2016/10/03

Choosing Your Candidate: There’s an App for That

While most of people have already decided whom to vote for in the presidential election, there are still quite a few undecideds. A new app, called ‘Voter,’ promises to help bring more clarity about candidates based on their stances on issues and one’s own beliefs. The app developers say it also automates what could otherwise be a laborious process of determining candidates’ credentials and positions, and reduces human bias in the selection process.

Hunter Scarborough, the creator of the Voter app, says plans call for extending the app beyond the November presidential elections to candidates in races for Senate, House of Representatives and governors’ seats. Next, he plans to take it further to local elections such as for city councils, and may also go international. He discusses the app’s features on the Knowledge@Wharton show on Wharton Business Radio on SiriusXM channel 111. (Listen to the podcast at the top of this page.)
An edited transcript of the conversation appears below.
Knowledge@Wharton:  The idea of creating this type of an app — where did it get started?
Hunter Scarborough: I was working very long hours in my last job, and when an election rolled around, I just didn’t have time to do personal research. I wasn’t keen to vote on a sound bite from a news anchor or a sound bite from an uncle at the dinner table. I really wanted to know who these candidates were, and I felt the only way to do that was to dive in and spend a lot of time looking at their track records, looking at the conflicting media sources that we have out there, and comparing everything — which is a lot to expect of your average voter, right? I figured, in the 21st century, there had to be a way to leverage technology to solve that problem. So that’s how we got started on it.
Knowledge@Wharton: Explain what is entailed in the app, and how it can benefit people that may still be undecided as to whom they’re going to vote for.
Scarborough: We usually call Voter “Tinder for politics.” (Tinder is a social media app that matches and helps connect mutually interested users with each other.) We say that because essentially, you’re answering a handful of questions based on your political beliefs. We’re showing you which candidates support the same issues that you do, and also have a track record to back that up. The reason we [call it] Tinder specifically is because we use that same swiping interface. So if you’re familiar with that, where you’re swiping left and right on significant others, in Voter, you’re swiping left and right on major political issues, and then being matched to candidates based on those responses.
“… When we automate things, we take out more of that human bias potential.”
Knowledge@Wharton: Right now, is [your app] specifically for the two main candidates — Hillary Clinton and Donald Trump? Or are you starting to factor in Gary Johnson in the mix?
Scarborough: Gary Johnson and Jill Stein, the two prominent third-party candidates, are in the app as well. We don’t see it as our job to play favorites in that sense. So when you match with a presidential candidate, you’re going to see the full list. Let’s say that, hypothetically, you matched best with Gary Johnson, and then maybe Hillary, and then Trump, and then Jill, or something along those lines. We show you that full spectrum. So you can say, “Okay, maybe I match best with this third-party candidate, but I feel like they might not have the best chance of winning, so I’m going to go with my second-highest match, who is one of the top two candidates.” We just want to make sure that you have all the information to make whatever choice you want to make.
[What] really sets us apart, aside from ease of use and the intuitiveness of that Tinder interface, is the fact that we do look at these candidates’ entire voting history. If they have held a principled stance over time on an issue, you’re going to match a lot higher with them than someone that may have waffled or flip-flopped on that same issue.
Knowledge@Wharton: So a lot of research has gone into this particular political election, because you have to look historically as well, correct?
Scarborough: That’s correct. Now, there is a lot of research that goes into it, but … I wanted to leverage technology to make this happen. For two reasons, we try to automate as much of this process as possible. The first reason is that it’s a lot easier for us if it’s not just man-hours and labor going into doing all this research. But the second reason is probably more important, and that’s that when we automate things, we take out more of that human bias potential. Because the second a human hand has to touch the research or the information, there’s that potential that it’s going to be skewed one direction or another. Our objective is to avoid that in every way, shape, and form.
We have partners that allow us to automate pulling in [the candidates’] voting history, [and] pulling in speech analytics. So if they said the word “immigration” three times in a speech and the word “economy” 50 times in that same speech, we can infer from that that they care more about the economy than immigration. We try to keep everything numbers-based [and] fact-based to leverage the power of all the political data that’s out there.
Knowledge@Wharton: How do you come up with the categories that people can [consider] and decide, that would lead them to one candidate or the other?
Scarborough: That’s a combination of our analyst team deciding which issues [will] lead someone to the correct result for them as quickly as possible. The first set of questions that you ask are very broad [and] polarizing [on] overarching political issues. You only need to answer eight questions to see your first set of matches.
“I wasn’t keen to vote on a sound bite from a news anchor or a sound bite from an uncle at the dinner table.”
But if you want to dig deeper, there [are] nearly 100 questions in the app at this point. As you go through more and more questions, the questions get more nuanced, more specific, and more policy-specific. [The] true policy wonks out there can dive deep, whereas if you want to spend 30 seconds to a minute in the app, you can still get value out of it. You can still see where you weigh in on some of the bigger issues that are being talked about right now.
Knowledge@Wharton: Do you have questions that you are adding as the course of this campaign has gone along, depending on something that a candidate has said in a speech, or during one of the conventions?
Scarborough:  We have been consistently adding to that list. We started with about 40, and now we’re up to around a little over 93 questions. We’re constantly keeping the app up-to-date with what the scandal or issue or topic of the week is.
The app goes beyond [the] presidential [elections]. We’re going to show you not just your presidential match, but also your Senate match, your [matches in the] House of Representatives, your congressional district [and] your gubernatorial match, if there’s a governor’s seat open in your state. We have about 12 states now that are having gubernatorial elections. It’s going to show you candidates for all those races as well. We’re working on going even further down the chain, so you can vote up and down your ballot confidently.
Knowledge@Wharton: So people in various states will be able to pick the races that they need to worry about?
Scarborough: Yes…. Thanks to location services provided by the phone, it makes it easy for us to see which congressional district you’re in, and we can serve you elections based on that. Soon, in Los Angeles, we’ll be going hyper-local, drilling down into some of those elections that no one has much of an idea about, which would include city council, school board [and] mayoral [and other local elections, where] we see about a 10% voter turnout.
Knowledge@Wharton: How are the downloads going at this point?
Scarborough: We launched on iOS last fall, and we’re around 160,000 users. We launched on Android …  during the Democratic National Convention, [and] the Android user count [was] above 7,000 a week later….
“If [a candidate has] held a principled stance over time on an issue, you’re going to match a lot higher with them than someone that may have waffled or flip-flopped on that same issue.”
Knowledge@Wharton: Is the fact that being able to aggregate some of this information on an app because so many people have smartphones a natural tie-in?
Scarborough: That’s why we decided to go with an app. We considered going the website route. At the end of the day, though, I built Voter for myself. I was personally frustrated with how fragmented political information in the news is and how conflicting a lot of those sources are. I was thinking, “What would I use, personally? What platform would work for me?” I certainly wanted a fluid, fun-to-use app. The integrity of the information is there, but on the surface, it’s very easy. We’ll probably expand to web in the near future, but Android and iOS were our two key driving points.
Knowledge@Wharton: Once you get past the presidential election, and we get into more of the two-year cycle [at] the local level, you’re looking to try and get as many states and as many cities and as many different voting cycles on the app or on a website as possible.
Scarborough: That’s correct. We have a three-pronged approach past November. Right after November is going to be sink-or-swim time for us, to see how we do beyond benefiting from the immense buzz around this election. That three-pronged approach is going to be keeping our users informed and up-to-date on legislation that’s relevant to them.
A good example would be — about a year ago in California, we had an anti-ridesharing bill that was going through the California state legislature. That would have made it super difficult for Uber and for Lyft to operate in California, which impacts our target demographic in a huge way. Millions of people weren’t aware of it at all, especially in our target demographic. Our goal is to bridge that gap.
The second [aspect] is local politics and local races, which come around quite a bit more often. The third is going international. We’ve had a lot of interest from other democracies around the world in doing a white label or some other version of the app to apply to their races.
Knowledge@Wharton: Is there a core demographic for your app? [Many] people would have said maybe a few years ago that a majority of people with smartphones are younger. That isn’t necessarily the case anymore.
“Our target demographic is really Pokémon GO players … millennials and younger voters.”
Scarborough: The demographic has expanded a bit. Our target demographic is really Pokémon GO players. Ironically, that probably isn’t totally untrue, because our target demographic really is millennials and younger voters. That’s not to say that the app wouldn’t be useful or helpful for a voter of any age, and we’ve certainly seen that. Someone maybe under the age of 40 or 35 instantly understands [the Voter app]. But [with] the older demographic, sometimes we have to walk them through it a little bit.
Knowledge@Wharton: Using something that says, “It’s Tinder, but for politics,” [is] a great way to promote it.
Scarborough: Sean Rad, the CEO of Tinder, is on our board of advisors.  We didn’t start out with him on board. It was a direct result of us calling ourselves “Tinder for politics.” We got a lot of fantastic press when we launched, and some of that press made it across his desk, and he saw the line, “Tinder for politics,” and he ended up reaching out to us. A couple of days later, I was getting lunch with the CEO of Tinder. It was a little surreal. We ended up helping them to build a mini version of Voter inside of Tinder, called Swipe the Vote. Sean joined our board of advisors, so now we can officially say we are Tinder for politics.
Knowledge@Wharton: I would think that there may be potential for other partnerships with that company.
Scarborough: We’re certainly keeping the door open. We have an open dialogue with them. There are a couple of other companies in similar spaces that we’re talking to, and looking at partnering with prior to November.

Publicar un comentario