Public Workshop on Competition in Television and Digital Advertising

By | August 28, 2019

DR. ATHEY: Thank you. Very good. So I’m very grateful to
have been invited here. And it looks like a really
terrific group of people gathered. So I’m going to start with
a fairly broad overview of a number of the issues
that will get developed substantially further. Before I kick off, I do
want to reiterate I’m here on my own as a
Stanford professor. But I have had a lot of
interactions with this industry, with a variety
of firms, including platforms as well as
advertisers and other industry participants,
including several of the cases that were
referenced just now. So to start out, I want
to really just take a few moments about the
role of advertising. And I think, you know,
most of the industry participants here
today don’t need to be convinced. But one of the frustrating
things as an economist, as especially I went out in
the early days to educate people about the issues
coming up in digital advertising, was that, you
know, so many people think about their experience as
a user and they think, you know, well these
advertisements, I never click on them. They don’t matter. You know, nobody
measures them. You know, they’re not very
important for the economy. And I think it’s just
really important I think to underscore the crucial
role that advertisers play in the economy. You know, and I
appreciated that, coming from the perspective of
advertising platforms. But I appreciate it even
more working with startups in Silicon
Valley and so on. And so, when you think
about being a young company going to get
funding, the first thing you need to do is
figure out your customer acquisition cost. And basically like you
don’t get funded if there’s not a path to
scale your business. You’re going to put in
a bunch of fixed cost investment. But you’re going to make a
plan for how do I expand, which markets do I go to,
which countries I go to. And small changes in the
cost of advertising can actually make or
break a company. And then, even with a
large, well-established, you know,
multibillion-dollar company, if there is a
change that affects, you know, the cost of clicks
on Google or, you know, the efficiency with which
you can reach consumers through TV, that
is material. It’s a board-level
discussion that affects the bottom line and,
again, it really affects strategy and investment. Another thing that I’ve
been involved in, both from the antitrust
perspective and from the industry perspective, is
just how does this affect consumers. The fact that these
advertisements are what’s alerting you to the
existence of new products. When a company — when
people don’t even know what your product is,
you need advertising. You can’t survive
without it. And then later, when
you’re doing price competition, again, it’s
a very important way. So we should
all know this. But like especially for
economists, I get very frustrated when people
think advertising isn’t important because how
do we think people get information? How do we think the
economy actually works without this
in the middle? And then, you know, it’s
also easy to forget the content producers and
what their incentives are. And I’ve actually spent
a lot of time myself studying the
news industry. And that’s something where
you have to get up every day and decide where to
send a reporter, how much to pay a reporter. And as we see, they have
stopped doing that quite so much in a
fairly larger way. Newsroom employment has
gone way, way down per unit of population. And so, when you — you’re
basically having to make an economy decision every
day for the return on investment of your R&D. You’re doing R&D and it
only pays off if you can get consumers. And of course, how you get
consumers is also affected by all of these platforms,
how many consumers you get and certainly how much
you make for consumers is affected by the
ad platforms. And so, generally,
like our democracy, our markets, at the very core,
advertising is crucial. And then, of course, we
can think about all these consumers of content. And how do you think all
these — why do you think all these things that
you find on the Internet exist? Well, they exist either
because they’re reaching consumers by finding them
by advertising and then they’re monetizing with
advertising once they get there. So the kind of canonical
economics view about the economics of advertising
is to think about it as a two-sided market. And some of the early
economic theories really were kind of motivated
by traditional media. And the idea is that, you
know, media is basically a platform for matching
advertisers and users and consumers/viewers. And so, the viewers are
just going to choose which content by how much they
like the quality of the content and maybe
the ad load. And the advertisers are
just trying to, you know, reach an audience and so
they look across all the content and see how they
can piece it together to reach the audience that
they’re interested in. And then, from the
perspective of the media, they’re basically trying
to get a hold of these consumers and then
basically auction them off to advertisers. And they’re basically
selling access to the advertisers. And of course, especially
in the news media, there were problems of having
sort of local monopoly. And in fact, you know,
one of the main roles of antitrust for news for
a long time was actually like they were badly
behaved monopolists instead of now kind of
barely surviving things that we need to subsidize. And it’s really
interesting how just the change in technology took
them from one extreme to the other. And the main thing in the
old world was that they, for a variety of
technological reasons, had sort of unique access
to a set of consumers. And when you have that,
you can then really exploit the advertisers
for access to those consumers. But one of the things that
changed in the digital world was the
fragmentation. So, you know, when I
think about TV, I started working in this digital
part of the industry in 2007. And at that point, you
know, things had gone very quickly with search and
display advertising. And so, there was a group
of people that thought that sort of TV was going
to follow very quickly. And there were very heated
arguments within the industry about sort of
whether TV would move sort of quickly to look
like digital or not. Most of the people who
had been in the industry a long time were on the side
that it was not going to move fast. And I think part of the
logic behind that was just that this was an industry
where the content producers and the
advertisers were already connected. They were connected
in an offline world. There just weren’t that
many of them, so, you know, fairly highly
concentrated, the top 200 advertisers now accounting
for 66 percent of TV spend. You know, it’s always been
pretty concentrated like that. So you just don’t
really need a lot of intermediation to come in
and solve the problem of connecting these small
groups of people. So if you bring in digital
or you bring in a platform in the middle that tries
to take a cut, like what’s the point. You know, that platform
in the middle would have a hard time taking much
surplus, which is really very different than the
environment in digital where things are so
fragmented, both on the advertiser side and the
publisher side, that the role of intermediation was
much more important from the very beginning. So, you know, when you
think about some of the things that change in
this model when you go to digital, first of all, you
know, people worried about how just there was going
to be so much content. There would be an infinite
amount of content. But of course what matters
is consumers’ time. So there could be more
and more Web pages. But there’s still a
limited amount to consume them. And I think through this
sort of first generation, kind of the time people
were spending was, you know, was the
limiting factor. But as mobile has come
along, we’ve actually gotten more addressable
time from people. You can consume content in
places where you couldn’t before. And so, that’s really sort
of changed the sort of supply of user
attention as well. And then, we have these
more recent trends of digital video
and cord cutting. On the advertiser side,
when you go out to this digital world, all this
fragmentation creates a bunch of possibility,
but also challenges. So the advertisers in
digital are typically, you know, perceiving the
possibility to be much more careful with
ROI measurement. And indeed when you’re
using digital to acquire consumers, you’re going
to be, you know — most sophisticated firms, the
good, big firms are all doing a ton of
experimentation, a ton of measurement to see how
effective is each ad on each channel and so on. But they are sort of
thwarted in this attempt to be good at measurement
by a whole variety of challenges. First of all, it’s
difficult to track users across these
different channels. We had cookie problems and
now these different types of changes in how people
are using devices makes it harder for advertisers
to do that. And of course solutions
to that then start running afoul of increased
privacy regulation. And so, we have to
be cautious about the unintended consequences of
regulations that make it difficult for advertisers
to go across channels. They can in principle
advantage the big thing, the big groups that can
track their users within that single publisher. And I have a paper in
Management Science
that fleshes out some of
the theory around this. They also have to worry
about duplication. When they can’t track the
consumers, then they keep giving the same consumers
the same ad over and over again to when
it’s ineffective. They have to worry about
attribution and fraud and safety. And so, it’s maybe not
surprising that as we’re going to get in sort of
potential issues of market power, that when firms do
have market power, there are quality issues that
can arise, both in terms of information that
are being given to the participants and their
ability to control their experience. And so, we’ve had all
sorts of different types of scandals around, you
know, advertisers’ ads appearing where they
didn’t want to, as well as advertisers paying for
things that were never viewed by a human. But they can’t really
escape all of this now because if they’re trying
to reach all consumers at all different points they
want to reach them, it’s hard to get out
of these things. So just really quickly,
this isn’t — this conference is mostly
not about digital. But just to kind of give a
little bit of grounding on the different types —
some of the different types of advertising and
the way that it occurs, you know, search
advertising was the early really super
profitable thing. As users, we all think
about that we do a search on Google or Bing and
then we see an ad there. So we can think about this
as sort of this is like first party, that Google
has an ad platform that serves ads directly
for Google. It’s also a form of native
advertising where the ads are specifically designed
to be effective on Google. And one of the things
that’s really special about search is the
consumer goes there hoping to click on a link. That is their goal, to put
in a search and click on a link. And so, if you show an ad
that gives you a link that might serve your need, you
know, you’re really in the mood to do something
right that second. And even if you see that
same consumer somewhere else doing something
else, if they’re surfing Facebook, they’re actually
not in the mood to click on a link to buy
auto insurance. That was two hours ago. And so, they have sort of
a temporal monopoly on the person’s attention
of a specific sort. They’re able to both
identify it right at that very moment. And so, they’ve been able
to really sell access to their users. Now, it’s less known to
the casual consumer that there’s actually, you
know, a whole platform on the other side where
there’s intermediation in search advertising. There are publishers
who are making money by putting search boxes. There are browsers and
so on that are funding themselves by directing
consumers to search engines. So there’s another side
of this market as well. But people — end
consumers don’t see it. But that’s a very large
part of the market and actually that part of the
market turned out to be very important for
entry in competition. Now, you could think
about the Yahoo deal with Microsoft where all of
Yahoo’s searches were served on Microsoft as a
way to acquire traffic. And it was really probably
the only way left for them that Microsoft could
achieve a break-even status by acquiring
more traffic. AOL was also
another example. But over time, there were
also lots of deals with, you know, distributors,
with companies like Comcast and AT&T. They all had deals
to distribute search. Intermediated —
first-party display advertising — I call
these walled gardens — have gotten really big and
popular today and they’re really getting a
lot of attention. So Facebook and YouTube
would be big examples where, again, you have a
special kind of content. And users are spending
enough time there that you can make ads that are
specifically for that content and sell
directly to advertisers. And of course you can only
do that because you’re big. Newspapers would
love to do this. But none of them are big
enough to really get the scale to get advertisers
interested or to be able to operate the platform
or do the targeting. So they all end up using
intermediated display advertising and going
through middlemen like Google’s Ad Exchange,
which was previously known as DoubleClick. And that’s a way to
intermediate between publishers and
advertisers. And competition at that
level can be very, very important because, if you
think about it, you have two advertising exchanges
competing against each other. They’re both kind of
giving publishers money. That’s like
perfectly suitable. Money is money
to a publisher. And the advertisers are
trying to reach users. If you have two of them
that are similarly sized and are monetizing
similarly, that’s very tough competition. In fact, like, you know,
it could be like kind of zero profit competition
where all the advertising dollars go straight
to the publishers. But because it’s — partly
because it’s so cutthroat and there are so many
economies of scale, you end up being very
vulnerable to having one company get big and
then taking a lot of the surplus. So the advertiser
pays money. But only a small amount of
it goes to the publishers. So competition among
these exchanges is super important. It’s a very
complicated industry. You know, you could spend
two hours explaining how the digital
industry works. There’s, you know, ad
exchanges, ad networks, supply-side platforms,
demand-side platforms. They’ve all been renamed
and rebranded a thousand times. But all of these parts of
the industry tend to be pretty concentrated. And so, we have to look
out for what’s happening in all of them and also
interplay between the different markets. But let me now move on and
talk about a broader set than just digital. When you think about why
companies are advertising, there is — this is
classic kind of, you know, marketing 101 class
type of material here. We talk about something
called the marketing funnel and it’s becoming
— also we can talk about a consumer journey, which
these are closely related. So if I’m trying to
tell people about a new product, like I’ve got
this great new marketplace and it’s a service and
people just didn’t know, you know, that you could,
you know, get this great service, then you need
to tell people about it. And one of the big ways
to do that is to do a TV campaign or some kind
of video campaign. And that’s going to give
you maybe 30 seconds to talk to your consumer and
really tell your story. That’s going to be —
then you’re trying to get people interested. And then, you can go all
the way down to intent, evaluation and purchase. Now, a consumer actually
isn’t going to go linearly through that. They’re going — they may
have a journey back and forth. And so, then the
advertiser is going to be trying to reach the
consumer at different points of the journey. But the message
will be different. And also the format and
the time when you want to find that user will be
different, depending on your purpose. So advertisers are
really not monolithic. We can tend to say, oh,
they’re a brand advertiser and, oh, they’re
a direct marketer. But a lot of the studies
that I’ve done, many large companies are doing both. The companies that I work
with are definitely doing both. And then also, as was
alluded to in the opening remarks, these things
are — not only are they different and not only
are they not directly substitutes, but actually
they’re often complements. And just in a classic
example, you can think about running a TV
campaign for a clothing brand and then
mailing out a catalog. And people are less likely
to throw your catalog in the trash if they
saw your TV campaign. And so, you’re typically
coordinating these campaigns an thinking
about one and the other. And in the digital world,
you know, the TV campaign can go along with search
engine optimization to make sure that when people
come and look for you, they then find
you on the Web. So these things often are
going together rather than being substitutes. So we can’t think of this
as just all one big soup where everybody’s
competing with everybody. Now, I’ve kind of put some
rough — rough, you know, labels for where different
types of advertising go. But I can say that, you
know, there’s exceptions to every rule and there’s
different ways the media can be used. But roughly people are
thinking about, you know, television and now digital
video in this awareness and intent part and
things like search and performance best response
at the bottom of the funnel. And of course in the old
world that was also going to be the catalogs and the
mailings and so on that are trying to get people
to buy at that moment. Some challenges and
changes in this view are really more and more of a
nonlinear journey and also the fact that you’re going
to be interacting with this consumer on different
platforms and really have a hard time knowing
whether this is the same consumer you’ve seen
somewhere before and therefore knowing exactly
how to talk to the consumer. So now that we’ve talked a
little bit about channels, let me give a little
historical perspective. So this is a chart
going from 1926 to 2012. And it shows the evolution
of advertising shares. And it shows — here’s
radio, this stripe at the bottom, which is like
amazingly constant over time. After the entry of TV,
it’s basically stayed about the same. TV came in and of course
took a big share of revenue. And then, the
Internet came in. It didn’t grow as fast
as TV did initially. But it’s certainly
getting much larger. And then newspaper share
has shrunk all the way through this period. Now, looking at a chart
like this makes you think about competition. But I want to be very
clear that we have to think about competition
on different sides of the market. So this could — if the
dollars per hour stay constant, then all — you
could think about there being not a lot of — this
could be consistent with not having a lot
of competition for advertising dollars and
just competition for user time. So going back to the
initial framing, you could — it would be possible
that you would see a chart like this. But the advertisers are
not treating any of these things as substitutes. It’s just that different
media are attracting the consumer time and then
advertisers are following the consumers where
their time is. So my colleague Matt
Gentzkow wrote a nice paper in 2014, so
the data ends there. But he tried to compare
the entry of TV in the ’40s to the entry
of Internet. And so, he shows over here
that when TV comes in, radio gets hit in
terms of time spent. There were people
listening to the radio at home and they stopped
doing that and they started watching TV. And so, radio lost time. When the Internet came in,
TV and magazines really seemed to — I mean,
newspaper and magazine really took a hit
in terms of time. But then if you look at
these, the top chart now is a repeat of this
minutes per day. The bottom chart is
revenue per hour. What Matt was arguing
was that when TV came in, radio’s revenue per hour
kept right on growing and it grew for another 20
years pretty steeply after they lost that time. So it wasn’t — it doesn’t
look like, you know, the competition between TV and
radio were pushing down ad prices. But when you look at the
newspapers, they really take a hit in terms of
revenue per hour after the entry of the Internet. And so, it looks like they
may have been hit more in terms of the competition
for ad dollars. Here’s a more recent chart
just put together with some eMarketer data to
show the revenue per hours in recent years. A few more things you can
notice, we do see again, you know, the magazine,
print and newspaper, the top ones all like, you
know, getting worse even faster. We see this rise of
social, which is going to be, you know, basically
you can think of Facebook is getting better and
better at monetizing it. First, like
they had no ads. And then, they had
really bad ads. And now, they’re sort of
getting better and better, growing their network,
more advertisers, better matching with advertisers
and consumers, video ads and so on. So they’re just making
more per hour while — and we also see this light
blue line, digital display and search
continues to rise. And video, non-TV video
also going up a bit in terms of revenue per hour,
again as we’re starting to see things like YouTube
and so on monetize. Now going into thinking
a little bit more about trends in TV and video,
we see that, you know, the average time per day
with video by U.S. adults has been
relatively flat. We’re seeing a bit,
especially coming in very recently, a sharper rise
in digital video versus TV. And then, we’re also
seeing changes in the mix. The top highlighted is pay
TV viewer growth, which is negative, and non-pay TV
viewer growth, which is positive. And then, I also want to
highlight that some things are changing actually very
rapidly very recently. So this picture is
just showing that how, according to PubMatic,
that digital video ad spend went from, you know,
83 percent desktop in 2017 to 54 percent in 2018, so
just a very rapid sort of shift of advertising
dollars just like in the last year. We’re also seeing other
trends of course, the rise of streaming services, in
fact such a proliferation that we’re starting to see
some people are claiming that they’re reinventing
cable because you’re going to re-aggregate all of
these different streaming services. But we also see a little
bit of a plateau and these are Statista projections. But they’re
projecting that the subscription-based video
like Netflix is going to be more flat
going forward. An interesting thing from
the advertising market perspective is that a
subscription-based service that takes consumer time
and doesn’t have ads on it of course creates scarcity
for the time that’s left. And so, in fact, that can
raise advertising prices just by sort of
restricting the amount of time that you can reach
certain consumers. And while that may be less
of a big deal overall, it can be a bigger deal
for different kinds of consumers. And of course you all know
the younger consumers, certainly my teenagers are
spending, you know, a lot of time on YouTube. Even my eight-year-old is
spending a lot of time on YouTube. And, you know, spending,
you know, less time and then on Netflix. And like, you know, they
hardly know what a TV ad looks like basically. The young people are like,
what’s that, you know? I don’t get it. So and then from a
marketer perspective, it then becomes difficult
to reach these people. They’re saying, you know
— they call them the unreachables, these
Millennials and Gen Xers that are mobile first or
mobile only can be harder to reach. And then, that means
that marketers have to go across a wider
range of channels. And in some sense, you
know, they might want to deliver them a 30-second
video, but they can’t because these people are
never in a context where you can actually shove 30
seconds down their — I mean, 30 seconds
down their throat. They’ll take 10, but
they won’t take 30. There’s like the browser
tab, and maybe they’ll take seven, depending
on where you are. So it actually — you
know, the advertisers’ demand for this just can’t
get met for some of these people. And so, then they end
up having to go across channels to try to
reach these people. And you can think about,
you know, there’s going to be some inefficiencies
because when they go from ad and desktop to TV,
they’re going to reach some people they already
reached and that’s going to be inefficient. And then, they
add in mobile. They’re going to
double the reach again. But they have to go there
if they want to get this other set of consumers. So these are some of
the challenges that advertisers would have. A few other kind of
trends to highlight, we’re certainly seeing,
you know, changes in monetization models
for publishers broadly. So for a while, the idea
was that, you know — and a lot of economists, I was
very skeptical actually about how successful
newspapers would be in reintroducing
subscriptions. You know, some of the
elite newspapers have been pretty successful
with that. And then, we’re also
seeing lots of other services using a
freemium model. And so, actually from
inside of a freemium company, they think about
the free users as actually part of a
marketing funnel. So they’re going
to attract users. They’ll show them ads to
kind of keep the lights on and pay for the services. But actually the main
reason to get all those free users is that it’s
a conversion funnel that some of them will then
convert into paid users or they’ll also attract more
friends to get paid users. And they need
scale economies. They may need lots of
users in order to optimize their algorithms that
they’re recommending music or something else. So, you know, they
need the scale. But they’re going to
basically not be making much money on those. And then, they want to get
them through this purchase funnel. So of course there’s
lots and lots of thriving freemium businesses
coming out. And just to sort of think
about these are just kind of sort of, quote,
unquote, “typical” numbers for like a young
freemium company. They might have 95 to 98
percent of their users would be free. But their revenue share
might be more like 20 to 50 percent because the ads
are just not monetizing them that well, while 2 to
5 percent of the users are covering the rest of
the revenue from the subscription. But from those kind of
numbers, you can see that if you have a middleman
that is taking a large share of the advertising
money before you get it as a service, that’s going to
be a big problem for you. And even cutting the
margins of all the intermediaries a little
bit can make a big difference for investment
in, you know, picking the right songs for people
and new innovations in providing these services
that we all love. So small increases in
efficiency can make a big difference. Here’s just showing
The New York Times
subscription revenue has
grown quite a bit and subscriptions have grown
for a number of players. A next challenge
is measurement and attribution. As I mentioned to start
with, advertisers are trying to get better and
better and really putting — you know, there’s
huge R&D teams. There’s actually like
probably a hundred PhD economists out there whose
full-time job it is to do advertising analysis for
top firms, especially tech firms. But they face challenges. Some tech economists
who work inside of tech companies have papers
showing that the signal-to-noise ratio
is just really hard. And, but then, ignoring
— running experiments is really difficult. You don’t get enough data. It’s hard to
measure success. But then, if you don’t
run them, you get really misleading results. So, you know, companies
are still using a lot of fairly crude tools. So just as an example,
when you’re going across channels, the standard
thing to do is what’s called a marketing
mix model. You know, I teach about
these in my classes but, you know, not with
a lot of enthusiasm. They’re not very rigorous
from a statistical perspective and can lead
to misleading results. But basically, you say
here’s my brand sales, like when did I do all of
my different advertising campaigns and I just run
a big statistical model to try to say, you know, gee,
did sales tend to go up around the time I did
this, that and the other thing. And it’s really
pretty rough. Now, one of the things
that’s been really exciting is that people
are trying to make that better and better
with microdata. So we’ve moved away from
like the very simplistic things from the past. But it remains a challenge
to actually do this well. And some of those
challenges arise from the difficulty getting data
and getting the right tools. And some of those are
not being provided by the folks that the advertisers
are trying to work with. So the advertisers
actually have to go and lobby and put a lot
of pressure on the intermediaries to try
to get them to give them enough data to actually
do the analysis. And sometimes that is
successful, and sometimes not. So here I’m arguing that
you can randomize and some interesting things is now
you can link ad delivery to purchaser. You can do that
potentially with lots of different data. Geolocation data is
becoming a new source of — you know, you show
somebody that and then you can see if that phone
actually went into the car dealership. But this measurement
again is imperfect. There’s lots of
different challenges. Advertisers might have to
put data into a company. But then, that data might
be shared with their competitors indirectly
in creating audiences. So when you put the data
in, you’re leaking some of your knowledge about your
customers to competitors. And then we’re having
again this big focus on location. But some people have a lot
more access to location data than others. The advertisers have also
been saddled with problems of targeting delivery. Like they ask for things
and it’s not getting done the way they specified. They have
viewability problems. So you’re paying for ads
that no human could ever have viewed. You have fraud, that
people are — publishers are just faking
impressions entirely. And then you — also
an advertiser might say please don’t show my ad
more than 10 times to a user. But in fact if you audit
it, they’re getting shown a hundred times
to the same user. And so, all of these are
sort of symptoms of also an immature industry
but also sometimes insufficient incentives
to provide high quality. And there’s been, you
know, a bunch of attention in the industry and
some of the industry associations have,
you know, pushed here. We’ve got this P&G tells
digital to clean up, lays down new rules. You know, and so on. And just here’s one of the
studies that was done that sort of kind of
targeted a lot of this. You had gross impressions. So this is what the
advertiser was paying for. The legitimate ones
was a smaller set. And they found — this was
an audit study that was done by — and only 43.2
percent of the impressions that were paid for
were actually even like possibly getting shown to
someone and forget about if they actually scrolled
down the screen and actually saw it, but that
it would even be possible for them to see it. So, you know, they —
these are — well, a little bit strong. There was some accounting
for that in the viewability, but a lot
of fraud and waste. So let me just — this
is of course an antitrust agency here. So I’ve been talking more
from an economics and business perspective
in the talk. But let me just close up
with a few thoughts from antitrust perspective. First of all, as has been
very well documented in search and in other
related markets, every time you look at these,
every textbook treatment of all of these things,
you start with economies of scale. More users, more
advertisers. More advertisers,
more users. More data, better
algorithms. Fixed costs of all the
R&D, all the science and so on. And in particular one of
the places where economies of scale are the most
important, and I really feel this acutely having
worked with a well-funded startup inside of
Microsoft that was a search engine that had 6
percent market share when I started working with it,
as well as, you know, lots of actual startups. You know, when you’re
early days in a platform setting, trying to get
enough scale is really crucial. That’s when you’re
very vulnerable. And anybody that does
something, either accidentally on purpose
that pushes you down — for some reason, Google’s
algorithm, you know, kicks you off the page for a
little while, that’s like make or break
for these firms. And so, anything you can
do to stop someone from getting — solving the
chicken and egg program, stop them from getting
a toehold is incredibly important. So I spent a lot of time
in search thinking where could you get a chunk of
users that could actually get that scale going. If you were somehow
excluded from getting those, that would be very
challenging to ever get started. Market power is
ubiquities, but not inevitable in ad tech. And so, we do have
examples where there are competing platforms. Even now, like Uber and
Lyft, a lot of surplus is going to drivers and users
while they subsidize every ride. You can read about it
in all of their filings, right? So in the process —
especially in the time of intense competition for
the market, both everybody anticipating, hoping for
the future to have more pricing power, they will
subsidize on the way up the scale curve. And stopping that
competition can be challenging for a
welfare perspective. And generally we just need
to be very cautious at the entry point where
somebody’s coming in and trying to gain scale
against an incumbent. That’s a very delicate
matter and they can be squashed pretty
easily along the way. And again, I’ve talked
about some of the consequences — low
quality, low transparency, high prices, high
take rates of the intermediaries, product
markets that rely on advertisings function
less well and are less competitive and entry can
be — entry of all the companies that we know and
love that have brought us so much surplus can be
jeopardized if these markets don’t work well. So this is just
a summary slide. I think I’ve covered
most of the points. But, you know, just really
to sum up, advertising plays this outside role. Competition and efficiency
are super important. Media has traditionally
been viewed as this two-sided platform with,
you know, monopoly power over users and selling
access to advertisers. Things have gotten
much more complicated. And that creates all sorts
of issues on both sides of the markets, the publisher
side and the advertiser side. Video is now this digital
battleground for consumer attention. But seeing what’s going to
happen on the advertising side of the market will
be quite interesting. So we can see the
consumers moving. But what is that going
to do to the advertising markets? That’s really the big open
question that we’re all grappling with. The advertisers are going
across media for balanced audience reach, but their
tools are still maturing. And I think a lot of these
quality and fraud issues are really just starting
to get the full attention that they deserve. And I’m, you know, hopeful
that as people go out and measure and they get
educated about what’s going on, they can do
a better job, you know, identifying problems and
getting, you know, better quality services for
all the participants. So I will stop there. Thank you. (Applause.)

Leave a Reply

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