Sep 1, 2020

The Problem With Polling

With national political polls in recent weeks showing a strong advantage for Joe Biden in the Presidential race, it may seem that the contest is pretty much set at this point. However, the experience of 2016 taught us the limitations of polling as science and demonstrated that knowing the minds of the electorate is no easy task.

During the polling post-mortem that followed the 2016 vote, several key threads emerged that underlined how pollsters’ biases may have affected their prediction.

Meeting Voters Where They Are

“Many voters instinctively shy away from discussing their beliefs with strangers, making it difficult to capture truly representative data,” says Josh Gillon, CEO of What If Media Group. “One solution to these issues is to meet voters where they are: not asking a series of questions solely dedicated to campaign issues, but rather inserting campaign-related questions into a wider survey covering a range of topics, from the impact of current news events to product marketing-related issues.”

What If Media Group, an award-winning performance marketing company, takes exactly that approach to its political polling, with results that suggest that, just like in 2016, the 2020 Presidential race is significantly closer than national political polling might suggest—and that it may once again be decided by unforeseen enthusiasm for candidates, and the late choices of undecided voters.

“We looked at voter enthusiasm, issues, and policy as well as each campaign’s ground game and came to some interesting conclusions,” Gillon said.

To see the results of the polls download “Reality Check: Election 2020 – The True Story of Biden v. Trump”,

For questions related to the surveys or methodology, please email [email protected].

Skip to content