National Rating Swings are Mostly Real -- It’s Reality That’s Unstable

Nielsen for example is investing heavily in research to finalize an algorithm that will enable its integrated panel and big data measurements to be rock steady without introducing biases amp8211 the stability of the big data with the validity of the panel Itamp8217s a math problem After months of thinking by a lot of smart people we are closing in on the ultimate integration of the methods developed during 2022 which threw off an added benefit of increasing the accuracy of smart TV data Iamp8217m personally confident that the new general methods generated in 2022 will integrate for maximizing validity and stability at the same time in both national and localI had requested some views of big data and Nielsen panel data and saw some unexpected patterns in them which changed the way I looked at the situation of stabilityThe settopbox data also known as Return Path Data or RPD and the Nielsen national panel data seem to move closely together going up and down at the same time and almost to the identical degree This chart below is a typical example This analysis is thanks to Nielsenamp8217s Cermet Ream using RPD data after application of the Nielsen OnOff algorithm which corrects for the set top box onTV set off conditionHereamp8217s anotherHereamp8217s anotherHere are a few more So far Iamp8217ve looked at 49 programs across 46 networks and the pattern of agreement between panel and settopbox data in terms of the movements up and down in program ratings from telecast to telecast made me feel more confident than ever in the accuracy of both methods Generally when you see two signals moving in parallel they are validating each other there is a single truth being accurately measured by both of themHere are the summary statistics across all the 49 program series graphsThere were a total of 2885 rating change observations for each source across the 49 series Half were daily shows and half were once a week shows We wanted to get as many observations as possible so we included daily shows as half the sampleThe telecasttotelecast change was in the same direction for panel and STB 94 of the timeAnother 1 of the time the two were almost parallel but not going in exactly the same directionThe remaining 5 of the observations showed different directions of changeamp160Despite a huge sample size advantage STB changes were only slightly smaller on average than panel which is another sign that these are mostly real changes with only a small contribution from sampling errorsThe size of these changes from telecasttotelecast strike one immediately as high instability yet the fact that both methods move together suggest that they are both measuring the same underlying reality and that in fact amp8211 at least in todayamp8217s era of more program choice than the average vieweramp8217s mind can comprehend amp8211 the volatility from telecast to telecast in the real specificprogram audiences is far greater than most of us have assumedThe fact that national panel is almost as steady as STB aka RPD appears to indicate that almost all of the changes shown are real since as Nielsenamp8217s Chief Research Officer Pete Doe says amp8220With sample sizes in the tens of millions there is virtually no sampling erroramp8221 I agree TRA only had a few million homes but there was virtually no sampling errorThe level of program rating change flattens out a lot when looked at on a monthtomonth basis as Iamp8217d asked Nielsenamp8217s Peter Katsingris to track Across 5911 series the average month to month program rating change is 1995 in the panel and 2005 in the Impact data panel big dataThis analysis also verifies the average program rating change parity between panel Production and panel big data Impact If the percent change is close between panel 42000 homes and panel big data 30000000 homes then the change cannot be coming from sampling error because big data at that scale doesnamp8217t have very much of that That leaves the only possibility which is the obvious one that most of these audience changes are realNielsen is as committed to big data amp8211 both STB and smart TV and server data for publishers who choose to make full integrations with Nielsen amp8211 as they are to their panels because the combination is the right thing to do That the industry still has national topline program level stability from the panel alone is good to know but its stability wonamp8217t extend to demo breakouts and it doesnamp8217t prevent having some zero audiences those two areas are where big data will add most of their value in national In local Nielsen is equally committed to apply the new integration solution there as well and as soon as possible because the pain of smaller sample sizes in this sort of fragmented media environment is much greaterThis stability finding is most unusual What it suggests is that the ragged ups and downs of ratings assumed to be the fault of highquality area probability panels with their sample sizes in the tens of thousands or less doesnamp8217t mostly represent measurement instability at all it represents swings in real audiences in and out of a program from week to week because of the beckoning effect of immense choice What we might call reality instabilityKey TakeawaysThis articleamp8217s findings have no effect on the top priority given by Nielsen to achieving the maximum stability possible in its national and local currency services It remains Nielsenamp8217s top priority based on client feedbackNielsen national panel and set top box data have excellent consilience in mirroring the vast majority of telecasttotelecast rating change directionality and magnitude This should give the industry more confidence in its audience measurement practicesThe volatility in a television seriesamp8217 actual audiences from telecast to telecast has not been a subject given much attention before but itamp8217s probably a subject worth studying and affecting by means of continuous tunein advertising for all series telecasts not just premieresSelfpublished at MediaVillage through the wwwAvrioB2Bcom platformClick the social buttons to share this content with your friends and colleaguesThe opinions and points of view expressed in this content are exclusively the views of the author andor subjects and do not necessarily represent the views of MediaVillagecomMyersBizNet Inc management or associated writers

SUBSCRIBE FOR FREE MEDIAVILLAGE CONTENT | BECOME A THOUGHT LEADERSHIP CONTRIBUTOR

National Rating Swings are Mostly Real -- It’s Reality That’s Unstable

Author

Publish date

Channel

Archived in

Bill Harvey

FEMA and the Ad Council Encourage AANHPI Communities to Enhance Disaster Preparedness in New PSAs - Ad Council InSites

More Videos:

SUBSCRIBE FOR FREE MEDIAVILLAGE CONTENT | BECOME A THOUGHT LEADERSHIP CONTRIBUTOR

National Rating Swings are Mostly Real -- It’s Reality That’s Unstable

Enjoying This Commentary? There's More to Love

Bill Harvey

TAGS:

FEMA and the Ad Council Encourage AANHPI Communities to Enhance Disaster Preparedness in New PSAs - Ad Council InSites

Cookies