CBS chief research officer Dave Poltrack says a video-on-demand
system-where commercials could be replaced every three days, and the ads
could not be fast-forwarded through-is a long-term solution that could work for
both the networks and advertisers as an alternative to DVR viewing.
One of the most active and outspoken TV network research
execs, Poltrack also believes that word-of-mouth conversations by viewers on
specific TV shows can be of great importance to advertisers, and the jump in
DVR viewing of programming during the first three weeks of the new broadcast
season might just be a result of viewers wanting to sample as many shows as
possible early on.
Poltrack sat down to discuss these and other issues about
the industry, along with some specifics about CBS.
Through the first
three weeks of the new broadcast TV season, the amount of delayed viewing via
DVR rose dramatically and has drawn concern from advertisers who believe their
commercials are likely being fast-forwarded through in time-shifted mode. As a
broadcast network research exec, why do you think DVR viewing has risen as much
as 40% in delayed mode for some shows?
This is a concern and we are monitoring it. Right now, we are thinking that
all the delayed viewing could be because viewers are recording many shows
during the first few weeks to sample, and DVRs are overloaded. We are watching
to see if, as time goes by, and viewers get to watch these shows over a period
of a few weeks, whether they will settle into a more traditional pattern of
watching more shows live.
If this turns into
a longer-term or permanent pattern, even with C3, which gives advertisers
credit for commercials actually watched in delayed viewing over a three-day
period, commercial viewing could go down significantly. What is the long-term
solution to that?
A video-on-demand platform that does not allow fast-forwarding through
commercials and which would also allow commercial insertion for new advertisers
after three days would be a solution. Research has shown that viewers look more
favorably at a VOD system that is free but ad-supported where they can't
fast-forward through commercials, rather than a per-per-view-type model.
How would a VOD
system like this work for advertisers?
Right now, many of the network shows offered via VOD on cable systems have
ads in them, which cannot be skipped through. Many contain the same commercials
that appear in the live versions of the programs. Nielsen is able to capture
that commercial viewing data and so the networks can get full credit for that
viewership over a three-day period. After three days we can then remove the
original commercials and replace them with new ones and start the process over
again. This would be better for both the networks and advertisers than DVR
playback. Support for this model seems to be gaining momentum.
At what point can
it become a mainstream way to sell advertising that can be captured via delayed-viewing
Well, right now, the cable systems still aren't where they need to be for
smooth ad insertion capabilities. But in the long-term, there is greater
potential for advertisers with a system like this versus DVR playback viewing
announced a partnership with the Keller Fay Group, which measures word-of-mouth
conversations. How is that playing into your ad selling?
Word-of-mouth is a critical component of how consumers make their buying
decisions and as broad as most think social conversations are online, they make
up only 10% of all word-of-mouth conversations. Keller Fay has a system to
measure the other 90%. We gave them our audience segmentation information,
which is based on more than just the traditional age demographics used to
measure TV viewing. Merging this data together, we can now tell advertisers
about what consumers who are watching each of our shows are
interested in and what they're talking about. It offers insight into which
shows are generating conversations about ad categories or specific ads or
products. It can help advertisers focus their ad buys better. We have that data
for our returning shows and now we will do it for our new fall shows.
Some have said the
Nielsen sample for measuring TV viewing is too small to be representative of
the entire country and needs to be expanded. How do you feel about that?
Doubling the Nielsen sample for pure TV viewing measurement at this point
is not productive and it's not where Nielsen's investment should be. If you
double the number of the Nielsen household sample just to measure TV
viewership, it will not only be very expensive to do, but there will be no
improvement in the viewing data. The current Nielsen measurement system for TV
viewing is still the gold standard. It is very well controlled and very well
managed and very accurate. However, if the data is not used correctly it can
become inaccurate. If you want to see which network has the most viewers, it is
very accurate. Where it can lead to invalid conclusions is where advertisers
try to buy demo groups within cable that have tiny .2 or .3 ratings in local
markets and try to differentiate one network from another. In that case you do
need broader sample bases, yet some advertisers and agencies make buys this
So you are not one
of those who are critical of Nielsen?
Nielsen is taking the right approach toward a number of new initiatives.
The one thing the industry needs to do is to work with Nielsen to move its TV
measurement system outside of the home. Right now Nielsen measures TV viewing
just inside the home while more and more people are viewing TV from outside the
home on various devices.
What's your view on
the information that the growing number of cross-platform measurement services is
Most of these services are just giving us reach extension information. You
don't just want to know if you are expanding your TV viewing reach. You want to
know if there is an increase in responsiveness to the advertising
online. To attempt to have one measurement that combines ratings across
platforms is a good first step, but we eventually need more return on
investment information. Another question is if one platform comes up way better
than another, does it make sense to sell with combined ratings or just sell
them separately? If ratings data for linear TV and online viewing are going to
be combined just to reduce the CPMs of online buys, then it's not a good thing.
What's happening at
your Television City Research Center in Las Vegas these days?
When we first opened it, about 99% of what we did there was program testing
through focus groups. Now about two-thirds of what we do there is specific
program-related and the other one-third focuses on more tech-related
and distribution stuff like VOD, mobile phones, tablets and second-screen
experiences. And we are working more with both media agencies and some
advertisers individually on projects there.
canceled freshman series Made
in Jersey after just a few weeks and the
new Monday night sitcom Partners is
struggling within an otherwise successful comedy block. How did Partners do in its focus group testing and how long
will it stay on if it continues to struggle?
Well, I can't speak for the entertainment
division on what decision they will make concerning the show, but I can tell
you that Partners is a work in
progress and we are still tweaking it and hoping it can find its audience. In
our testing, we found Partners had a
more responsive younger audience than the other comedies on that night, which
skew a bit older. Unfortunately, it is going after the same audience that [NBC's]
The Voice is targeting in the time
period, so we are trying to be patient and find ways to get the sampling up. How I Met Your Mother started off slow
for us similar to Partners but we
promoted it via the Internet and we started to motivate the 18-34 audience to
watch. As you know, that series' audience grew very nicely. With Made in Jersey, it did very well
initially and feedback was good. But it started to slip and Friday night is a
night we have been winning for several years and we want to stay on top on the
night. We felt bringing back Undercover
Boss at 8 could do better than Made
in Jersey so we brought it back early to lead off the night.