Nielsen Unveils 'National' Out-Of-Home TV Service Based On 44 Markets

by Joe Mandese , October 24, 2016

Weeks after unveiling a plan to introduce out-of-home measurement of television viewing in local TV markets, Nielsen today announced a plan to also introduce a “national” out-of-home TV measurement service, but it did not indicate whether and how it would actually be nationally representative.

Nielsen originally announced that it would begin measuring out-of-home viewing of TV as part of a plan to roll out so-called “portable people meters” -- handheld devices carried by panelists that to date have been used mainly to measure radio listening. Nielsen acquired that technology as part of its acquisition of Arbitron (now Nielsen Audio) in 2013, and has been licensing it to rival comScore as part of a consent decree related to that merger.

Nielsen said it was rolling out the PPMs to increase the sample size of its local ratings in 44 markets, and that it would also offer a measurement service of out-of-home viewing as part of that. Historically, agencies and brands have resisted attempts by television companies to make out-of-home viewing a formal part of their ratings currency, asserting that out-of-home audiences were always implied as a part of what it was buying, whether it was measured or not.

In today’s announcement, Nielsen said it was also rolling out a “national out-of-home” measurement service, although it states that viewing will be “based on data from over 75,000 PPM panelists across 44 local markets.” There are 210 markets in the U.S. TV advertising marketplace.

It said the sample will enable Nielsen to "project out-of-home viewing in over half of [the] U.S. population," but did not say whether that would be an extrapolation or whether it would use some other method -- such as modeling -- to make it representative of the national marketplace.

It’s not uncommon for Nielsen to mix local and national panels as part of its TV measurement samples. Its national TV ratings currently are comprised of a combination of some households participating in its local people meter panels, and separate “national” people meter homes.

media measurement, nielsen, out-of-home, ratings, tv

9 comments about "Nielsen Unveils 'National' Out-Of-Home TV Service Based On 44 Markets".

Check to receive email when comments are posted.

Ed Papazian from Media Dynamics, October 25, 2016 at 9:31 a.m.
Nielsen is not measuring OOH "viewing", Joe. All that the PPMs can tell you is whether a person wearing or carrying one of them is within electronic "earshot" of a screen transmiting program or commercial content. "Viewing" is assumed, not measured in the PPM system--which, I believe, will be a boon to ad sellers as it will artifically inflate OOH "viewing" projections---especially for ads.
Reply

Brent Lightfoot from iHeartMedia, October 25, 2016 at 11:26 a.m.

Will be intrested to learn how they will credit viewing - example if I'm dining at BW3 and 5 different games are on the 20+ screens in the resturant, but only one game has audio playing, am I creditited to the game I can hear, or the game I'm watching, or not watching as the case may be.

Tony Jarvis from Olympic Media Consultancy, October 25, 2016 at noon

Ed is quite correct as usual. Added to his concern for "within earshot" as a surrogate for TV measurement, we need to underline that not only does the PPM approach not measure "viewing" but that the current Nielsen network TV measurment technique whether at the Live; C3; or C7 levels also does not measure "Viewing" despite Media Post et al using the term. While I defer to Ed or Steve Sternberg I believe it actually measures "likely presence". As a long time supoprter of passive electronic measurment and the PPM, it clearly offers a piece of the OOH video/audio measurement puzzle but only a piece based on the reservations outlined - there are many more.
The perfunctory service that has been Nielsen's Spot TV measurement, which has also managed to be called "viewing" by the trade press, has been referred to frequently by many of the media research cognoscenti in many Media Post comments. As annouced recently Nielsen is also planning to use PPM (2017) to "improve" their measurement of local TV when diaries are finally eliminated in 2018. Similar issues?!

Ed Papazian from Media Dynamics Inc, October 25, 2016 at 12:23 p.m.

Brent, it's my understanding that in the case you described the embedded audio signal could only be "heard" by the PPM you are carrying or, more likely, wearing, if the content's audio is on. Then again, if the encoded audio signal can be "heard" by the PPM whether or not the "viewer" can hear it, I suppose that Nielsen might have something of a problem, though the ad sellers certainly wouldn't mind the system counted you as "watching" all five sets at the same time. That way everybody---except the advertiser---wins.

Look at it another way. If my assumption is correct about how the system "counts" its "viewers", you might be credited as "watching" the screen which had its audio turned on when you were actually watching another silent screen---and channel---or not watching at all.

John Grono from GAP Research, October 25, 2016 at 11:03 p.m.

Ed, you have valid concerns. The PPM and similar devices only credit a single source (generally the loudest or the most recent). Five stations can't 'win' the same minute. Key are the data validation rules - which would also include removing 'listening' due to long duration (e.g. 'listening/viewing' at work when truth be known you were dragged off into a meeting).

There are valid issues like sound-down but viewing, that (in part) counter-balance sound-up but not viewing (e.g. in a pub or club).

There are other valid concerns (especially with radio - and dependent on the metering device) of listening via headphones - this particularly relates to iPods and smartphones.

But the biggest issue I have seen is carry-rate. Whether it is a PPM or a smartphone app, once you don't wear or carry the device there is no way to go back and 'fill in the gaps'. Ironically, a diary allows this to happen! In essence the device, let's say left at home or on the office desk, will capture data which will generally be edited out as there has been no movement. The real listening as the person goes to work, or goes home etc. is lost. All data I have seen so far indicates that TSL/TSV for a diary > PPM > smartphone. Tony do you broadly concur?

Ed Papazian from Media Dynamics Inc, October 26, 2016 at 8:07 a.m.

Thanks, John. In addition to the more technical matters we are discussing, my main issue with the PPMs or, for that matter, any purely electronic measurement of "listening" or "viewing" is that device usage---the TV set being on; the radio receiver being on; the smartphone being on, etc. is not a measurement of audience---only potential audience. As a result, electronic indicators can vastly overstate the real audience and the overstatement is not a constant across platform types, content types, or even demographics. Attempts to get at this by adding more electronic indicators---like time spent----are not a solution. The only way we can get an initial fix on "listening" or "viewing" is to ask the panel member, via some sort of promt, to categorize himself/herself as a "listener" or "viewer" and even if this were somehow possible, it couldn't be applied to average minute findings, let alone to ad noting, no matter how precise the electronic device usage data seemed to be. The only practical solution would be to organize an ongoing commercial recall/sales motivation study that cut across all types of content and all kinds of platforms---clearly an impossible dream. In short, there are some things that simply cant be measured with any degree of precision or relevance by electronuic means. All you actually get are indications or "directions", not the kinds of numbers that the time buyers and sellers love to crunch, accepting the findings down to decimal points as a transactional "currency".

Joe Mandese from MediaPost Inc., October 26, 2016 at 8:25 a.m.

Good points about "viewing." In truth, I think it could be argued that Nielsen's national currency -- the button-pushing people meter -- doesn't measure viewing either. Just that respondents pushed their people meter buttons when the people meter's prompt came on. That methodology may have been sufficient when it was introduced in the "viewing" environment of 1987, but does it qualify as a proxy for viewing today? Also note that Nielsen's recent neuromarketing study of "second screen" distractions for the Council for Research Excellence found that 25% of people meter respondents did not push their people meter buttons because they were distracted by their second screens when the people meter prompts came on, albeit this was not a statistically significant or representative sample. It does say something though about what constitutes viewing in today's television environment. Watch for a MediaPost report today about TVision -- an MIT-incubated company that measures how people's eyes are focused on their TV screens -- getting a round of funding to scale its operations.

Tony Jarvis from Olympic Media Consultancy, October 26, 2016 at 11:44 a.m.

Joe: Thanks for embellishing my earlier point regarding the fact that Nielsen's national ratings plus its LPM HH ratings in 25 markets (as accredited by MRC) are also not "viewing" but as cautioned a measure best described as, "likely presence" or as Ed identified, "potential audience" which is still an OTS or gross impression. As America's leading media editor-in-chief you should also have noted that the "Council for Research Excellence" is funded exclusively by Nielsen and its membership, agenda and detailed research-on-research reports are not availbale for assessment by non-Nielsen clients! BTW: The percentage of people not pushing buttons in meter homes varies by demo and by time period.

John & Ed: I do broadly concur with your concerns regarding PPM specifically and passive electronic measurment generally. However passive electronic technolgies are evolving to reasonably get us to the next level of measurement: Likelihood-to-see; Eyes-On; contact; or exposure. However even these very sophisticated techniques will require integrated solutions to address the inadequcies of any one passive electronic mesurement technique. Based on the evidence regarding the horrendous weaknesses with recall, it shall not be part of that integrated approach. Yes, I said "shall not"not "should not"!
Sound like TVision may address the true measure of "viewing" for TV. OOH/DOOH measurement is already there with its Eyes-On ratings approach via GeoPath along with companies like Quividi, Glueck, etc. that have face recognition technology when a screen is viewed. The latter companies offer increased precision plus real time albeit only for those DOOH networks using the service on all their screens. So they do not conform to a syndicated comprehensive harmonized audience measurement of the entire media channel that would provide comparable "face towards" audiences/ratings.

John Grono from GAP Research, October 26, 2016 at 6:16 p.m.

Great discussion.

In fact you could say that all personal electronic measurement have an element that is recall based ... do they recall that they have to wear or carry the device!

I haven't written off recall yet, as it includes an element of cognition, which as Ed points out is vital.

One thing I am certain of though is that a single methodology will not cut it. I'm all for a hybrid solution where you use the best tool for the 'component' of the measurement.