The accepted norm in polling is now higher costs as fewer people are willing to pick up their phone – even their cell phone – and talk to a stranger on the other end of an unfamiliar number.
Even so,it is still very possible to get good samples of voters via live phone calls if you’re just looking at the demographics – for example: 18 females under the age of 35 in a congressional district.
What’s increasingly being ignored is non-response bias, because we can still fill a phone sample in larger districts. However, the type of person who answers that unfamiliar call looks different than the individual who ignores it. The same goes for other modes of research like online panels and email surveys.
So why would anyone in today’s fragmented communication landscape rely on a single mode of communication to conduct a voter survey, especially when so much is riding on the results?
What campaign only does TV ads or only invests in direct mail? None.
That’s why Cygnal spent a significant amount of time and resources last year developing a reliable multi-mode approach to polling. In several hundred actual surveys in the 2018 cycle, we figured out the challenges and opportunities each type of non-phone mode brings.
I’m not saying that these other modes of research will supplant political phone research – for now. What I am saying is that multi-mode communication produces more representative surveys, which is imperative for the 2020 election cycle.
Think about the “Trump Effect” (or scientifically known as social desirability bias) where individuals were unwilling to say on a phone call that they support Donald Trump. Do you know where they were willing to let their preference be known? Everywhere – from IVR to text to email to websites – except to another live human being.
Cygnal now incorporates voter ID-validated text message, email, and online panel responses alongside the phone component of our polls. It is incredibly time-consuming and difficult to conduct multi-mode research, because you’re basically fielding 3 to 4 different surveys at the same time to a unified same sample frame. This is why traditional pollsters who rely heavily on outsourced call centers to conduct their sampling and generate their reports are going to be many steps behind the imminent change in how polling is done.
We are investing in multi-mode research because our desire is to create more accurate surveys at a lower cost so campaigns can focus more resources on actually communicating with and persuading voters.
The investment includes the development of our own peer-to-peer text messaging (P2P SMS) platform, because existing tools aren’t built for extreme deliverability and conversational speed. Our full-time software development team has already built a complex quota generation tool that allows us to better control for responses by stratification and mode. They are working on tools to improve multi-mode fielding and the entire polling process end-to-end – from scripting to data management to reporting.
Cygnal is also transforming how polling data is consumed and understood. In the last thirty years, the biggest innovation in polling reports is moving from printed binders to the exact same data and formatting in a PDF instead.
Now, we provide an interactive dashboard that allows you to see the toplines visualized alongside the most important crosstabs. A “filter” function is included that allows you to instantly see the survey data with any combination of demographic and geographic filters applied.
For example, you need to understand where female voters in the Boomer generation within 3 congressional districts see the race. Just click, “Female”, “Boomers”, and the 3 congressional districts.
Voila – the topline and key crosstabs are now showing the filtered data.
Isn’t that so much easier than trying to figure it out yourself in the hundreds of pages of crosstabs or waiting on your pollster to reply to an email when you need to make a decision right now?
As you can see (and probably already felt), polling is in dire need of innovation. And Cygnal is driving that change.
What race do you have that you can’t afford to get wrong because of the old way of polling and viewing poll reports?