Your coffee order can tell a campaign who you’re going to vote for

National campaigns are headed towards Florida, so be careful what you buy.

The time, temperature, and sweetness of your Pumpkin Spice Latte speaks worlds about who you’ll vote for in this upcoming election.

Joe Caccitolo, a political campaign manager, is the dude who can explain how.

He’s a professor of psychology at the Chicago School of Professional Psychology who began working on political campaigns in Illinois in 2000. He got fed up with how unreliable public opinion polling was, so in 2010 he and his friends began started talking about how to improve the predictive behavior of polls.

Around the same time, the school he taught at was starting to talk about behavioral economics, a field that uses psychology to explain why we buy the things we buy. Caccitolo thought this could also be used to improve the predictability of polling.

For example, what does buying fair trade coffee say about your values? Does it mean you sympathize with farmers over big business?

We asked him to demystify the process and help us understand how pollsters come to conclusions about you based on your habits.

Interested in hearing more? Caccitolo and a panel of experts will be speaking at Venture Cafe at 5:30 p.m. on Thursday. You can RSVP here.

This interview has been lightly edited for clarity and length.

How is a campaign designed?

The first question you have to ask when designing a campaign is who is going to vote for you and the second is how are you going to win. Everyone isn’t a target for you. There are some people who won’t vote for you no matter what.

If you have a majority you’ve got your core group, but there are some people in the middle and you have to somehow identify what subsegments you’re going to appeal to. How you break that data apart gives you the idea of who you’re going to reach out to. Who is more likely to support that message rather than the other side?

What are some techniques behavioral economists use to actually get people to the polls? In Miami, it’s not so easy. Our voter turnout is pretty low.

We see a lot of it already. The two most effective and common techniques are around the idea of social pressure and the notion that how we see ourselves can be influenced by how we see others and how others see us. If you see yourself as thoughtful concerned citizens, well, those people vote. So that’s been one important factor and way of talking with folks that aren’t registered to vote.

The second is an idea of loss, which is based in behavioral economics. The framing for getting people to register and get out to vote uses that “loss” framework. Loss is a much different motivator. It’s a lot more about “Think about the bad things that could happen if…”

I just saw this while scrolling through facebook. A sponsored ad on my newsfeed showed celebrities who were all giving testimony to the idea that we need to vote and register to vote, and all of it is geared toward what could happen if you didn’t and they were using specific candidates names. It used both of those things, social pressure and people you would know, recognize, and respect, who are thoughtful, engaged and socially aware individuals. And if you want to be part of that, you take their message and heed it.

OK, knowing that campaigns are using these techniques, how can we become smarter voters or consumers?

It stems from identifying that it’s being done and recognizing that it’s always being done. We can identify it depending on how it’s framed: does it seem like a gain or a loss? Each of those has a different impact on how we process information, from a consumer perspective.

For example, with the advent of mobile technology, your coffee shop knows you like to buy an iced latte at 2:45 p.m., so maybe they’re gonna hit you up on your phone. That’s manipulation, in the idea that you’re getting what you want. Part of it is knowing it happens and being able to identify what’s really good for you and what’s not.

That doesn’t sound so bad. If someone knows what coffee I get then gives it to me, well that’s kind of convenient. What am I not thinking about here?

Everybody has free will for the most part. But just as your coffee purchases are predictable, campaigns have gotten wise. They look at your consumer profile, your Pandora playlist, and make predictions about what you’re going to respond to. The information is out there. They can reasonably reconstruct “you” from the data. Then, by talking to people with similar profiles, they can get an idea of where you are politically, within some range of error.

For example, someone who has firearms and subscribes to the NRA, those are just a couple of pieces of data you need to say they’re probably a conservative voter and a Trump supporter.

The question is, how many pieces of data do you need to make good predictions? How do you get that data? Campaigns now have access to that data and now have computing tech to crunch those numbers. When they’re knocking at your door, they know exactly who the consumer is.

What are some examples of some products that say something about people’s priorities or political leanings?

Products themselves tell us what people prioritize. So if someone was using fair trade coffee, they might be more [concerned with] the plight of farmers and [against] big companies who have driven them out of business. They might care about environmental issues, a perspective that leads them open to renewing our trade policy, might be against free trade. Those are the kinds of things you can take away from a product like that. That’s theoretical though.

Pandora Radio is also an interesting one, you can distill demographics and the music you listen to by demographic groups.

Magazines are also very indicative of what people’s priorities are. A typical one is gun owners, people who subscribe to gun magazines tend to vote more conservative.

There’s also the saying that Democrats drink more whiskey and Republicans drink more bourbon, but then you also have to think about the fact that there’s more bourbon in the South, and more whiskey in the city.

Another one is that Democrats have Subarus and Republicans have more Fords and Chevys, and the idea is that there’s a focus on American-made products in that party.

But people aren’t robots and are often unpredictable. How do you design campaigns around humanness?

Yeah, there are a lot of people who change their mind. Or they might not be feeling the same way now, or their voter profile doesn’t match their predicted profile. They are still likely to be persuadable. You set out your initial plan and say “Okay, these are folks I am going to talk to. These are the issues. These issues are going to get me X votes down the road.” But you have to be able to pivot. If you go into a campaign thinking “I‘ve got enough people,” you’re not going to win.

Are you saying politicians only advocate an issue so they can get a certain subset of votes without any real intention of seeing that through? Can you elaborate?

Well one of Obama’s key campaign speeches, one of his first policy initiatives, was to close Guantanamo Bay. I’m not suggesting he was not sincere, but certainly it didn’t happen. Also George Bush said he was going to tackle immigration in 2000. But Obama is a very interesting discussion [for other reasons]. If you look at the Iowa Caucus, Obama didn’t have enough votes to win, but he thought about who might be potential voters not registered yet and went after those students. They were able to register enough people for the Caucus based on the profile of the people that weren’t registered yet.

For the most part, people who fit your demographic profile purchase similar things at similar stores. You have outliers for one reason or another, but for the most part most folks are going to act between one and two standard deviations.

But, there’s still free will and you can change your mind.