Polls Can Be So Wrong – And Your Success Depends on Them

December 15, 2016

Over the past few decades polling and survey techniques have become more scientific and we are led to believe more precise. However, the recent presidential election proved that this “exact science” can result in terribly wrong results. How else can you explain that just about every poll had predicted with confidence that Hillary Clinton would be the next president? All Americans now know that the outcome was very different than the polls had projected.

Why is this? There are some basics to polling. They are:

• Is your sample size large enough?
• Are you getting a representative cross-section of the population?
• Are you gathering the proper data?

In review of the 2016 presidential election, pollsters are mainly focused on points two and three. Some say they missed important voting blocks, while others say they did not gather the proper information (i.e. ask the right questions).

The point is simple. Those that have made a complicated science of polling blew it.

So what does this have to do with radio broadcasters in 2017?

Our success and failure relies on a single poll – the Nielsen Audio survey. As scientific as the PPM and Diary services are, they are not perfect. Every survey, broadcasters need to ask the same three questions.

• Is your sample size large enough?
• Are you getting a representative cross-section of the population?
• Are you gathering the proper data?

While tracking and reviewing overall sample size and proportionality are important, this does not assure us that the data is being properly gathered. More transparency on what is being collected would help build trust and restore faith in the survey reliability.

Let’s remember, the pollsters were confident that we were about to elect the first female president of the United States. Nobody questioned the quality of those polls until after November 8th. Is it time to begin to examine Nielsen to make sure they are doing the best for our industry?

Nielsen’s recent situation with 8% of the PPM meters losing connectivity combined with other recent SNAFUs (Situation Normal: All Fouled Up) makes these questions even more important. We at Research Director, Inc. continuously track Nielsen’s sample performance in both their diary and PPM markets, and we alert our clients when we see unfavorable results.

For more information, reach out to us here.

-Charlie Sislen, Partner

• Tweet