It's funny because I feel like we have done a lot of work that
advertisers can benefit from, even if they might be seen as ad
announcements. For example, you gotta start with content. All the
acquisitions we made -- whether it's Discovery Networks, we have every episode
every created of "South Park," we just announced every episode of
"CSI" ever created is with us -- so you have all that content, and
then you have our originals. We're doing a big project with Stephen
King about his novel "11/22/63" produced by J.J. Abrams and then
James Franco is going to star in that. You can't talk about big TV
without considering that. Then even if you look at the other part
of the spectrum, which is all the YouTube creators, we did a deal
with Freddie Wong. And we'll have more of those announcements soon.
We might save them for the actual [NewFronts] event on April 29. So
that is really good for marketers because I can sell them
sponsorships and adjacencies and all that good stuff.
What I've been really working on quietly in the background is
building our programmatic stack. And for us programmatic means not
just media automation, but more interestingly for marketers is the
ability to use big data to make better targeting decisions.
In what ways?
A lot of marketers want to bump their databases up against our
database, find the right person and serve the right ad to the right
person. That kind of innovation to me may not be an interface
innovation, so much as it's a targeting innovation. The user may
never see it, but hopefully there's just this feeling that I'm
always being served relevant ads.
And that's a big deal because it's for ads against TV
content?
Right. Relevance is what it's all about. People who say 'I hate
ads' -- they don't hate ads, they hate irrelevance. And if we can
solve that problem with data, that's great. Then they're not
intrusive, they're entertaining or informational. So from
programmatic to data to the content, all those things allow me to
go to market with a lot of confidence.
So how are you planning to sell "11/22/63" to
advertisers?
"11/22/63" will probably be out around this time next year, if
not a little earlier. Maybe January. So we've got time. The
principal photography starts in May in both Dallas and Toronto. A
lot of things we're thinking about include pretty conventional
ideas like integration, sponsorship and looking at -- for lack of a
better word -- shoulder programming, or webisodic programming. Just
short-form stuff. Our menu would be things like "Anatomy of a
Scene," "Anatomy of an Episode," director's cuts. None of that has
been set, but that's the menu we'll be looking at with our partners
at Warner Bros. They know that we cut the show with conventional
commercial breaks. And we're going to want to do what most big TV
does, which again is integrations, sponsorships and shoulder
programming. There's one thing that I'm thinking about that I can't
reveal before the upfronts.
Hulu's been doing originals for a few years, but the
slate keeps getting bigger. How are you thinking about selling
originals compared to the shows Hulu gets from TV networks? Do you
think about them differently?
Well, a lot of our content providers have agreements that we
don't sell shows specifically. A lot of the content we acquire from
companies, they've got their own grown-up sales forces. So I think
the companies like that we're a distribution outlet, but they don't
want us to conspire against their sales forces. As a result, we
sell audience-centric packages, genre-centric packages and big
bundles of shows. And the market seems to be fine with that. With
the originals, I am allowed to sell show-specific, and that gives
me the opportunity to sell sponsorships and integrations in a way
that we don't into the content that's acquired.
You mentioned earlier the ability for a marketer to
bring in its own dataset and match it against Hulu's. How does that
work?
To be clear, that's the technology stack I'm building. We're
very much in progress in building a programmatic ad stack to
transact in a programmatic video marketplace. Is that too much
lingo?
Eh, I used to cover ad tech a lot.
So the promise is any marketer who has a dataset, the theory is
you bump up their dataset against our subscriber base, we find the
matches, we serve the matches one piece of creative and serve the
people who aren't a match another piece of creative.
Hulu has two audiences: the people who watch videos on
the free service and those who pay a monthly subscription for Hulu
Plus. Both groups are shown ads. Do you sell them
differently?
The Hulu audience is the Hulu audience. We sell all together.
More to the point, we don't differentiate between a desktop, mobile
or connected-TV user. When we go to market, we sell those
impressions across.
So the number of ads people are shown is the
same?
Yes, the ad load is the same on different platforms.
What about free Hulu versus Hulu Plus?
Free Hulu is a heavier ad load than the subscription model.
Are there major differences in that ad
load?
It's the same advertisers. On the free service, it might be
three ads per pod [commercial break]; on the paid service, it might
be one or two. And that fluctuates with time of year, and frankly
it can fluctuate with pockets of unsold inventory. That goes back
to yield management. Where can we go ahead and lower rates to
stimulate demand and where can increase rates because we're sold
out.
I spoke with a media buyer recently, and Hulu came up.
And this person said they're not able to get as much inventory as
they want from Hulu. And other people I've talked to said a big
push for Hulu is to get the free Hulu video player embedded on as
many sites as possible to increase the number of impressions for
the free service. What can you tell me about what they're referring
to?
Who's the buyer? Because I want to make sure they know I can
sell them.
So you look at an election year. Last year there are candidates
in New Hampshire that want [to buy ads against viewers in] Southern
New Hampshire. This is all I have in one town in Southern New
Hampshire, so you get constricted that way. Depending on the
geography or the demo, maybe we run into that.
On the second question, we were built on proprietary technology
because we wanted to control the user experience and make sure it
was beautiful for the user. And we needed to build for all these
platforms, rather than be reliant on a third party to help us plug
into all the living-room devices. So when it comes to having other
people's players take our content, that gets back to talking with
the content owners and making sure everyone's comfortable. There's
opportunity there, but it's still in progress.
Hulu rolled out an 7-second pre-roll ad a couple years
ago. Now Hulu's former CEO Jason Kilar's new streaming video
company has adopted a 5-second pre-roll. Is there a trend toward
shorter video ads?
We were early with it. The market still likes their 15s and 30s
for a couple reasons. One, they run on TV as well as in the digital
space, so they can travel. Second, it's a media math challenge. If
you're an agency planning group and you're trying to figure out
you're gross rating points, you're currency is 15s and 30s. So if I
show up with 7s or 8s or 10s, they're like, 'This doesn't fit our
model.' So there's a little bit of inertia built into the entire
system for the old model, and different-length ads could be
creatively called for. But it's still hard in an era of
accountability for people to say, "Well, how do I make sense of
what a 7-second ad costs and how much value I should put on
it?"
You mentioned advertisers taking their TV spots and
repurposing them online. But is there also a need for advertisers
to think about digital-native or even Hulu-native ads?
It's a reasonable conversation to have. Most people want to
spend their money on media and not creative, if they can get away
with it. You can give me 10 cool, new companies, and if they all
have 10 different ad models, that doesn't scale very elegantly. The
nice thing about 15s and 30s in video environments is that you can
run those in a lot of places.
Late last year Hulu
restructured its branded content division and stopped creating
standalone programs specifically for brands. That seems to go
against the tide as publishers like BuzzFeed and Vox Media open up in-house agencies to create
content for brands. Can you walk me through the
rationale?
![]()
Everyone's running their own business, and we decided the
business we want to be in is more branded commercials. When I say
commercials, you hear that and think 15s, 30s and 60s, as opposed
to branded content. We're creating television shows, and there's a
lot of people creating branded content. So I'm not saying we
couldn't be an outlet for longer-form branded content; it's just
that we're not going to create it. That's not our focus. We will
create integrations, sponsorships and branded commercials, but
long-form branded content, there's so many people doing it so well.
We could potentially be a distributor for that. It's a subtle but
important shift we made.
Lastly, NewFronts. What can you tell me about the sales
strategy for this year's NewFronts talks?
The NewFronts presentation is one day, an hour and a half. Then
the upfront market goes throughout the summer. We're very much
talking to broadcast buyers and their clients about our core
offerings, and the idea is to figure out a way for them to put
money down in advance. And we try to secure their commitments, and
in exchange we give them fill-in-the-blank, cost-per-whatever. It
could be completions, viewability, sponsorships, seasonal
tentpoles. Pretty much anything scarce is where we'll try and
transact.
One thing I remember from talking with media buyers
during last year's NewFronts season is that many of them wait until
after the TV upfronts to get their NewFronts deals done. The
window's closing each year, but there still seems to be this
holding pattern to the NewFronts until the TV upfronts. Is that
something you expect this year, and what can you do about
it?
The people who transact in the upfronts, if you transacted last
year in the upfronts, you're already having conversations right now
about what to expect, what do you need, what should we be talking
about. Everybody's doing a needs assessment. Some of our customers
will go in course with TV, and some will go broadcast, then cable,
then digital. It's really case by case. So as a publisher we go out
and assess the market and where the market's going to turn.
What's your assessment so far for this
season?
I think TV's going to continue to be fine. Viewability could
also constrict supply and prop up TV. That might offset the
explosion of available content in digital. So many people --
whether it's magazines, TV networks, pure-plays like us --
everybody's creating more and more video. They're trying to attract
TV replacement dollars. So as there's this explosion of avails,
there's also a contraction because of viewability. At the same
time, television continues to be very time-tested medium. That's a
long way of saying, I think there's going to be another healthy
upfront in terms of dollar volume. And as far as CPMs, the good
guys should get higher CPMs, and we're the good guys.
To hear more from Peter Naylor on TV, digital video and the
future of ads, come to the Ad Age
Digital Conference on April 14 and 15 in New York.