Our panel of experts including Rose Kapolczynski (Rose Kapolczynski Consulting), Brian Kind (Target Point), Ben Krompak (Tulchin Research), Tom Patras (EMC Research), and Ashlee Rich Stephenson (WPA Intelligence) took a closer look at recent trends and discussed what you can, and should be doing for your local level races, the influence big data analytics is having on traditional polling and why you shouldn’t throw polling to the wayside just yet during AAPC’s September 6^th webinar. In addition to the summary, login to watch the full webinar recording!

Brian’s Lessons:

1. Use large, generic, random samples in modeling so that all geographic areas are represented.
2. County lines can be really misleading. Try to get a good sample from around the state–the reaction of voters is very different in small counties vs. larger.
3. Bias is the great enemy of polling and modeling. You need to have an understanding of any potential bias before developing sound surveys.
4. Exit polls are heavily biased. Make sure you are looking at the bigger picture in terms of the entire campaign and not just on the day of the election.
5. Polling is so much more than just a horserace to predict a win. Most importantly it is about developing a narrative and story-telling.

About Brian: Brian Kind is Target Point’s National Director of Geo-Strategic Analytics and oversees the operations of its Office of National Geo-Strategic Analytics in Madison, Wisconsin. Brian is long time Wisconsin data strategist having served with the Republican Party of Wisconsin and the campaigns of Governor Scott Walker and U.S. Senator Ron Johnson. In addition he has provided support for a variety of state and local races.

Ben’s Lessons:

1. Declining poll participation rates affected the predictability of the 2016 Presidential Election tremendously. It is much harder to be accurate when there are genuinely fewer people responding to surveys.
2. You should monitor the polling environment on an ongoing basis and adjust on the fly to get an accurate representation.
3. Cellphones are no longer just for younger voters. They are for everyone. Results will not be representative if you only use landline surveying.
4. Exit polling is never completely accurate. It is more important to qualitatively understand why people voted in the way they did.
5. Focus groups get a general sense of how people talk about issues in their own words. It is important to do this kind of deep dive before developing your campaign strategy.

About Ben: Ben Krompak is the Vice President of Tulchin Research and has conducted hundreds of surveys and focus groups across the country on issues ranging from education to wildlife conservation to healthcare to criminal justice. Krompak was part of the team that helped put U.S. Senator Bernie Sanders in serious contention for the 2016 Democratic nomination for president, conducting focus groups and polling in 46 states.

Tom’s Lessons:

1. It is very hard to do good polling without a large budget. Remember this when creating a campaign budget.
2. The 2016 Presidential Election taught us to look at results without any preconceived notions.
3. Make sure to talk to your pollsters about the sample. Are you targeting the right voters? If you aren’t talking to the right people then it doesn’t matter what your survey looks like.
4. Internet-only polling truly depends on geography. We are seeing a lot more web surveys than in previous years and it’s still growing.
5. It is important to remember that voting is more of an emotional reaction than a practical one. It is key to consider this while also studying quantitative facts of a poll.

About Tom: Tom Patras has been bringing his laid-back, open-minded approach to EMC since joining the firm in 2005. With his diverse background in the private sector, government and on political campaigns, Tom brings first-hand experience of how his clients utilize research data. He is especially skilled at helping to develop clarity in their project goals and providing research-based insights along with recommendations that can be used in the real world.

Ashlee’s Lessons:

1. The 2016 Presidential Election was not as surprising for pollsters as everyone thought. It was the culmination of the continued magnification of polling research that has been coming for a few years.
2. Cellphones are a far more expensive method with which to poll, but it is important to make use of them to ensure you have the appropriate balance between landline and cellphone surveying.
3. Everyone always wants a poll to predict the outcome if the campaign were to end today. While this is important, it is imperative to use polling to find the pieces of your narrative that are most impactful for voters.
4. Make sure you are strategizing about how to reach voters whose support you’re confident you already have, as well as those that are on the fence.
5. You’ll want to ask pollsters this question before hiring, “how do we know we can trust your numbers?”

About Ashlee: Ashlee Rich Stephenson is a pollster and general consultant who has guided major election victories for her clients and has served as the General Consultant for Congressman Jason Lewis (MN-02) and as a General Consultant for Senator Joni Ernst’s (R-IA) historic win in 2014. Stephenson has worked for and consulted on behalf of several presidential campaigns, including serving as the Polling Director for Mayor Rudy Giuliani’s Presidential Campaign and moderating focus group research among independent women in swing states on behalf of Romney for President.

A huge thank you goes out to our speakers for sharing their expertise with the AAPC membership!