During these final few weeks, campaigns (and their consultants) are looking to finish strong. Ideally, the countless number of doors knocked, direct mail pieces sent, and digital ads viewed have drawn clear expectations for Election Day. The campaign’s message has resonated with voters, GOTV efforts are strong, and a clear strategic plan for the last three weeks is in place.
Unfortunately, this isn’t a typical election cycle, and it certainly isn’t ideal. While October surprises are common in competitive races, the onslaught of controversy surrounding presidential candidates Hillary Clinton and Donald Trump have thrust turmoil on down-ballot campaigns in even the safest red and blue districts.
With dwindling days until the general election, the key to victory will be asking the right question. The question you should be asking isn’t IF the latest Clinton email scandal or Trump’s deplorable comments toward women is affecting your down-ballot race, it’s by how much?
As a consultant, you can trust your gut and gamble on your campaign strategy, but wagering on voter behavior in the climate is a risky bet. The only way to accurately gauge where your campaign stands is by polling. Polling now and polling often.
Buyer Beware
As a research firm, we often here consultants complain about inaccurate poll results. Polls – like Hillary Clinton – have a trustworthiness problem. While there are many “research” firms that will sell your campaign a poll and promise accurate results, few at the mid-range price point will put in the time to back it up.
Here’s the bottom line, polling can be trusted IF it’s done scientifically. Getting 500 people to answer a few questions and then calling it a poll is easy, but it won’t yield accurate results. Scientific surveys must be timely, representative, and unbiased.
Protecting Your Investment
Before you spend hard-earned campaign dollars on a survey, ask your pollster these five questions to ensure a scientific approach:
1) What type of methodology will be used?
There are four major methods of collection: IVR (robocall), live, online, and hybrid (a mix between live and IVR).
For a cheap poll with inaccurate head-to-head results, go with IVR-only or online. Legal restraints prevent IVR polls from securing cell responses, and although online polling will be used heavily in the future, the lack of targeting accuracy holds this method in the R&D department for now.
Our advice: go with hybrid or full-live for best results. While these methods will cost more than IVR, the difference in sample quality easily overcomes the price. Plus, hybrid and full-live are the only methods that can produce an accurate sample of a cell phone-reliant population.
2) How do you avoid bias in script design?
It’s tempting to want to use the opportunity of contact with voters to frame your candidate in a negative light. However, if you do this, your survey results will be vanity metrics.
Understanding where the electorate stands and what strategy to use to win on November 8th requires an accurate, unbiased measure of where your candidate stands today. Any script must reflect that goal.
For example, instead of saying: “John Smith raised taxes. Does that make you more or less like to vote for him?”, ask: “If you knew a candidate raised taxes, would you be more or less likely to vote for him or her?”
3) How will the targeted universe be designed?
Designing a representative universe is more than just balancing age, gender, race, and party affiliation. Recent voter history and the use of landlines versus cells play a significant role as well.
Not all registered voters will vote. The accuracy of your survey will hinge on sampling a representative population of likely voters. Make sure your research partner is doing their due diligence in this step.
4) What is your data source?
Most pollsters partner with outside vendors to procure their voter data. The quality of this data is vital to the accuracy of your survey. If you try to bake a cake with 3-month old eggs, it won’t matter how great the recipe is, the cake will taste like garbage. The same goes for survey research.
While using a voter file from 2014 sounds like a great way to save money, it’s a fast track to garbage results, and any polling firm worth their salt won’t use it.
Ask your research partner where they get their data and if that data is an up-to-date, random sample of the target universe. If it’s not, no deal.
5) How accurate are your past results?
Quality research firms have a track record of success, and, trust me, they’ll be happy to show you. Don’t view this as the firm bragging, view it as your opportunity to vet how thorough they are.
If you’re planning on fielding a survey that will determine your campaign strategy for the last days of this election cycle, you should want seasoned veterans executing it. Make sure they have the track record to back up their claims.
Conclusion
The current tumultuous political climate surrounding the presidential candidates is bleeding down-ballot. The only way to ensure your strategy is effective and will lead to victory on November 8th is through polling – accurate polling. Before spending thousands of dollars on a poll, ensure the quality of the results by vetting your potential research partners.
At Cygnal, we’re proud of our accurate approach to polling and we have the track record to prove it. Don’t let the drama surrounding the race for the White House hurt your down-ballot race. Let us help you uncover exactly what it will take to win.