The Democrats’ Data Ecosystem

Episode Summary

How do political campaigns gather voter data online, and how are they using it to change hearts and minds? To find out, Tara spoke with Nell Thomas, the Chief Technology Officer of the Democratic National Committee.

Episode Notes

How do political campaigns gather voter data online, and how are they using it to change hearts and minds? To find out, Tara spoke with Nell Thomas, the Chief Technology Officer of the Democratic National Committee. FWIW Podcast is a product of Courier Newsroom + produced by Wonder Media Network.

Episode Transcription

Tara McGowan: Welcome to For What It's Worth, a new podcast that takes you into the digital spaces where the race for the White House could be won. I'm Tara McGowan and we're glad you're here. For this week's episode, I had the opportunity to sit down with someone who works behind the scenes every day to transform and modernize the way the democratic party operates and how they service state parties across the country. This is no small feat. That person is Nell Thomas, the Chief Technology Officer of the Democratic National Committee or what most of us know as the DNC. She's the first woman to hold the position. She's also the second CTO ever of the Democratic Party. So Nell, thanks so much for joining us.

Nell Thomas: Thank you. I'm so excited to be here.

Tara McGowan: So I just kind of want to start by giving our listeners a sense of you, your background, who you are. You're relatively new to this very big role, which I'm excited for our listeners to hear about. So yeah, tell us a little bit about, um, how you got to this role and what you did before.

Nell Thomas: Sure. So I'm a data analyst and data scientist by training. I've been working in the technology space for about, I don't know, 15 years. I got my first toe into political analytics in the 2016 campaign when I joined Hillary's headquarters up in Brooklyn and absolutely fell in love with the work and the people and the mission. And so I was really gearing up between 2016 and now to get back involved in the presidential cycle. So most recently I was actually at Facebook and I came to the DNC about three and a half months ago and was thrilled when the team over at the DNC reached out and asked me to get involved. I was very excited for the opportunity to work at scale across all campaigns and not just work on one but building infrastructure that's used by all the presidentials on the eventual nominee and all the down-ballot candidates.

Tara McGowan: So was the Hillary Clinton campaign in 2016 your first foray, professionally, into politics?

Nell Thomas: That's right. Yeah. And I had done a little bit of volunteering and I was involved in college and some political organizations, but my career experience before that had been in finance and tech startups. So I was definitely a whole new world. I was lucky to work with amazing colleagues who taught me so much about the quirky and crazy world of political strategy, both an education and a challenge, but also just, it was like finding a love I've been really excited to continue to learn more and keep in my understanding of the political space.

Tara McGowan: And what made you decide to transition in your career to work on a political campaign, which as somebody who's worked on political campaigns, that is a crazy idea because it is a startup that is essentially built to dissolve after a number of months. So what made you decide to take that leap?

Nell Thomas: So I had hired a couple data staffers from Obama 2012 to Etsy where I worked for five years and grew and built the analytics team and obviously the 2012 campaign had a lot of coverage of the digital and tech investments and the team. And so I had read about it and talked a ton to one of the analysts that is just a fabulous person, shout out to Samarth Bhaskar, and it kind of inspired me and made me intrigued. I find resource allocation problems fascinating and I campaign is one of the hardest and greatest resource allocation problems there is. So, I mean, obviously there's a mission there. I care deeply about progressive values. I care deeply a gap about having our country on the right track, but purely from an intellectual standpoint, it's also just really, really fascinating work. And I definitely experienced both sides of the mission and the intellectual challenges in '16 and now I'm doing it in a whole 'nother way at the DNC.

Tara McGowan: Yeah, absolutely. And I'm excited to dig in on that. First just kind of zooming out, 90,000 foot view, you're a data scientist by trade. You're now a chief technology officer of a national party. How in your words would you describe the relationship between data and technology?

Nell Thomas: So the state parties collect data that people provide when they register to vote and when they vote and the state parties all funnel up those lists of people and contact information they get at the state level up to the national party where you work and then you are overseeing all of the technology that is being used to take all this data in. You're organizing it, you're cleaning it, you're making sure that it's the correct information and that there's as much information about voters as possible.

Tara McGowan: And this is where I'm sure some of the listeners are perking up and being like, and to what end, what are you doing with this information? I just kind of want to ask you two questions. One, what is the value in the national party having this information in a clean and organized and applicable format for campaigns? And two, what are some of the new data streams that are now available given so much of our voter contact happens online as well as off.

Nell Thomas: Great. Um, yes. So as you mentioned, we at the DNC are responsible for building the infrastructure that collects all that data, that standardizes it, processes, augments it. We then build a data warehouse where it's made accessible to campaigns, state parties, and sister committees. So the DCCC, the DGA, et cetera.

Tara McGowan: All of the acronyms.

Nell Thomas: All the acronyms. That's really where the programs begin is at the campaign level or at the committee where it campaign has a candidate who is trying to reach voters and the campaign uses the data that we provide them to determine their field program, which includes: what doors to knock, what phones to call their paid media programs. So who they want to send mail to and who they want to try to reach on TV to the extent that you can kind of target that, who they wanna reach online, which demographics and maybe even which individual people using list matching. And so again, they're kind of doing this calculus between of we want to turn out voters that we know will support us and we want to convince voters who we think will turn out but may not support us. Let's use all the data we have about who people are to build as targeted program as possible. That will give us kind of the highest return.

Nell Thomas: The thing I'll say about data and analytics generally and then I will get your other question is that I think sometimes people talk about it like it's a magic bullet or that it's some special sauce that changes fundamentally what a campaign does and I really don't believe that. And I'm saying that as a data person, fundamentally campaigns are about talking to people and about getting people to hear a message. Analytics and data help you do it a little bit more efficiently and it's marginally more efficiently. So maybe you get an extra five calls out of your volunteer or maybe you're able, you're able to speak to, you know, one or two voters you wouldn't have otherwise in a rural area. That's not easily contactable, but these are marginal efficiency gains. They're really, really, really important. But it's not necessarily where the campaign is won. The campaign is won people talking to people about who they're going to vote for and why and what analytics now does is it just helps make that a little bit better.

Tara McGowan: Yeah. And I think it's a really popular misconception that data is the strategy. Data is not the strategy. There's a strategy and then data is a tool to help you implement a strategy efficiently.

Nell Thomas: Exactly.

Tara McGowan: And so something that we talk about a lot on this podcast is about how, you know, the evolution of the media and information ecosystem to be more distributed, to be, to happen more online on your social media news feeds, on your cell phone, et cetera. Obviously implied in that is there is a whole lot of new kinds of data and data breaches both very popular topics in the press today. And so I'm really curious, you know, I'm so excited about your background generally that you come from tech and digital startups and e-commerce and Facebook and, and so you really, really understand digital data as well, which is something that we talk about a lot is digital generally , another adjective that means nothing, but how, you know, taking a legacy organization like the Democratic National Committee and evolving it to really meet the demands of a digital ecosystem and marketplace. Um, and that is inclusive of the new data there. So I'd love to hear a little bit about that. What is digital data? How do you guys store it? How do you use it? How do you protect it?

Nell Thomas: So I'm going to start off by saying that I really believe that we're at a moment right now where there's a lot of new questions about the ethics of what data is used and how it's used. And it's something that we take incredibly seriously at the DNC. So we believe it's really important to think ahead to make sure that we're protecting our users, that we are not taking advantage of data, that we don't think that people are consenting to give in any way. You know, we really want to be mindful of it. So I'm just getting that out of the way because I think it's an important point that isn't talked about enough when we talk about like, Oh, the world of digital data.

Tara McGowan: Absolutely. Also the data that is publicly available that you're using is to encourage civic participation in democracy.

Nell Thomas: Absolutely.

Tara McGowan: Not to sell you toothbrushes, though that is also fine.

Nell Thomas: Well, it's funny, I think a lot of people don't even realize, I mean I'm sure you know this, that they're, the fact that they're registered to vote is publicly available information.

Tara McGowan: Right. But the more people who do realize that, we also know from much experimentation, can lead to more people voting.

Nell Thomas: Yes. In terms of digital data opportunities, I think that there are two people talk about a lot. So one is just reaching people online, which is how do you find someone on the internet, on their phone, or in an app. The other is getting information at what people are doing online and using that to feed models. I think personally that we as a kind of technological society have made a lot more progress in the first than the second, despite what people might say. So on the first one, you know, the Facebooks and the Googles of the world are getting much better about what we call list match. So you upload a list of people you want to find online and then via their fun systems, they're able to match those, to a user and send them a targeted ad. So in that case, one of the things that campaigns try to do and we try to do is, is have the best possible list to match against. So the more information you have about someone, if you have their cell phone number, if you have their email address, it increases your match rate. And so that's something that we're always thinking about is how we make sure that you're giving campaigns the tools they need to find people online the same way they can find them via mail, at their home.

Tara McGowan: Right, and in the interest of the voter, the consumer, it means that they're going to get more information from the DNC that's relevant to them, to their interest, to where they live, to their communities. And we think that's a positive thing.

Nell Thomas: Absolutely. I think sometimes people, you know, especially in the wake of Cambridge Analytica and I was at Facebook in the wake of Cambridge Analytica, so I spent a lot of time thinking about this. People think that campaigns or you know, Facebook is giving over who clicked on what ad or they're kind of an enabling, people do this crazy microtargeting based on psychographic information, you know, buzzword, buzzword, buzzword. I think a lot of that personally is smoke and mirrors and that it's, I mean, it's probably be like a way more interesting for me to be like, we're doing this magical, you know, algorithmic psychographic targeting. I think by and large a lot of that is a distraction from what really matters, which is having really good basic information about people and being able to talk to them with authentic messages. And so, you know, on the, on the future of like where will things going in terms of the data that people provide online. I'm sure there'll be more and more sophisticated ways of collecting that and utilizing it. So if you signed an online petition, is that then incorporated into the view of the person, you know?

Tara McGowan: And these challenges are not unique to the political landscape at all. This is a challenge in every industry right now and that sort of what's possible and what companies and parties are doing with data. There needs to be regulation to inform that it's a new landscape across the board.

Nell Thomas: 100 percent. And I think, or like the reason I usually talk about campaigns also like marketing campaigns is because I think that's the best compliment to how we think about this, which is a lot of the tactics used by campaigns are also used by Warby Parker or Clorox or you know, insert your favorite Instagram direct to consumer product, which is also trying to use a lot of these same tools and technologies to sell their products. The other thing I'm just gonna mention quickly, and I think you'll appreciate this, is I think a lot of times there's confusion between fundraising digital programs and what are called voter-oriented programs like turnout and persuasion. And it tends to be that fundraising is way more optimizable. You have an outcome online that you're trying to track much like an e-commerce company. You know, at Etsy, I worked on optimizing and experimenting on how to get someone to buy something on the platform. With fundraising, you have the same thing you have, you're trying to optimize and experiment and target around how to get someone to donate a dollar.

Tara McGowan: And it's black and white if it's successful. They either donated the dollar or they didn't or they clicked and decided not to donate and you learn that about them and you figure out why didn't they click.

Nell Thomas: And you know, oh people like this like to click and donate on these things. And that allows you to feed, you know, it also allows companies like Facebook and Google to target and optimize themselves as well because you have an outcome you can track really concretely.

Tara McGowan: And I also want to say because like part of the reason that targeting specific voters, that for instance like to give to candidates or parties online, get hit with more asks to give money is because there is an efficiency objective here. You can't spend money to reach everyone every single day. So you want to reach the people where that content or that call to action is going to be relevant and desired where they are going to actually engage.

Nell Thomas: Yeah, I mean, so fundraising, lots of interesting stuff happening actually. I think there's way more innovation on persuasion. 

Tara McGowan: Way more dollars being donated to candidates and parties than ever before, really exciting.

Nell Thomas: Yes and way more dollars being spent, as you guys so, dutifully track in the newsletter. On the persuasion side. The outcome is so much harder because it's an outcome that happens in someone's brain, right? Are they slightly more likely to like a candidate because they saw a video of them speaking about healthcare. And I think that on that front it is just, it's, we're still in a little bit of unknown territory about how we really optimize those programs and how we really leverage digital and especially in terms of the measurement piece, right? Because that part of it is really, really hard. And there's some fun and interesting experiments that can be run and I'm sure there are campaigns out there right now doing fantastic and interesting programs and experimentation.

Tara McGowan: ACRONYM may also being be engaging in some of that work.

Nell Thomas: Love it. Please do. Please do. But it just tends to be a harder area. And obviously the overall outcome you're optimizing for with any campaign is a binary outcome that happens once. Right? And so it's very hard. It's much, it's much harder problem than selling toothbrushes, to use that example again, because your kind of outcome is you're up for two years, nine months, three months, however long it is. And then, you know, in one day, yes you won or no you didn't. And you have no idea if had anything to do with a given ad you ran.

Tara McGowan: Changing hearts and minds is not as simple as selling toothbrushes, it turns out.

Nell Thomas: Absolutely. That's why it's a way more fun space to work in.

Tara McGowan: Exactly, and you know, arguably more meaningful, although we all need to brush our teeth. I'm really curious just kind of pivoting from how sort of campaigns think about collecting data and using it to how organizers who work on behalf of candidates they support and believe in use technology to be able to recruit more supporters and volunteers. And I think that's a really exciting space because it's evolving very quickly. So I'm, I'm really curious from your perspective what that landscape looks like today.

Nell Thomas: Yeah, I think there are a couple great tools that have emerged over the last four years and some of which were around in 2016 and early stages but have really matured, which is peer-to-peer texting. So organizers will be able to text instead of calling, which is obviously much faster and easier. And you're able to do it remotely so you don't have to be in a state necessarily. And we can talk about whether or not that's as effective, but it certainly allows you to hit scale. I think relational organizing tools as well. So for those of you that aren't familiar with the term, it's really about allowing people to understand who in their network already is politically engaged, could be politically engaged that's not yet and encouraging people to reach out to their friends and family to get them to register or to volunteer or to vote for a candidate.

Tara McGowan: Right, because you're more likely to take an action if somebody you trust or love or care about, asks you to do it, then a cold call from a campaign.

Nell Thomas: Absolutely. And it's probably more effective use of my time as opposed to going in and calling through a random group of 50 strangers to spend the time thinking about the 50 friends I have who could be more politically active than they are and who I probably know a lot about without having to rely on some database.

Tara McGowan: And innovation and technology in that space I think is so exciting. I think it could be game changing.

Nell Thomas: Yeah. No and I think we'll see a lot more experimentation around these new platforms happening in 2020, these presidential cycles and the midterms as well, but tend to be great innovation labs. We see such cool technologies coming out of them. I love the work that like Higher Ground Labs is doing and encouraging companies to really invest in the space and be creative about their solutions.

Tara McGowan: And bring innovation into the campaign space. And okay, last question and I cannot interview you and not ask this question 'cause you are the first ever woman CTO of the Democratic Party, which is really exciting. Also sad it's taken this long, but I'm glad it's you. Congratulations.

Nell Thomas: Thank you!

Tara McGowan: I've got to ask you, you know, what does it like to be a woman in the tech space? Because from what I've heard, you know, it's not that common and it's not often the best. But I'm very curious about your experience, certainly being the "minority" gender, still.

Nell Thomas: Yeah. Yeah. You know, it's, it's interesting cause I don't have the, you know, I'm an analyst, so I was thinking about do I have the experimental data and I don't, I only know what it's like to be a woman in this space. And so it's hard to know what it would be like if I was a man, but I would say I do ask myself that question way more often than I probably should. You know, I've been incredibly lucky in my career and I've had great opportunities and wonderful mentors. But I do think I question, how would this be different if I was a man or you know, do people perceive me this way because I'm a woman. I've been doing a lot of fundraising recently and some of the donors asked me, what are you planning to have children, or...

Tara McGowan: Love those questions, from my mother and the donors, all around.

Nell Thomas: I know, I would, I literally say, please just talk to my mom about this. But you know, I think there are those types of kind of moments that happen and I'm like, this is a little weird. But not necessarily enough to ever make me feel like I don't want to be doing the work but enough that it kind of causes a little bit of extra energy and mental space that I think probably it wouldn't have to spend otherwise.

Tara McGowan: And do you, do you feel as though more women are getting involved in data and technology roles and sectors?

Nell Thomas: Definitely. I think that there's a huge shift in the, especially kind of the generation coming out of college right now where you actually have data science programs for the first time and there's a real focus on obviously engineering diversity and I think also as data. So I do think that gender diversity is getting better. I think that there's still a huge problem with people of color not being underrepresented. And that's something that I, we try to be very mindful about in all of our hiring and recruiting. And I think it's something that we just need to continue to really spend the time on, you know, talking about, thinking about and acting on because I do think that that's a space that I don't see the change happening as quickly and I'm, I'm nervous about it.

Tara McGowan: Yeah. And so, sorry, last, last question. What would you say to people who are listening who work in tech or data, and I've never worked in politics, but might be interested in getting involved in this election in particular.

Nell Thomas: Do it. I mean, get involved. I would say start with, I mean this is, I'm going to say things you'll probably hear from other people start with your local races. See if there's something nearby.

Tara McGowan: Please, all politics is local!

Nell Thomas: Absolutely. And if you can walk into your state party's office and volunteer with them, help them fix their Google Sheets. Everyone has problems with Google sheets.

Tara McGowan: There's no job too small, people.

Nell Thomas: Help them update their Google G-Suite security settings. There are lots of little tasks that go a really long way. Be humble. Don't go in and say, I'm going to solve all your problems. But you give yourself the opportunity to learn from people who are doing the work and start in the field, start in state party offices, start with local campaigns. I think that's a great thing to do and go from there. And I think there's huge opportunities. Also donate to all of your favorite organizations.

Tara McGowan: And get all your friends to vote and get them registered and turnout.

Nell Thomas: That's right. Mostly importantly just like be engaged and pay attention.

Tara McGowan: Absolutely. Well, thank you so much Nell for joining, it's such a pleasure.

Nell Thomas: Thank you, this is awesome.

Tara McGowan: Have you back soon. That's all we have for this week. If you want to take a deeper dive into the state of digital politics, and if you're not already a subscriber to our weekly newsletter, also called, For What It's Worth, you can sign up at anotheracronym.org/fwiw.