Is March a 28-Day Survey? – Surprisingly, NO!
April 8, 2020
Next week, Nielsen will be releasing the March survey data (survey dates February 27-March 25). While many consider this a 28-day survey, in reality it isn’t; it’s really 28 one-day surveys. Typically this is not a significant issue, but as I will explain shortly, it might be with this survey.
While this is an oversimplification, every day is tabulated separately and then the 28 one-day surveys are basically averaged together to create a monthly. In other words, every day has equal value, regardless of that day’s in-tab level. Typically, this is not an issue, as long as the sample size for each individual day does not greatly fluctuate.
So why may this be an issue with the March survey?
During the March survey, COVID-19 altered the lifestyle and habits of just about every American. As our typical day has changed, it has impacted meter keepers’ behavior and thus the number of hours per day that they are carrying their meter. How has this impacted daily in-tab levels? As a much larger portion of our workforce started to work from home, were some meter keepers never undocking their meters? Panelists who do not carry their meter are not included in the daily in-tab.
Based on what we suspect is happening with meter keepers who are working from home, daily in-tab will go down, and listening from the meter keepers who are still meeting daily carrying requirements will be weighted up. Higher meter weights from a smaller daily sample will result in more fluctuation in the data.
Are we sure that significant fluctuations in daily in-tab levels will occur with the March survey? No. That’s why we use our proprietary tools to review Nielsen sample performance for our programming clients.
If you’d like more information on how we can help you determine the impact of COVID-19 on your March numbers, contact us here.
-Charlie Sislen, Partner