Nielsen: Long-Term TV Viewing Changes Part Of Larger Trend
- by Wayne Friedman , April 9, 2021
Amid concerns of alleged undercounting of TV usage, Nielsen says while total TV usage has declined, this is part of an overall long-term trend. It is not related specifically to COVID-19 pandemic disruption of servicing its TV household panel.
“A shift has been happening for the last several years,” says Mainak Mazumdar, Chief Data And Research Officer at Nielsen, speaking with Television News Daily. “And what we see is that the shift has been accelerated over the last four months.”
For example, in the first quarter of 2021, an average 19.1% of persons 2+ were using TV across total day -- down from 20.8% in the first quarter of 2019. This data includes all TV use -- traditional, live, time-shifted and streaming.
The VAB, the TV advertising trade group, has slammed Nielsen's undercounting of TV usage, due to a 20% decline in the number of TV homes in the Nielsen panel. It blames the lack of field agents regularly servicing Nielsen in-home TV panel of around 40,000 homes.
advertisement
advertisement
But Mazumdar says field agents have been working remotely with its TV panel homes, as well as shifting to other new technology. “We love to be at the peak of our sample size, but we are not there yet,” he says. “Despite all our efforts, we see a decline in the sample size.”
Although Nielsen’s standard error level did witness a slight uptick, he adds: “Statistically speaking, there is no material difference in the measurement of our ratings, across demo rates, across geography.”
The VAB says the 20% drop so called “in-tab” TV homes -- usable TV data -- amounted to roughly 7,000 homes, lowering the Nielsen panel to around 29,000 in the first three months of this year.
Nielsen disputes this, claiming the current TV home number is higher. In addition, Mazumdar says, it has kept the sample size stable to that of the U.S. population.
Starting in October 2020 -- the traditional commencement of the TV season for many big prime-time networks -- Nielsen says there has been a dramatic shift, with a drop in traditional TV viewership.
Nielsen points to TV production disruptions -- 13% fewer new episodes on traditional TV in October compared with the same month in 2019. Also, there was aa 75% increase in repeat programming.
Nielsen says there were significant changes in total TV viewing consumption in the third quarter of 2020 versus the second quarter of that year. Total day TV use of persons 2+ declined to 18.5% (versus 18.9% in third-quarter 2019). In the second quarter, consumption was up 21% (versus 19% the year before).
But in both periods, usage on TV connected devices grew -- 5.6% in the second quarter (vs. 3.8% the year before) and 4.9 in third quarter (vs. 4.1% the year before).
Mazumdar also says there was more disruption and fragmentation last year of non-video media, including rising podcast usage. Streaming audio grew 39% between May 2020 and January 2021.
Given all the changes in media, Mazumdar says: “We feel pretty good the panel is measuring what it is supposed to.”
Wayne, there's a lot of smoke blowing about on this one. It's not a question of sample size as a smaller panel would probably not result in the recording of less viewing per panel member---indeed, it might actually raise it as heavy viewers could be better represented in a sample where lots of folks---inlcuding lighter than normal viewers--- suddenly opted out or ceased providing data. The real question is whether Nielsen---for whatever reason---or reasons---was counting people who should have been taken out of the panel due to moving to a new abode, sickness, the acquision of new TV sets which were not plugged in to the system, etc. etc. as if their data was acceptable. Since many of the possibilities could have lead to less recorded viewing than was normal for those involved or no viewing being recorded even though there might have been some, that's what the VAB is contending---I think.
As I keep asking, why isn't somebody who subscribes to the following alternative sources for TV set usage or viewing---Nielsen's local market reports, the MRI-Simmons studies, Comcast's set-top-box panel, Comscore's meter panel and TVision's "eye camera" panel---- compiling their findings for the periods in question to give us a comparison with Nielsen's national peoplemeter data. If the results are the same, that's fine for Nielsen but if not, that makes the VAB's case somewhat stronger---doesn't it?