Tip:
Highlight text to annotate it
X
MATT SCHECKNER: Good morning, everyone.
We tried to get you out of the rain as soon as we could.
We apologize.
My name's Matt Scheckner.
It's my treat to serve as executive director of
Advertising Week.
And this is the kick-off off to nearly 100 seminars that
will take place over the course of the next five days
here at the Times Center, which will be our main
headquarters, as well as at BB Kings and the Paley Center,
and all over many other locations with our sister
organizations who are staging events during the week.
I want to be brief and get on and get off.
We've got great stuff coming.
I just really wanted to thank our dear friends and partners
at Nielsen.
It's a joy having them involved this year, for the
first time in a real meaningful way.
You will see Nielsen, members of their esteemed posse and
intellectual high-powered group throughout the week.
So I want to thank Jenny and her team, my friend, Sabrina
Crow, who I'm not sure is here.
I know John Burbank is here.
And just really give a warm welcome to the Nielsen group
as part of our Advertising Week family.
And now, on with the show.
Thank you.
[APPLAUSE]
STEVE HASKER: Well, welcome, everyone.
Thanks very much for joining us.
My name's Steve Hasker, from the Nielsen Company.
And it's a great pleasure to be able to be in front of you
today and introduce Nielsen's Online Campaign Ratings.
At Nielsen, we have a two-part strategy in and around our
media business.
And the first part of that strategy is to improve all
forms of measurement within and across
TV, online, and mobile.
And the second part is to help our advertising clients and
our media clients improve the effectiveness of their
advertising.
Online Campaign Ratings is a very big step, we believe, for
Nielsen, we hope for the online advertising space and
for the media marketing space more generally.
So it's with great excitement that we have a chance to
announce this product and get moving on it.
What I want to do first is very quickly introduce the
folks that are here on stage with me.
Next to me is Mike Murphy from Facebook.
Next to Mike is Joel Lunenfeld from Moxie.
Next to Joel is Suzy Deering from Verizon.
Next to Suzy is Kate Sirkin from Starcom MediaVest. And
next to Kate is Ari Paparo from Nielsen.
I want to make one extra comment on a panelist who's
not with us today, and that's Lucas Watson from P&G. Lucas
and his wife are expecting a baby any minute now.
And he felt it inappropriate to fly to New
York and join us.
So he's with us in spirit, and Kate will make a few comments
about Lucas and about Procter and about how the folks in
Cincinnati think about this.
So what we're going to do with the next few minutes is, I'm
going to sit down and be quiet.
Ari's going to stand up and walk you in brief through the
product, give you a sense of what it is and how it works.
We're then going to ask Mike to give the publisher point of
view, so he's going to stand up on behalf of Facebook, but
also online players more generally,
and talk about that.
And then Kate is going to say a few words about the
advertiser perspective and the agency perspective of
Nielsen's Online Campaign Ratings.
Once we do that, we'll all sit down and we'll give Suzy and
Joel a chance to join the group and answer a few
questions from me, then I think, more importantly, a few
questions from you all, should you have them.
So without further ado, I'd like to introduce Ari.
ARI PAPARO: Thank you, Steve.
I've been very passionate about online advertising, and
about online advertising measurement, for years.
And so the opportunity to improve the ability to measure
online, especially with respect to brand advertising,
is something that I think is really exciting.
And that's why what we're talking about here, the Online
Campaign Report, is especially groundbreaking.
Over the next two days, and over the course of Advertising
Week, you're going to hear a lot of people.
I imagine, if you keep count, it's almost
like a drinking game.
You know, take a shot every time someone says, online
isn't getting its fair share of advertising dollars.
I guess probably somewhere around 15 to 20, would be my
over-under.
And then the second thing you're going to hear over and
over again is that brand advertisers are reluctant to
spend online because of, among other things, measurement,
among other things, the question of, what do the
metrics mean?
How do I compare them to online? how do I compare them
to other media?
So that's what we're trying to address here.
And what I'm going to do is I'm going to play a short
video that explains the concept.
And since the video will probably be very
self-explanatory, I'm going to spend about three or four
minutes explaining it afterwards.
So with that, I'm going to [UNINTELLIGIBLE].
[VIDEO PLAYBACK]
-Here's the problem with online ads.
How do you know who really sees them?
Here's the answer.
Nielsen has a way to turn impressions into a clearer
picture of your audience without
compromising their privacy.
Here's how it works.
Your ad gets a Nielsen tag.
Thousands view your ad.
Through the tag, Nielsen records what's being viewed
and by how many.
The tag also pings major online publishers, who take
their data of who's viewing the ad, although the sites
never see it or know where it ran, and send the data to a
calibration engine.
Here, Nielsen fuses the who with the what and the how
many, creating a more detailed profile of the types of people
who viewed your ad, which enables us to deliver Online
Campaign Ratings, comparable to TV and other media,
featuring key metrics, such as reach, frequency, and GRP with
results that can be organized by various criteria, including
deeper data sets for marketing mix models, all in reports
that can be generated in just days.
Meaning you get direct in-flight feedback, better
campaign measurement, a common approach for online and TV,
and ultimately, improved marketing ROI, plus clearer
audience insights without compromising privacy.
Find out how Nielsen can deliver online campaign
ratings for you.
[MUSIC PLAYING]
[END VIDEO PLAYBACK]
ARI PAPARO: So Online Campaign Ratings, often people say,
well, I can already measure everything online.
I can measure through my gross volume on impressions, or I
can get a report saying roughly
what audience I reached.
So this is a change in scale for the ability to measure
audiences that's really important.
Because most of what's available currently is the
ability to measure 30 days after a campaign runs, with a
minimum of a very large campaign size, with a limited
ability to slice and dice the data.
So the three really distinguishing characteristics
of this program, I think, take it from being a nice to have
to a real important part of the brand ecosystem are
accuracy, timeliness, and the scalability of the program.
So I'll talk briefly about each of those three.
First, accuracy.
So panel-based measurement, as everyone knows, is very
accurate, but also has error rates.
And the error rates vary depending on
the size of the campaign.
Through our new methodology, we've already seen a 14%
improvement of accuracy of measurement, versus a
panel-only approach in an online measurement context.
So that makes a really big difference, when you're
looking at these numbers every day to make decisions, to
understand the value of media.
Secondly, timeliness.
Post-campaign reporting is used to understand how to plan
your media campaign.
It's not used in-flight to change the decisions you're
making and to optimize campaigns.
So being able to deliver audience data in-flight on a
daily basis is an order of magnitude change to what's
been available previously.
And finally, the scalability.
So sample size is always a problem with any type of
measurement.
And being able to do audience measurement on much smaller
campaigns that are using all the advanced techniques of
targeting and other mechanisms that are online is a
requirement to be able to count on online audience
measurements.
So the system we've developed works on dramatically smaller
sample sizes than any other system that's entirely
panel-based.
So with that, I have one scary chart.
My general rule of thumb is that if a chart's really
complex, you should put a headline
that says, it's simple.
But I didn't have the room on the slide, so you can assume
it's simple.
But basically, it is simple.
So when an ad is run on a website, any website, there
are different ways to measure it.
So the top measurement, census measurement, is a pixel.
So Nielsen records the total volume of impressions that are
seen by a given ad.
That's straightforward.
And the bottom measurement is also straightforward.
As probably everyone knows, Nielsen has opt-in panelists
across TV and PC that allow us to measure the audience
of an online ad.
So that exists currently.
That is the backbone of this system.
But what's unique here is that we are also working with
third-party publishers, such as Facebook, to, in real-time,
make decisions about which audiences we're seeing on the
given impression.
So the ad is requesting a pixel from the Facebook.
Facebook is aggregating on a totally anonymous basis the
impressions by demographic group, by age or by gender.
And then it goes into this audience calibration engine
that takes the total aggregate from our publishers along with
the aggregates from our census measurement and the aggregates
from our panel and combines them into one report of
region, frequency by demographic group that allows
the advertiser to see at a very accurate level who is
being exposed to a given campaign.
So the calibration engine is a unique technology we've
developed that is all working on aggregate data, no
individual data, not even PII, no individual data at all.
And that information is then compiled into our online
campaign rating.
So I'm available for more questions after the panelists
and also after this entire session to talk about it in
more depth.
But the way it works is kind of a very important part to
why it's different than what's come before.
So with that, I'll stop talking about
the details and go.
And I'm happy to introduce Mike Murphy from Facebook.
Thank you.
[APPLAUSE]
MIKE MURPHY: It's an exciting time to be in the advertising
industry, as more and more marketers are looking to shift
from the information web to a social web.
That creates some really interesting opportunities.
As marketers are looking to reach people instead of just
eyeballs, the opportunity continues to get better and
better with regard to creating cross-platform integrated
opportunities.
Old analog metrics, like impressions and clicks, aren't
doing anybody a favor in the marketing community.
So with Nielsen and with this project, it's really important
that we come together and come up with something more--
you know, a simpler process that both planners, strategic
agencies, and media agencies can use.
Last year on this stage, we announced our first
relationship with Nielsen on a product called Brand Lift.
Since then, we've done well over 200 studies together
where we've seen significant changes in the way that we
measure performance, and are finding more and more ways for
marketers to do the same.
And we believe this is a great opportunity for us and for
publishers.
We look very forward to working with our agency
partners and also the brands to come up with something
meaningful here, and are excited to be a part of it.
Next up, Kate Sirkin from Starcom MediaVest to give a
perspective from both the client and the agency.
[APPLAUSE]
KATE SIRKIN: Good morning.
I'm speaking on behalf of Procter & Gamble as well as
Starcom MediaVest Group this morning.
And as you know, he made the right decision to stay at home
with his wife who is about to give birth.
As a mother and a wife, I know he made the right decision.
[LAUGHTER]
KATE SIRKIN: I have two children, and both were
delivered early, so happy to be here and
speaking on his behalf.
We had a call on Friday afternoon and we realized we
were both saying exactly the same things.
So I think it is as well in the interest of not
duplicating our talks today.
It's good for you guys as well.
So the challenges of digital measurement--
it seems so simple, doesn't it?
Digital is the ultimate measurable media that we all
have. Everything's tracked, right?
So I know none of us is so naive with this audience this
morning, but what does the digital ad buying-selling
process look like?
Why is it not so simple?
Why is it hard to measure?
Why is this Nielsen and Facebook announcement so
important today?
This is why.
We plan, buy, and post on different metrics in the
online world.
I want to acknowledge, first of all, the work that the
Counsel for Research Excellence has done.
This is a chart I've taken from a recent education
program that they ran.
The CRE is a Nielsen-funded organization working to
improve the quality of all audience measurement.
It's independent from Nielsen, but obviously we share the
results back with Nielsen, with the view to improve
audience measurements over time.
And the CRE recently presented this to a group of Nielsen
clients as part of its education role.
And I think the chart sums up the problem perfectly.
The primary currency for digital sales is the ad
impression, as we all know.
Plans are developed right now in all kinds of systems and
agencies and advertisers' workplaces, though we know
[UNINTELLIGIBLE]
Nielsen online do have penetration across most of
those places.
And the posts are a combination of impressions and
clicks and custom survey work and all kinds
of different things.
Audience measurement, in its traditional world, reach,
GRPs, and frequency, constitute such a secondary
part of today's online buying world that it's just hard to
make comparisons.
It's hard for advertisers to think in similar currencies
across all of their media buys.
And that's what this is looking to solve today.
And in case you thought the digital workflow that I just
showed was simple, here's a snapshot of the more
complicated one that, really, we all exist in.
And if you just look at all of those measurements, these
companies and products, all together, form the marketplace
that we buy and sell in.
And it's a bit like the Wild West. Times are evolving so
fast that probably most of this is out of
date already anyway.
Many firms with many, many different metrics, processes,
and algorithms, the data collection, data matching, ad
serving, and reporting.
And we know that many of them provide significant benefits
for the advertising brands that use them.
But it's really hard to know which one and how to
understand that, and how to speak about it in a language
that both CMOs and regular people, not digital people,
can understand.
And again, that's why I think this initiative this morning
is so important.
So what came out of the CRE's education process were the top
three things for the digital industry to do to help the
rest of the world understand what it was all about and how
it could help brands to increase the reach against
their target consumers.
And audience, measurement and accountability metrics were
the top three, as silly as that might seem.
What we're really trying to understand is very simple,
basic questions--
who's seeing the ad, how many times, and where?
But we need to understand that across the digital spectrum.
We need to understand it across the hybrid measurement
solutions that Nielsen has.
We need to understand it against the behavioral targets
that we all are very excited about using, and across
long-tail sites.
We need to understand them, though, at all stages of the
planning and buying process.
So what this initiative can do for the industry today-- and I
think it's incredibly positive that it's happening now-- it
will provide a way forward for the online industry, a way to
better compete for money against television, although
from an agency and advertiser's perspective,
that's not what I'm looking for.
I'm looking for a much more media-neutral way to allocate
budgets for clients.
It'll provide a much better understanding of actual
impression delivery amongst the target, the demographic
targets and the different kinds of targets that we need
to look at.
It'll be delivered faster, so that campaigns can be adjusted
in real time.
It can be used to supplement the information that we put
into market mix modeling.
Anybody that's been doing that for digital campaigns just
knows how awful that process is today, and how much money
the online industry is losing from that bad methodology that
we're all using.
It'll provide much better understanding of the actual
duplication between television and online campaigns, and
online within publisher and by channel.
And it'll provide a much better understanding of the
most efficient way to build reach
against our target consumers.
And that's really from our side, what all we're looking
to do is quite simple, all of which is good for the
advertising buying, planning industry.
I think it's about time that the online world be given the
tools to be able to compete effectively with the other
media within today's marketplace, and the same for
the advertisers, to allocate efficiently between those
different media and [UNINTELLIGIBLE] at the end.
So I think we're going to go back to the
group and do Q&A now.
[APPLAUSE]
STEVE HASKER: So just as a reminder, I'm going to ask the
group a couple questions, and then we'll open up to any
questions that you may have for each and any of the
participants today.
So if I could start with Suzy.
Suzy, from a Verizon standpoint, as you think about
your marketing spend and the allocation,
what does this mean?
SUZY DEERING: Needless to say, I think about our marketing
spend quite often, given the fact that it's a
pretty hefty amount.
And I look at this as a great opportunity, because it's the
accountability that we need, it's the
transparency that we need.
And I also look at it as a right step.
So there's always steps in these types of programs,
because as we look at it, I don't look at it as one silo
that it's answering for online.
It's actually helping us across the board.
And any of that education is always helpful, given the fact
that we have a tough role in an advertiser's side to ensure
great little buzzwords like ROI and overall measurement,
and so forth.
So I think that this is definitely,
again, it's a step.
It's not the sole answer, but it's a step
in the right direction.
STEVE HASKER: Right.
And Joel, from a Moxie perspective, will this hurt?
Will it help?
Will it hinder?
What's it going to mean for you guys?
JOEL LUNENFELD: From an agency perspective in general, this
is a big step forward, as Suzy was saying.
I think, there were earlier, three points about this that
got me thinking.
The first was that this is a step forward, and as the rest
of the media world becomes only more digital and TV
becomes only more IP-based and our magazines become only more
IP-based, it's good to establish a base of
measurement foundation now.
We've had the GRP for many, many years, and I look at this
as a refresher and a more accurate way.
The second thing that's really important is, especially in
the digital world is, this is an important measurement but
not the only measurement.
So for digital marketers, yeah, this isn't going to
replace any of the action metrics or trackability that
you want in the space, but it is going to definitely give us
a good denominator to look at media across the board.
And the third point is really, the digital hippie in me, when
I first heard this, got a little nervous hearing about
shoving online to GRP to be honest. And I thought about it
for a while, and I got really excited to see that we are
actually coming out with a standard.
And it's important to have a unit of measurement.
And I still measure my car power in horsepower.
And it's important to have a standard.
And this is really going to only help us allocate
intelligently moving forward.
STEVE HASKER: Just to talk to Mike, as Joel talks about the
allocation, how does Facebook think about working with some
of the other types of media, so working with--
whether it be television networks or other sort of more
traditional forms of media?
MIKE MURPHY: Yeah, I think this week, there's a lot of
buzz in all of the press right now about digital being a
direct-response vehicle and TV being a branding vehicle.
I don't think that you'll hear much from Facebook as far as
trying to convince marketers we're taking money out of TV
and moving into digital.
A lot of marketers are already doing that.
What we want to work with marketers to do, is to make
their TV campaigns perform better.
So we believe that TV plus Facebook is greater than TV.
And the way to show that is to run those campaigns together,
make those campaigns more social, find ways to measure
every post. But to be able to plan that in advance and to
have metrics that you can plan against create a far better
opportunity.
But for us, it's not about trying to convince marketers
to take more money out of TV. At this point, it's we're
trying to work with marketers to make sure that all of their
different platforms perform better.
STEVE HASKER: Thanks, Mike.
Kate, you talked a bit about the allocation process.
Is that how you think about this platform?
KATE SIRKIN: I think so.
There's a couple of different ways to think about it.
One is, what's the strategic objective of the campaign?
How can you make that work harder?
How can you make it bigger?
There's various different metrics you need for that.
But at the simple start of any budget allocation process, you
need to know how many people does your channel reach, how
much does it cost, and how is it likely to work?
And if you don't have those things, it's
hard to move forward.
STEVE HASKER: Very nice.
KATE SIRKIN: And right now, digital doesn't have any
credible figures in that space.
STEVE HASKER: Let me ask one last question, then I'll open
it up to Q&A. Suzy, you talked about steps, and this being a
great step.
Where would you like it to go over time?
So what would be your perspective as you make hard
decisions around Verizon's marketing spend?
And what would be ideal for you?
And how could Nielsen or other folks help?
SUZY DEERING: Well, I think Joel alluded to this, which
is, the steps for me are, one-- and I have to laugh,
because I also look at it and go, kind of in the first
discussions, it's like, can you go to GRPs?
And you think, there's a lot that I still wish that I could
understand about my TV audience.
And I get a little nervous when I think about the fact
that, from an online audience perspective, I can understand
or know so many things that I don't want hold them hostage,
just because of the fact that I can get more learnings from
my online audience.
I also, from a staging or steps perspective, look at it
and think, you know, it is exactly what Joel said, which
is, we are in a digital space.
So given the fact that I don't consider vehicles or media
being consumed in just silos.
And looking at it from a cross-platform perspective,
we've been doing cross-platform promotions and
programs now for, oh gosh, at least the last two, possibly
even three years.
And I know that we've been talking about it
for the last five.
So now I feel like I have this step in the right direction to
say, how do we really deliver that across the cross-platform
kind of space, and actually have true measurement there,
and understand exactly who's consuming what and when?
Because I don't know about you, but the last time I sat
down to watch a TV program by myself, without any other
device around me, I don't know when that was, because it's so
far and so long ago.
So besides the fact that it's either my laptop there or now
with the whole tablet conversion, not to mention,
obviously, something near and dear to my
heart, which is wireless.
So that just in itself, just as we as advertisers haven't
pushed the marketplace hard enough.
And I think this gives us a chance to do that, because, on
the flipside of it, the media partners will sit there and
say, but I don't know how to talk to you about it.
There's not a common currency there.
STEVE HASKER: Joel, does that make the digital
hippie in you happy?
[LAUGHTER]
JOEL LUNENFELD: Yes, yes, the digital hippie is dancing.
[LAUGHTER]
[INTERPOSING VOICES]
STEVE HASKER: Well, unless there are any other comments
you guys would like to make, why don't we open up and see
whether anyone here has any questions for any one of you.
AUDIENCE: Yeah, [INAUDIBLE]
the new model measurements impact [INAUDIBLE].
STEVE HASKER: Yeah.
So I'll let Ari pile on here.
Suzy made the point, this is the first and
very important step.
This measures the age and gender reach and frequency in
a way that will be comparable to other types of media.
At Nielsen, our product roadmap is to then add on the
effectiveness measures.
And I suppose the good news, selfishly from our point of
view, is that we have those measures ready to go,
[UNINTELLIGIBLE] and the brand of product
that Mike talked about.
So we can relatively easily overlay those.
ARI PAPARO: I will jump onto that.
If you think about the entire funnel that we always talk
about, the top of the funnel is reach.
And there are a lot of tools to measure the rest of the
funnel in terms of engagement, things along
those lines, or surveys.
What this also allows is that, we now have a common
denominator for all those equations.
We've only been calculating the numerator and then
dividing by impressions, let's say.
Now we can say, what was the receptivity or the reaction to
a campaign, based on who we actually reached, which is
probably the most fundamental question.
STEVE HASKER: Yeah, in the middle here.
AUDIENCE: I got the mic, so I guess that [INAUDIBLE].
[INTERPOSING VOICES]
AUDIENCE: Yeah, I keep one in my pocket at all times.
[LAUGHTER]
AUDIENCE: It's David [UNINTELLIGIBLE].
I agree, this is an amazing first step, but it's a step.
And one of the critiques that we hear from our clients--
[SNEEZE]
AUDIENCE: That was Don.
[LAUGHTER]
AUDIENCE: I just want to be clear.
STEVE HASKER: We do that when we don't like the way the
question's going.
JOEL LUNENFELD: Can you rephrase it nicer?
AUDIENCE: So what we hear from our clients, though, in
context of this being first step, is that, while reach and
frequency measures across platforms are absolutely
critical, sample sizes need to come up in order for us to,
the industry, have a better handle on where consumers are
in a model that, unlike television, is so extensively
[UNINTELLIGIBLE].
I'd just like to throw it out to the panel about how the
sampling size, how the audience sizes need to
increase to make this really, I'll say the next step of this
first step.
ARI PAPARO: Let me dive in.
That's exactly why we built this product, because having a
panel-only solution has a sample size problem.
And as you indicated, the most cutting-edge techniques in
online are fragmenting the size of the buys rather than
going broad.
We think this solution actually solves that problem
by having a methodology that's not just hybrid, but combines
three sources of data--
the panel, which has a high degree of accuracy, the
census, which gives the volume, and then a group of
publishers who have very wide coverage in terms of the
number of people reached.
We think we can overcome that.
In our work to date, we found a 14% increase in quality in
terms of reduction in the error rates of panel-only
versus this new methodology.
And we think we can push that quite a bit further, so that
we can measure with a very high degree of accuracy even
the smallest campaigns.
AUDIENCE: Do you all buy that, by the way, on the panel?
JOEL LUNENFELD: Yeah, I mean--
KATE SIRKIN: Yes.
From my perspective, I'm excited, and I applaud Nielsen
for looking outside the box for a
solution to this problem.
I think in the traditional research methods, you would
never have got to something that would provide that sample
size to measure the kinds of things that we're buying.
SUZY DEERING: We asked the same questions, and got the
answer that we needed before we were ready
to sit on this stage.
JOEL LUNENFELD: Yeah, I think it's fair that this would be
the most accurate way compared to other ways of measuring
measurement.
And that's a good thing, because, as I come from social
web, we're signing into our phones now with things like
Facebook and Twitter, and we'll soon be signing into our
TVs with the same type of social log-in.
So we'll get that measurement and that scale.
MIKE MURPHY: Yeah, just from our perspective, we're excited
that, with our reach within the US, it's significant.
But we believe that other publisher partners are going
to deliver together with us significantly better reach in
the audience that we can measure more broadly.
STEVE HASKER: Yeah, we had a question here.
AUDIENCE: Carol Geller with [UNINTELLIGIBLE].
Ari, you opened up the conversation saying that the
center of this is going to be a Nielsen tag, which it seems
will be a tracking pixel.
How does this fit in with all the various pixels and tags
and other objects that are now inside of advertising assets?
We start to see a real plethora of these things, and
they're starting to actually impact the speed and the
weight of individual advertising assets.
So can you talk about that a little bit?
ARI PAPARO: Sure.
I think that's another issue that's dear to my heart, but
most people don't care.
So we're not solving that problem.
The proliferation of pixels in ads is an operational issue
that the online industry is facing.
And everyone cares about latency because it matters and
it impacts user performance.
I think that this is something that, as a whole, we still
need to work on in terms of intelligently only showing the
pixels that we need as opposed to showing 20 or
30 or however many.
I don't have much comment as to how to actually solve it,
because I think it's beyond the scope of
what we're doing here.
AUDIENCE: Are there any [INAUDIBLE] issues with model
adjustment with mobile versus [INAUDIBLE]?
STEVE HASKER: Today's discussion is in and around
the online space.
We're working *** the mobile exchange.
We have a question in the center here.
AUDIENCE: How do you [UNINTELLIGIBLE]
duplication?
So I read this week that there's 110 billion users
[UNINTELLIGIBLE] market, [INAUDIBLE]
population 25 million.
I use my iPhone touch.
I use my mobile.
I might use my laptop.
I might not always be logged into Facebook.
So how do you deal with that aspect, when you've got
tracking pixels, you've got retargeting, that spread
across five platforms for one user?
STEVE HASKER: I think it's a particularly bad problem in
Australia where we have multiple personalities.
[LAUGHTER]
STEVE HASKER: Throwing it to with Ari here.
ARI PAPARO: I get all the questions here.
Well, there's a couple of things.
First, the problem of the same user logging in multiple
places or having multiple devices, that's actually
handled in two ways.
Our online panelists are prompted to ask who is
actually using the computer at a given time.
And it is installed across computers
on their home computer.
And secondly, with our publisher log-ins, that is
consistent across devices potentially.
Now, going from online to mobile is a challenge that we
haven't tackled yet.
So this is online.
But we expect to continue to invest in that area, because
mobile's obviously of key importance to online
measurement moving forward.
AUDIENCE: I have a question about the demographics.
When you registered in Facebook, you put in demos.
I'm assuming in the Nielsen panel, you're putting demos.
But with all of these aggregate data forms from the
publishers, there aren't always demos.
So for this entire database, are some of the demos simply
further ascribed?
Which is what I'm guessing.
Is that correct?
ARI PAPARO: Me again.
No.
This is entirely based on demos, either from Nielsen
panelists or from registration data from a publisher partner.
Just to be clear, because I know there's been some
confusion on this point, this program works for
advertisement across any publisher partner, whether
they participate by providing data or not.
So the ad can run anywhere.
So there's no limitation where this is only applicable,
besides to have demographic data.
But the measurement is dependent upon
registration-based demographic data.
We're not using any ascription or inference models.
STEVE HASKER: We've got a question way up
there in the back.
I don't know if we can get a microphone up there.
AUDIENCE: Yeah.
I'm Mark [UNINTELLIGIBLE] from E.W. Scripps.
I just had a question.
How do you see this system working in tandem with ABC,
which handles and serves print products, particularly
newspapers and magazines?
Have you had any discussions with those organizations,
trying to blend in what you're trying to do here?
STEVE HASKER: Yeah, we've certainly had discussions with
the online publisher components of a number of the
major newspaper brands and players about, so you're
talking about how their online properties would
benefit from this.
We haven't yet extended that into the print realm and how
those two things will come together.
We haven't done that yet.
AUDIENCE: I have a question over here.
STEVE HASKER: Sorry, I'm having a hard time turning.
AUDIENCE: It's OK.
So you're tagging advertisements digitally now.
And does this impact your combination with monitor plus
tagging and TBS so that you could do across media, this
person was exposed to x amount of TV, x amount of online, and
then do some kind of analysis across that?
ARI PAPARO: It's certainly the vision, although it's not
directly tied to monitor plus tagging.
Our online and PC households have panel software on their
computers as well as their TV, so we have a single-source
panel that can go across the media.
So while the primary goal here is to allow us to provide
measurement that's exactly compatible with TV, a
secondary role is to actually do crossover measurement
between the two media.
STEVE HASKER: OK, does anyone have a question for someone
other than Ari?
[LAUGHTER]
STEVE HASKER: Yeah, just here.
AUDIENCE: For Verizon, how do you manage the moral hazard
working with companies that offer measurement or
something like that?
How do you make sure that they're being accountable in
an age when people are serving invisible ads and
that kind of thing?
SUZY DEERING: From a privacy standpoint, or from just a--
AUDIENCE: From an accounting standpoint.
SUZY DEERING: Can you take this?
JOEL LUNENFELD: Yeah, I mean, I think it speaks to a few
different topics.
We talked about tagging and the different views there, so
actively working with and pursuing universal
tagging-type places.
Having ad-serving information and using that as an answer
key when in doubt.
Also, ad verification services.
We work with a few different ones as well
on behalf of Verizon.
So there's a multitude of different types of auditing
for, again, for protection things like invisible ad and
cookie stuffing or anything like that.
We look at this more as a pure kind of reach and frequency
measurement answer key across the board.
There are definitely other measures that Verizon
[UNINTELLIGIBLE PHRASE]
I suppose.
STEVE HASKER: If I could just add to that.
I think one way to think about Nielsen's role, or for that
matter, any other third party, independent measurement
provider, is it's analogous to being an umpire or a referee
on the field.
And if in any way, shape, or form we favor one advertiser
over another or one publisher or media company, one media
over another, then the buying and selling community pretty
quickly come to the conclusion that our metrics aren't as
valuable as they thought when they started to make decisions
based on them.
So from our point of view, our model doesn't work if we favor
one over the other in any way.
And that's why, in launching a product like this, we think
hard about where the data was coming from and how we're
putting it together, and we think hard about how that will
potentially be tied with our other types of measurement.
JOEL LUNENFELD: We don't want this to be yet another
competing measurement.
We really want to look to this as one that the whole
community is helping to build.
STEVE HASKER: Yeah, we have a question over there.
AUDIENCE: How would you compare this solution to
advertisers who are more apt to go towards a cost-per-click
or a cost-per-acquisition model, assuming that's a
little bit of a lower risk upfront?
What would you say to those kind of advertisers?
JOEL LUNENFELD: It's what your goal is.
SUZY DEERING: Yeah, exactly.
JOEL LUNENFELD: If you have a ground-up goal in a budgeting,
where you'd say, this is the action I want and this is what
I'm willing to pay for it, and you can go there until you
have a point of diminishing returns.
Or you can start the other way, which is a reach goal.
I want this percent of the population, because it's a new
message or a new campaign.
SUZY DEERING: Yeah, I don't see it as [UNINTELLIGIBLE].
It's definitely not a replacement.
So it's definitely looking to say, it's just another piece
or another point from a measurement perspective.
I mean, when we go into many of our campaigns, we're
looking at campaigns we know on the back-end what we want,
and how we're going to-- you know, obviously the key
measurements that we're going to be wanting
to look at and evaluate.
So I look at this, again, as those steps
that we talked about.
This is a step towards that direction of saying, OK, how
do I make sure that, from a cross-platform perspective,
I'm really delivering on what the--
Am I reaching the folks that I said I was going to reach, and
I'm also delivering on a little bit richer, from an
information perspective, as far as, really, who are they
from an attitudes or a behavior demographic kind of
information?
STEVE HASKER: We have time for one last question.
Yeah, up here.
AUDIENCE: I would assume that each publisher has a different
way or a different process for registering.
So how would the different registration
processes impact the data?
I assume that some sites, perhaps people will give up
before they complete registration.
ARI PAPARO: I guess I'll take that.
So that big amorphous cloud that's called the calibration
engine is part of our secret sauce here, where we look at
each partner to research the reliability of the data and
the ways in which it's represented, the ways in which
it's available, and calibrate, essentially, against the
Nielsen panel to optimize the accuracy.
STEVE HASKER: OK, one last. Yeah.
AUDIENCE: [INAUDIBLE]
STEVE HASKER: I think, let me have a go at that.
And then maybe Kate, Suzy, and Joel can add on.
For the way we sort of set out about this, is it's initially
designed for branded advertisers.
It's designed for the display component of the campaign.
So I think to the point made earlier, whether it's looking
at more of a cost-per-click or a cost-per-action model, I
think it sits alongside those and addresses different
marketing needs.
And that's the way we sort of thought about it initially.
And obviously, the two need to come into the overall
marketing mix, and maybe be looked at.
KATE SIRKIN: Yeah, I'll add on to that.
As I mentioned before, when we start our budget allocation
process, we need to know the reach of each of the options
that we're looking at.
And we'd still push in the industry to think about how
that might work for search.
Some challenges that I think are that we don't track a lot
of the information.
But I have to believe that all the smart minds, certainly on
this panel and in the audience, there is a way to
get that data so that we can have more comparable metrics
across all the of the different options
that we've got today.
JOEL LUNENFELD: I won't to tell you that this is really
an argument against planned reach.
So you set out to hit this goal, and then you have a
means of saying you hit it or you didn't hit it, and holding
publishers and everyone accountable for it.
With search, it's definitely more of a-- if they were
asking for it at that moment in time,
it's going to be delivered.
And it's hard to tell Google that we want more people to
search for a certain keyword.
As much as we want that.
STEVE HASKER: OK, well, we've run out of time.
But thank you, everyone, for attending.
It's been [INAUDIBLE].
[APPLAUSE]