time consumers spent online and where
the ad dollars were. And ad dollars were still with
traditional television and traditional media. So that’s caught up and
there’s been a handful of drivers to that. But I think you have to
look at digital as it’s IP-delivered. It is the ad executions on
digital are display, are video and search. And that’s simplifying it. And then, you have sort of
walled garden platforms like a Facebook or Instagram or
Snap that are all digital. But they’re bringing a slight
— coming from a slightly different angle. Anyway, you know, Sling
basically was born, you know, four or five years ago from
the need of consumers wanting another option than, you know,
the sort of standard cable or satellite television where
they have, you know, hundreds of channels and they’re like,
well, I watch 20 channels and I want, you know, what is
commonly referred to today as a skinny bundle. And so, it really —
it came from consumers. I mean, consumers
demanded that. And Sling, you know, or DISH
made, you know, an acquisition of a company called Move
Networks and that was the — that’s the foundation
of Sling today. And, you know, our stats on
Sling are public in terms of how many subscribers we have. But in a nutshell, Sling is
live television delivered OTT. And OTT, you know, again,
just to sort of level set, you know, part of what we do
every day is make sure that we educate and define OTT. It is over-the-top. So it is content that is
delivered through someone’s device. You know, for us, 90-plus
percent of our consumption comes from the big
screen in the home. And that is — that is
delivered via an IP signal and not — and it is not going
through, you know, sort of the traditional cable or
satellite subscription. So Sling, to any consumer that
has it, looks and feels to the consumer like live television
because it is live television. It is the live television feed
that you would find when you watch TBS or TNT or NBC or
whatever station you’re tuning into on Sling. But it acts and it delivers
like digital because it all comes IP. So we’re at a really
interesting section. I know in the earlier panel
they talked about, you know, TV is not digital. And we actually are right in
the middle of the interaction of television and digital. And I believe that that
intersection is all sort of tied to addressable
advertising and is tied to live television
delivered over the top. And just to sort of, you know
— this isn’t all about Sling. I mean, DIRECTV now, YouTube
TV, Hulu Live are the four main, you know, companies that
are in the live TV OTT space. And so, for digital, you know,
we actually — we view it as however our customer
wants to buy. If someone wants to buy
via IO, that’s great. If someone wants to buy
programmatic, we are fully connected into all the
technology that enables someone for the first time to
buy live television in, you know, a nanosecond. And so, for us, digital
advertising, it gives us the flexibility to leverage
either, you know, existing third party data or consumer
— or sorry, brand first party data along with our subscriber
data to deliver targeted advertising, to deliver
addressable advertising. But in the end, it is all
digital delivered, data-driven video that we provide. MS. LUBELL: Thank you. I want to go back to some of
the discussion took us to some of these intermediaries that
play a role and particularly in programmatic and
the transactions. And so, Preston, I was hoping
if you could talk to us a little bit more about the role
of programmatic and the ad networks. DR. MCAFEE: Absolutely. So I want you to start by
thinking about if you want to buy let’s say a collectible
Pez dispenser, and those things exist, on eBay. You go. There’s a seller on eBay. eBay is a platform. You connect to that seller. You put in a bid. If you win the
auction, you’re done. They send you the — you give
them money and they give you a Pez dispenser. This is not at all how
advertising works, not at all. So first, there are at least
five ad networks — AdX, Amazon, AppNexus, OpenX and
Facebook Exchange all play the role of eBay. That is to say, in principle,
an advertiser could be connected to one of those. Say I would like to ad. Publishers say here’s
an opportunity. An opportunity is something
like you go to Breitbart. They then tell the advertising
exchange here’s who’s on my site. Here’s what
they’re looking at. Here’s what I know about them. What they know about them —
what they know about you could be a lot because maybe they
have access to data from Experian that tells you
— tells them what your IP address, what your household
income is and so on. So, and that could go through
any of those five exchanges. And that’s how it
worked circa 2012. Starting shortly after 2012,
we see the rise of what’s called the
demand-side platform. A demand-side platform
aggregates buyers. That is to say, advertisers,
into larger groups. If I want to make an auction
not work very well, have monopoly power, right? And that’s what demand-side
platforms were all about is if I get all the advertisers
together, rather than submit the highest bid, the second
highest bid, the third highest bid and so on, I
just submit one bid. So that was a — it was
a beautiful vision, the demand-side platform. We’re going to aggregate all
the advertisers and we’re in the process — in a way, we’re
going to help them also do their job. But then what happens is
there are multiple demand-side platforms that compete
against each other. There are at least
five of those. So that is to say the
advertiser puts their ad into a demand-side platform. The demand-side platform
sticks it into not one, but all five of the exchanges,
hoping to get a publisher. Meanwhile, the publishers —
so what this process did to the publishers is cause
them to lose money, right? They’re not making as much
money because now there’s power on the
advertisers’ side. How did the
publishers respond? They did the same thing. Those are called containers. A container is a group of
publishers who run their own auction. So now what we have is a
sequence of auctions being run by the different players in
the marketplace with different levels of data and so on. It has made it
somewhat symmetric. But then the thing that you
need to understand about advertising is that everybody
thinks they deserve half. So each of the — each of
these parties — so that is there’s a publisher who’s
supposed to get the money in the end. But meanwhile, they run their
ads through a container, which is just basically collecting
the publishing opportunities and running its own auction. It’s getting bids from
exchanges, each of which are taking a cut, generally
substantially less than half. Those then are getting bids
from demand-side platforms, each of which is
taking roughly half. And then finally, there are
— well, they’re actually advertising agencies because
there are very few advertisers who actually ever
participate in these markets. That is to say, they’re
represented by someone who also has creatives
and all that. And so, in this process, what
we now have is basically three sets of auctions when one
would have sufficed and been more efficient. So — MS. LUBELL: Thank you. MR. ARRIX: I would jump in and
say that that is — that does not describe the programmatic
world that we live in at all. And I think it’s important
for the audience to understand that there are
different flavors. I think what Larry and what
you guys are describing is open auction. And open auction is like
everyone is putting in and it’s display. Like there is — there is so
much display ad inventory, it’s insane. You know, when you go to a
webpage, do you see one ad? No, you see five, right? And so, it’s so commoditized. So I think what you’re
describing is open auction. And it’s messy. There’s middle players. But in our world for video, it
is very clean and it is very effective. And I’m not an advocate
for programmatic. I’m an advocate for whatever
solution my client wants. And we have one partner that
we run — that use their pipes to connect to the demand side. And that partner gets paid a
fee by us which is, you know, about to become transparent,
you know, because the industry is asking for it. And then, also gets
paid by the demand side. And it eliminates
the middle person. It eliminates all those, you
know, four or five middle players. It’s very effective. It’s very clean. And I think programmatic
is very — it’s hard to understand. I mean, it’s complicated. But it has evolved. It started out as open
exchanges, very messy. But there are other flavors. There’s personal auctions. There’s programmatic
guaranteed. You can control however you
want your inventory to exist. But I do think that the deck
is stacked not in the favor of the display publisher in
programmatic open exchange. MR. LOWCOCK: I think there’s a
couple of things we just need to — and Kevin’s spot on. There’s a couple of things
that need to be clarified, which is content and
environment still matter. So while we’re chasing
audience, quality environments matter. Reach still matters and
effective reach still matters. So there’s PMPs. There’s private marketplaces,
which drive a lot of programmatic and, you know,
while the media landscape is fragmented and consumers
are going everywhere, you categorically know that
there’s — for any brand or advertiser, there might be
10 to 15 sites where you can reach 80 percent of the
audience quite effectively in one go. And again, that’s why you have
to factor time into all of these decisions. The other thing that needs
to be factored in is on the transparency side. There’s a lot of work
being done on supply path optimization. So you’ll hear the term
SPO bandied around in the industry. And that’s about eliminating
all those middlemen that are like clicking the ticket
all the way through. And it’s come from this
perverse belief by I’ll say the publisher side of the
industry, which is every eyeball deserves to be
monetized, and that’s not true. THz end result should be not
every eyeball deserves to be monetized. Not every page deserves
to carry an ad. It should be an ad should
only be served if it’s highly relevant and needs to be
there for the consumer. Otherwise the publisher needs
to make a decision not to carry an ad. And the reason why the
industry has moved in the way that ads tend to be everywhere
is so little of the traffic is actually direct. So once upon a time, in a
magazine, you turned the page, turned the page and you’d
sell 20 pages in, you know, a 60-page magazine. What happens now is publishers
get traffic from search. So then they’re like
desperately trying to monetize when someone comes in, reads
an article, disappears and none ever see them again. So these are sort of the
dynamics that are going on. MS. LUBELL: Before I switch
gears and ask you all about data, I just wanted to give
everybody at this side of the table an opportunity to, if
you had anything to add to that. No? Okay. You know, I want to go back to
something that you had started to touch on, Preston, when
you were talking about the different — all of the things
that they know about you, these various intermediaries. And so, you know, it
sounds like ad networks are responsible for
collecting a lot of data. So perhaps you could describe
some of the data that are being gathered and maybe even
if you could tell us a little bit about what is shared with
advertisers or publishers and how the data is
used generally. DR. MCAFEE: So, sure. The first thing is there’s
almost nothing about you that can’t be known. From your IP address, both
Acxiom and Experian sell — they make money
on selling data. So if you want your financial
wherewithal, have you bought a car in the last year, things
like that, the stuff that would be in your credit
report, that data’s available. The data, actually from
an academic researcher perspective, is really a
delight because you’ve seen your own credit report. It’s very extensive
about the big things. It’s also quite expensive. So there are many advertisers
that don’t use it. Second, it’s pretty easy for
an advertiser to track what you do on their site — or
excuse me, a publisher to track what you do on their
site, what pages you visit, how much time you spend. By the way, this isn’t —
there’s academic research that was published by people who
were at Yahoo and then later at Google that say we have
a pretty good sense of what you’re looking at on a page. Like you probably think,
well, it’s the whole page. How could you know
what I’m looking at? But it turns out people
move their mouse. At least on a desktop, they
move their mouse where they’re looking, or at least enough of
them that we get a pretty good sense from the page
what you’re looking at. With mobile, we actually even
have a better sense because you don’t — probably don’t
know this, but people only use — 85 percent of their eye
gaze goes to the top half of their screen. That is, you don’t read the
bottom half of your screen. You scroll or
read the top half. So we know what you’re — we
have a pretty good sense of what you’re looking at,
what pages you’ve visited. And then there’s — let me
tell you about cookie sharing. And cookie sharing is — and I
think Larry mentioned this — so a cookie tells — like they
write into a cookie what you do on a site. So it tracks in your browser. That is, it’s on your
computer, not stored necessarily off your computer,
what you’re doing on that site. And then, cookie sharing
means if I searched for United Airlines and then I go to some
other site, that other site can access the cookie that
United left on my computer and say, oh, he looked
at this flight. Let me offer him another
flight, a competing flight or a similar flight or maybe
even exactly the same flight. That’s what you get
with Amazon, right? They just ask — you know,
show you the same product, sometimes embarrassingly. That’s not actually quite as
bad as it sounds in the sense that it’s not necessarily
the advertiser seeing that. And then, let me finish with
just Facebook and Google. Facebook and Google basically
don’t let data off their own sites. They are — if you want to
access Google data, you have to advertise on Google,
not outside of Google. And that actually
protects the consumer. And the same is
true with Facebook. That is, the Facebook data —
and I’m not on Facebook, and yet Facebook still has
extensive data about me and that’s because of the
“Like” button, right? So the “Like” button means
that when I go to a site that has the Facebook “Like”
button, Facebook — this is called pixeling, by the way. If a publisher — excuse me,
if a company controls one pixel on the page, they get
to see everything that the browser supplies, including
your IP address and so on. So that allows them to track
your behavior and some pages have as many as 2,000
people tracking you. One thing that’s true — so
Facebook and Google require you to use the
data on their site. They don’t let the data out
and that is consumer-friendly. But it also is sort of
monopolization-friendly because they have
such extensive data. Let me say — let me finish
with there are really very simple things for consumers
to do to limit all this and almost nobody does. So that is to say the cost to
a consumer of having cookies not be written — like you
can use the Brave browser, B-R-A-V-E. That’s the name of
a competing browser. It just prevents all
those cookie writing. So consumers have a really low
cost remedy that they don’t take. And so, that’s a — so on the
other side of all this, like they can see all this. They have tons of data. But we don’t do the first
thing mostly as a society to protect ourselves. MS. LUBELL: Thank you, and
because you brought it, unless Wendy — DR. MOE: Yeah. I’d love to just add
to that a little bit. MS. LUBELL: Please. DR. MOE: So, I mean, Preston
describes a process of collecting what pages you’re
looking at, what you’re searching for, et cetera. But I think from a marketer’s
perspective, the value in that data is what you can infer
about the person in terms of their preferences, their
lifestyles, et cetera. And so, the idea is to if you
can observe someone behaving across the Internet, you
can infer what this person’s preferences for type of
car, preferences for type of computer, political leanings,
sports preferences and interests. And so, that is
all extractable. However, there’s still
accuracy issues with that. So I always use my — when I
teach this to my students, I always use myself and
my son as an example. So for myself, I read
a lot of technology. I read a lot of
business and finance. So all of my devices and all
of my machines think I’m a man. We won’t get into what my son
— that’s not necessary to share. But the point is there are
definitely accuracy issues. And the larger the ad
networks, the more accurate — the more visibility they
have on your behaviors. And the more visibility they
have on your behaviors, the more accurate they become. And that’s where we start
seeing — I’ve thought a lot about Facebook versus
Google and the size of their footprint across the Internet. So Google has a large
footprint across all of their sites, as well as other
content sites and anything that carries an ad. Facebook has actually an even
bigger footprint in terms of their ability to observe
your behavior within their platform, anything that
carries that Facebook “Like”, any page that caries that
Facebook “Like” is visibility for them into your behavior. And on top of that, they
have the social network, the behavior of the friends
on your social network, et cetera. And so, I don’t know what
the solution is in terms of privacy and tracking. But it’s just an interesting
thought process that I always take my students through in
terms of Google is the poster child for tracking. But if you think about what
they — what all of these firms have insights into, it’s
not clear that they have the biggest footprint
on your behavior. MS. LUBELL: Thank you. Yeah. I wanted to continue
that conversation. You both obviously brought up
Google and Facebook and people have talked about increasingly
Amazon as accounting for the majority of digital ad spend. And, you know, some people
argue that they’ve been able to maintain their dominance
through this walled garden approach. So, you know, I would love to
hear sort of how people think, you know, the role of data
and what actually that walled garden approach means. And I think, Greg, you
mentioned that you had some thoughts on this as well. MR. STUART: Okay. DR. MCAFEE: You do now. MR. STUART: So yeah, I mean,
listen, I’ll just stake a position and maybe other people
can sort of disagree with. You know, so at the end of the
day, I represent the marketers. So I sit with the marketers
on it a little bit. And the thing that’s caught me
interesting about the walled gardens, I mean, in some
regards, you know, Google and Facebook have built very
strong advertising businesses that really work
for advertisers. They wouldn’t use
them if they didn’t. And, you know, if they — they
have decided, for whatever business reasons, that they
don’t want to share everything out beyond that. And it seems to me that, you
know, eventually the economic marketplace may decide that
they can continue to do that or not. I don’t know what’s
going to happen. I do know that there are a lot
of companies, publishers, not that far below in size, not in
the kind of sort of usage that we see. But if you look at, you know,
companies like CBSi, who has extensive Internet properties
and mobile properties, or Merida (ph), there are a
number of other sort of companies out there. You know, they sit at just,
you know, if Google and Facebook sit around 210
million unique, according to what comScore says on a
monthly basis, they sit in the 180, 190 million range. They’re also equal size. And they will share
the information. So there are
options out there. And so, I don’t know. I kind of — I get the — I
guess in some regards I get the hysteria that might exist
because somebody’s not sharing something back with me. But for the moment, it seems
like a business decision that they’ve made that for whatever
reasons they’ve decided to do that. And so, I’m sure that’s going
to be contradictory to a lot of other people. But I’m surprised that
marketers don’t find other options. If they really wanted that,
then go develop the other options. MR. LOWCOCK: I’d like to weigh
in there because there’s a couple of — there’s sort of
— there’s two parts of the conversation really which
is there’s the supply of inventory and there’s
the supply of data. And media and media buying
and planning is really the convergence of both of
those things these days. And, you know, there’s
companies like Amazon, Google, Facebook, you know,
they need to be named. They have a concentration of
power on both sides and that is advantageous for them
in the entire ecosystem. And then, I want — I
don’t want to paint them as monopolies. That’s up for the
DOJ to decide. The real question that needs
to be sorted out is it’s the consumer behavior that
concentrates and the buying decisions that flow from that. So even if they were
independent companies or the fragmentation in the
marketplace, because consumers spend so much time on YouTube,
on an Android device, on Google search, that’s where
the power comes from and that’s where they get a
disproportionate share of the spend. And that’s the same for the
Facebook ecosystem and, quite frankly, that’s the truth for
the Amazon ecosystem as well. If you unstitch all that and
say, well, they need to share the data outside of their
walled gardens, that’s interesting and, you
know, a powerful argument. The flipside challenge to
that is it doesn’t change the consumer behavior which
is consumer behavior is concentrated on a handful
of platforms anyways. So even if I can pick up
Facebook’s data or Amazon’s data — and, you know, Preston
mentioned Facebook Exchange and Amazon sells media
elsewhere and Google sells media elsewhere. It doesn’t change the fact
that consumer behavior is dominated and consumer
interest is dominated on a few core environments. And then, everything else is
sort of the mid- to long-term. MS. LUBELL: Thank you. DR. MOE: And I think that’s
the traditional argument for targeting. And so, there are larger
discussions about food deserts, right? So you have valuable consumers
in high income neighborhoods who are — who everyone
wants to try and reach. Alternatively, you have other
consumers who — and because everyone’s targeting those
high value consumers, there’s price competition. There’s a lot of competition. They get good offers. They get good prices. Alternatively, when you look
at some of the less served neighborhoods, you see they’re
not as valuable as consumers. They don’t have a
lot of buying power. And so, there’s not a lot of
competition in those spaces. And then, as a result,
prices are higher. Not a lot of competitors
and retail go into those locations. And as a result, you have
these food deserts where there are no fresh groceries
for those communities. And you can see there’s
a parallel, right? So it’s the outcome
of extreme targeting. If you start targeting more
and more, you’re going to — there’s a gravitation toward
the valuable consumers, however you want to define
the valuable consumers. And so, that goes back to
my earlier discussion of broadening some of your
targeting efforts to try and reach serendipitous consumers
who don’t fit the profile. MR. LOWCOCK: I want to just
— I want to take that because there’s something that Mark
touched on earlier, which is there’s a separate
conversation which is advertising provides an
important social function of making content available to
people for free and funding important, you know, public
services and public goods. And that becomes I guess
the next challenge for advertisers, which is making
the — like not everything is — programmatic is
a buying method. It’s not a way we think
about buying things. And we do do brand
advertising in digital. It’s not sort of all
over-targeted and driven that way. But there’s a bigger
question which is where does advertising — where does the
bulk of advertising money get weighted and spent and where
do other organizations like news organizations miss out
because money is just shifted away from them. And how do you ensure that
advertising funds everything? Because as much as I’d love
infinite advertising budgets, I don’t. I have finite
advertising budgets. Audiences are finite as well. And campaign
cycles are finite. What I am increasingly aware
of is how do we ensure that we fund appropriate content and
serve the communities which buy our products, as well as
ensure that we don’t drive dominance of opinion or views
into certain pockets of the advertising industry. MS. LUBELL: Oh, yeah. I’m sorry. Preston? DR. MCAFEE: So I want to agree
with Joshua about one thing, which is advertisers do have
alternatives, especially about data. But from an antitrust
perspective, if Google winds up being a necessary part of
a buy and Facebook winds up being a necessary part of a
buy — that is to say, there are customers that you just
can’t reach without using that particular platform, that is
— that gives them dominance in a market. And then, that says there’s
lots of things they can’t do to extend that dominance. One of those things might be
buying other companies that would actually give them even
more power, things like that. So even if advertisers
actually do have a fair amount of alternatives and both
Google and Facebook achieve their market power through
legitimate means — that is, not through something illegal
— nonetheless, there may be implications from the
necessity that advertisers face in using those platforms. MS. LUBELL: Larry? MR. SOLOV: So I would just
— again, coming back to programmatic and ad networks
and maybe talking a little bit about what Josh was raising. I mean, the bottom
line — and Kevin too. I mean, the bottom line is
display advertising online has been completely commoditized. It’s practically dead, okay? And the reason you’re reading
about BuzzFeed laying off 200 people, even though they were
the darling of the advertising world, is I am sure in no
small part because of things like that. Programmatic has allowed a
concentration of power in the hands of a handful of ad
networks, which in and of itself might not be bad. But let me give you a
Breitbart-specific example. But please understand this
example is not limited to Breitbart. After President Trump was
elected — this may come as a complete shock to people in
this room — a lot of people were very, very, very —
and I mean very upset. So maybe you know
some of them. Some of those people run
and control ad networks. In November of ’16, there was
a customer that did almost 10 million impressions
on Breitbart. To be clear, they did
not place direct buys on Breitbart. These were programmatic buys. But the customers they wanted
to reach were on Breitbart. Their advertising on Breitbart
was successful and it had been for years because these
were longtime advertisers. So clearly that’s where the
ad networks representing the advertiser was bidding
to reach those customers. I looked at that same
advertiser in November of 2017, that advertiser did
only 6,000 impressions on Breitbart. It took me a long time. But I got to people who could
answer the question for me at this advertiser. And here’s the bottom line. This advertiser never
boycotted Breitbart and never complained that its
ads were on Breitbart. They had no idea that their
advertising was being blocked from Breitbart and not
reaching their customers on Breitbart’s website. So what happened? After 2016, a bunch of ad
networks got together and for openly and admitted political
reasons blocked programmatic buying on Breitbart. And they did so without
telling at least some of their advertisers who had
never complained. And they did it despite the
fact that advertising on Breitbart had been
successful for years. And it was for purely
political reasons. Breitbart has —
is a news site. Their editorial position
is openly conservative. And people were
pissed Trump won. And frankly, they wanted
to take it out on somebody. So stop and think and take
Breitbart out of that and give it to your favorite news site
or your favorite opinion site, right? I think that should
bother everybody. That was maybe an unintended
consequence of programmatic. But that should
bother anybody. I told you in the beginning
I approach this from a very simple standpoint. I like the free flow of
information and ideas. Ideas should be iterated
upon, discussed. And that makes a
better democracy. And I think Josh was referring
to — and I could be wrong, so jump in — the fact that a lot
of news companies are not able to sustain investigative
journalism on digital dollars, digital advertising dollars. But I would strongly suggest
that if solutions are looked into that, those solutions
not be Facebook and Google or anyone else saying, well,
we’ll fund the investigative journalism. Now we understand. Because those are all
companies with strong viewpoints. And that would — I believe
that would result in a terrible result. A free media that relies
and can fund itself through advertising is a safeguard. We should cherish it. We don’t want big companies
coming in and solving the problem that they’ve helped
create by saying, well, we’ll give these people money
because we like their journalism. We won’t give people money
because we don’t like their journalism. MS. LUBELL: Thank you. I have — there are two other
topics I want to make sure that we cover before we wrap
up in the next 10 minutes or so. But before I do that, does
anybody have any responses on the power of the ad networks
or anything like that? MR. SOLOV: One thing I would
point out, it has happened recently that major companies
have bought ad networks, major content companies. I’m referring specifically
to the Time Warner merger. A lot of you might know
a company, AppNexus. It was bought by the company. And so, I think I’d just
put a placeholder in it. I think that raises a number
of questions as ad networks begin to get bought and folded
into larger companies that are also content providers
— MS. LUBELL: Thank you. MR. SOLOV: — for another day. MR. ARRIX: I’ll
just add onto that. Again, you know, in the spirit
of clarification, like I’m not going to comment on the
backlisting, you know, that’s — I’ll stay clear from that. But in terms of ad networks,
so a company like AppNexus and, you know, The Trade Desk,
I mean, it’s important to understand sort of my
definition of an ad network is a company that is aggregating
supply and arbitraging. And they’re sort of
identifying a gap in the marketplace and they’re
saying, okay, I’m going to cobble together a network and
I’m going to sell it to the demand side. And I’m going to take a piece
of what I bring back to the publisher. And I think AppNexus and
others that are in the programmatic space are
not only — you know, they aggregate supply. But they are the technology
pipes that lead to the demand side. So it’s really important to
differentiate between is it demand side, is
it supply side. And, you know, the
programmatic space used to be three pieces. It used to be demand
side, exchange and supply. And supply and exchange have
consolidated in the last few years. And so, I think it’s
important to note that. I also think just objectively
— and again, like DISH doesn’t have a horse in
this race in terms of like display-based websites. But having worked in that —
on that side of the business, it’s really — display
digital advertising is really challenging for publishers
that are sub whatever million. It’s really hard. I mean, that market is
dominated by players that have scaled audiences, that have
ease of solutions in terms of buying. And it’s tough. It’s very difficult. And that has driven, in my
point of view, sort of what’s been happening to BuzzFeed
and Vice and folks like that. MS. LUBELL: Thanks, Kevin. So I want to ask you all about
something that came up several times during the
earlier panel. You know, and maybe we’ll
start with you, Joshua. You know, numerous observers
have raised concerns about digital advertising. And during the previous panel,
we heard from Marc Pritchard about P&G’s concerns about
ad fraud and brand safety. And, you know, how do those
problems play into your decision on which channel
to use, digital versus traditional media. And what is — do you have a
sense of what the industry is doing to address
those problems? MR. LOWCOCK: Okay. So they’re two — they’re
two separate issues. So on the ad fraud, I’m going
to encompass in that also things like quality, such as
viewability and ensuring the ad is — like even if your
ad is seen by a human, is it actually being seen in a
way that has an impact. It’s not the 0.7 of a second. So we use — so the industry
uses third party verification companies for that that
are accredited by the Media Ratings Council or
others around the world. We negotiate the contracts
with media partners so that neither our clients nor we are
on the hook for ads that are delivered to non-humans. So we call it invalid traffic,
IVT or SIVT, sophisticated invalid traffic. And what happens is if the
third party verifications detect fraud, so that comes
out in the report in a buy. We say we’re not paying for
those fraudulent impressions. And then, I get back to
that time constraint. We either have to get those
impressions credited back to make good for the advertiser
or the bill gets reduced so that we don’t pay for it. And that covers
quality as well. So we might buy against
viewability metrics. And we say that the ad has to
not only be seen by a human, but we want the ad seen for
a certain amount of time. There’s all sorts of third
party tools that account for that and we might pay a
premium to ensure that’s delivered. On the brand safety side of
things, that’s a lot more complicated because, again,
we use third party tools to manage for brand safety. What we talk — one of the
things I developed in the industry is the Advertiser
Protection Bureau, the APB. And that’s a consortium of
all the agency holding groups through the four As. We actually share information
about brand safety incidents. So if an ad appears in front
of an ISIS video, it’s not just the advertiser that gets
exposed that has to worry about it. All the agencies get alerted. And that way, we can hold a
publisher accountable and get that video taken down. Safety covers though a myriad
of sins because you have the ISIS video which no one
wants to be against. And then, you might have —
the oft-used example is the airplane crash that airlines
don’t want to be against. And both are brand safety
concerns for various advertisers. That becomes much more complex
and difficult to manage. And when you get to
user-generated content on YouTube or Instagram or
Snapchat, it becomes another problem. And so, one of the big
conversations that we have at the moment with the industry
is advertising is a privilege, not a right. I don’t have infinite budgets. You need to have safe
and quality environments. And a lot of the pressure
needs to be taken off not only just the agency and the
advertiser, because we’re trying to police
for all of that. A lot of the pressure has to
also go back to the publisher and the advertising
environment because, when I grew up in traditional media
in TV and print, when we had a breaking news event, we would
go dark in advertising because we knew it wasn’t
a safe environment. We need much more
responsibility taken on the digital side for that. But I also would say you need
a lot more responsibility taken on the regulatory front
side of things because — and I’ll point to the Christchurch
example, which was a terrible tragedy. There was no ads running in
front of that content, which was — you know, from an
advertiser perspective, that’s great. The platforms were slow at
taking that content down, which is terrible. But then you’ve got bad actors
that continue to publish that content. You need a lot more regulatory
action being taken against individuals that publish bad
content online because bad actors will always be
incentivized to break rules. And unless you take those bad
actors and punish them for their bad behavior, like
I’m constantly playing whack-a-mole. And the platforms will be
constantly held accountable for people that have got bad
behavior, that will not care that the platforms are
being held accountable. You need to get the bad
actors penalized as well. MS. LUBELL: And Greg
— MR. STUART: Yeah? MS. LUBELL: — are mobile
marketers thinking about these issues the same way? MR. STUART: Yeah. Listen, you know, it’s funny. I think, you know, the digital
world — I’ve been doing digital now for over 20 years. And it’s a wild
and wooly place. There’s no question that
there’s a little bit of an Old West mentality there that
complicates sort of marketers as we figure out new things. And as I think Joshua makes
a good point, you know, bad actors, for whatever reason,
have now been given access to all sorts of things that they
never would have access before Internet. We actually have launched
both brand safety for CMOs to really help them figure
out how they manage that significantly better. I’ll be honest with you. I’m a little surprised the
number of CMOs who did not have a brand safety. I mean, to the best of my
knowledge, there is only one brand safety officer at a
major corporation who has that actual specific title. Joshua, maybe you
know a couple more. But it’s limited in a way
that I’m surprised at. The other thing around fraud,
I mean, fraud has really been a desktop issue and it is a
very serious one that needs to be addressed. The data suggests that we
guess it will probably come mobile’s way here as mobile
has now become sort of much more dominant. And we’re still trying to
figure out sort of — there’s just a lot of creative people
who want to do bad stuff. And so, I think there just
needs to be an increasing concerted effort — I’ll
call out P&G and others advertisers. Unilever is my chairman, and,
you know, the work that we do there to try to aggressively
work against that. It’s hard to keep
up with, admittedly. But I think that there’s
been some progress. It’s funny. These verification companies
didn’t exist five years ago. And they’re each now —
there’s three of them that are probably 600, 700 employees. So they’ve come a long ways
and are sort of getting better — too slow, but
getting better. MS. LUBELL: Thank you. So we are nearly out of time. But there is just one last
topic I wanted to make sure that we get to. And that’s convergence. So maybe, Kevin, I’m going
to go back to you on this because, you know, companies
like Sling are obviously — or DISH are in an interesting
position since, as you said, they really are at the
crossroads of traditional TV and digital TV. And so, I’d be curious to
hear what kind of convergence you’re seeing between online
and offline TV advertising. MR. ARRIX: Well,
we’re seeing a lot. I mean, I think, you know —
I think it ties predominantly to, you know, another topic
that we’re covering, which is data. And, you know, there’s
obviously — you know, there’s an issue around data leakage. But there is a big opportunity
I think for organizations that do have compelling first party
data to, you know, partner together or to properly mine
that and leverage it with their advertising partners
and agency partners. And that is kind of the
critical piece to all things advanced television
advertising-related. As it speaks to DISH, I mean,
you know, we were first to the market in terms of selling
cross-platform addressable. So what does that mean? That means that we work with
any given marketer or agency and we put together the DISH
footprint and we put together the Sling footprint. And if someone is looking for
an in-market truck buyer, you know, we cobble it together. We don’t sort of — we don’t
differentiate between, oh, this is sort of linear
television and satellite-based and this is OTT. We just look at it as we have
— we have very, very little duplication. But we have a subscriber
footprint that we put together that we bring to
the market as one. So I think that’s one
version of convergence. I do think that — I know
measurement was mentioned in the earlier panel. We have been working with
comScore on a cross-platform addressable measurement,
which is great. So there are things that are
coming that allow the demand side and marketers to
view something sort of holistically. I think that’s really,
really important. But we see convergence as
something that is happening and something that will
continue to happen. And actually another level of
convergence I think is — you know, when you look at the
power of Facebook and Google and Amazon, it’s — you know,
again, it’s one platform that has massive scale. You know, that’s why we need
to start getting our, you know, measurement and
definitions all sorted because there needs to be — you know,
there needs to be a consortium — you know, continued sort
of consortium effort around television and OTT to bring it
together because television is still super effective. And if you can make it easier
for the market to buy, whether it’s local, regional or
national — so I think convergence — I look at it
as convergence of linear and digital, but also convergence
of — and I think technology will be the key catalyst for
that — convergence of what have been competitors to sort
of consortium collaborators. MS. LUBELL: Thank you. And any final thoughts
on convergence? MR. LOWCOCK: I just want to
— I just want to add that I guess the last pillar of that
convergence is the convergence of commerce because we’re
seeing the integration of like data, media and then
connecting it to the commerce side of things. And that’s going to be
the next big battleground. MR. ARRIX: Yeah. DR. MOE: And I think this goes
back to this idea of how much visibility you have on
the consumer, right? So the key really is the data. So we’ve been talking about
Internet, occasionally about mobile. And that’s starting to come —
that data is starting to come together. And you can get a
cross-platform or cross-device profile of a consumer. Once we start bringing in
TV, you have a smart TV, and Google is on that. All the TV viewership behavior
starts coming in as well and that adds to the accuracy
of their profile for the consumer. One thing that we haven’t
talked about as much are the broadband providers. Someone mentioned Time Warner. They also have
access to the data. But they haven’t been playing
the same Internet game along with Google and
Facebook up until now. DR. MCAFEE: Well,
they do sell the data. So others are buying the data
to do — even like magazines and so on. DR. MOE: Well, that’s a
whole different issue. DR. MCAFEE: Yeah. MR. STUART: I mean, Karina, in
some regards, this is kind of I think — if I maybe even
summarize a little bit, I think the challenge that we’re
having is that a business that’s deep in innovation and
change, I mean, it is more dynamic — I’ve been
doing this a lot of years. And I’ve been in
the heart of it. And I am shocked at how much
it changes and how quickly. And so, we do need to find
some balance in that sort of innovation development for the
greater good, the things that were talked about on the
previous panel, versus I think, as Larry was trying
to point out and others, you know, there are unintended
consequences to some of that. And so, the degree to which we
sort of manage against those — I don’t even profess
to have those answers. But it’s definitely a very
interesting business right now. There’s no question
about that. MS. LUBELL: Sure. DR. MCAFEE: Well, smartphones
were adopted really fast. MR. STUART: Amazing. DR. MCAFEE: But smart TVs
are going really slowly. MR. STUART: Yeah. Yeah, right. Yeah. MS. LUBELL: But I think
part of the — obviously the dynamism of this industry is
one of the reasons we wanted to make sure to
hold this workshop. Given the hour, I think we will
forego audience questions. I apologize for that, but
hope you will all come back tomorrow for the second
portion of our competition and television and digital
advertising workshop. And before you go though, I
just would like you to thank all of our panelists
for their presentations. (Applause.)

Leave a Reply

Your email address will not be published. Required fields are marked *