Mobilize Your People Like Obama: Applying Lessons from the 2012 Campaign to Your Everyday Work

Introduction

JULIA TAYLOR KENNEDY: Welcome, everyone, to the first Carnegie New Leaders event of 2014.

It's always a really exciting time. For those of you who aren't Carnegie New Leaders, we look at different themes every year. This year the Carnegie New Leaders' themes are lining up with the Carnegie Council's Centennial theme, because this year the Carnegie Council is turning 100, which is pretty exciting.

There are six themes that the Carnegie Council is looking at. Tonight's event grew out of citizenship and difference, one of the themes. We're going to be talking a lot about citizenship and bridging differences, in a way.

Now let's get to it. I'm Julia Taylor Kennedy. I'll be moderating this evening. I'll be talking to David Osborne.

David and I have known each other for almost 15 years, to be honest. We went to college together. Something that I noticed about David at that time was that he could talk to his roommate, who was an engineer, pretty easily, he could talk to me about Pablo Neruda and Chile, and then he would go on the radio and be a co‑host on a show about arts and culture. So this is a guy with diverse interests, who is able to toggle back and forth and carry his weight in a lot of different conversations.

His life since college has reflected that. Instead of getting a job straight out of school or going to grad school, he decided he would just move to Barcelona for two years and wing it, which he did. Then he came back and worked for the Department of Education here in New York, became a program director at the Gates Foundation, and then, when he got a call to join the Obama campaign, did so and built one of the key tools for the Obama campaign called Dashboard and helped organize volunteers across the nation in key battleground states. Then he decided to enter the battleground and moved to Richmond and helped organize volunteers there, and turned around a really tough team.

In that process, he learned a lot about the intersection between data and technology, and connecting with people and bringing them together. That's a lot of what we're going to talk about tonight.

Now what he does is he helps different organizations implement the kinds of strategies that he learned on the Obama campaign. Again it's a really wide array of organizations doing all kinds of cool things, but having David help them do it with data.

I want to welcome David Osborne to Carnegie New Leaders. Welcome.

DAVID OSBORNE: Thanks, Julia.

Remarks

JULIA TAYLOR KENNEDY: Let's start with the Obama campaign. Why did you decide to join?

DAVID OSBORNE: I had also done some campaign work before, just after Spain. Moving from Barcelona to Iowa is quite a thing to do, speaking of schizophrenia and mental illness. I had done that and then kind of sworn it off and said I would only get involved in the campaign if it really was special and important. Election Day in 2008, I had volunteered. That was one of the most exciting days of my life. Some you may feel the same way. Having seen the issues that followed—economic recovery that didn't appear to be recovering—and throughout I just thought Obama was just amazing every time I turned on the TV. In spite of all the challenges, he just was steadfast and smart and level-headed. I said, if I ever had the opportunity to help, I would.

Lo and behold, I got a call from a friend who had helped run the technology group in 2008, who said, "We need someone to step up. We've got this new role called product manager. We need someone who can kind of figure out what it is on the fly, and you're good at figuring stuff out."

It turns out that's a blessing and a curse.

So I said, "Sure." Chicago's nice in the summer, as we know. We went to college there—no excuse. So I went out. That was May 2011.

JULIA TAYLOR KENNEDY: How did you start using data and why was data so important to this product that you managed?

DAVID OSBORNE: We can start talking about the technology. I think the biggest thing for me was that I was the—I don't want to say "dumbest person in the room," because that may have been true, but that's not the point I'm trying to make—I was the least informed. Bad combination. But I was around a lot of smart people, on the one hand, people who were engineers—up to that point, I thought they drove trains—and then also—

JULIA TAYLOR KENNEDY: Come on.

DAVID OSBORNE: I did not really know what they did. I had worked on some technology projects, but not a lot, so I was really swimming in a deeper end.

Then also these people who ran the Obama campaign, who had won in 2008. I like to joke that they had Obama on their "fave five." These were people who were running things.

So I was trying to traverse both of them. It's kind of where I found myself—I walked into an office that could have about 1,000 people, and there were 25, in Chicago, which was cool, because you could throw a ball really far and everything echoed, and it wasn't as hot as it got later.

JULIA TAYLOR KENNEDY: What kind of data did you have about voters that you were using?

DAVID OSBORNE: I think this is a good opportunity to actually take a step back and just talk about the data operation for the Obama campaign. We're going to talk about a few different experiences. When I was first formulating this, there are a lot of similarities and differences. I like the difference theme with citizenship. I identify with that.

This first thing we're going to talk about is moving big groups. Really, the Obama campaign was focused on moving the whole country. At a minimum, we had information on every single voter in the country. That information was largely derived from something called a voter file. This is a quick primer in how national elections work from the data side.

A voter file—every time you vote, it's actually a public record. Whom you vote for no one knows. But the fact that you voted—and that's including in a primary. Your registration is also public. Maybe you're not registered as a Democrat, but you voted in the Democratic primary. Gotcha. You voted at all. Gotcha. You can learn greater degrees, and that's a good example of knowing more or less about a person.

But if this were the whole country, we knew if you guys had voted, we knew if you had voted in primaries, your registration. So that was sort of our starting point in terms of what we knew data-wise on people.

The innovation of 2012 was not that we had lots of data, though. We had lots of data—2008 was really the first time we mapped. We had all this data, and we actually started appending to it. If you guys are familiar with basic data structures—Excel, etc.—imagine additional files coming on. Did you vote? Great. Did you vote in all four of the last elections? That means you're a four-of-four voter, which is as good as it sounds. Maybe you skipped the off-years. That's okay. But then you were still a presidential voter. So we were like, you're probably going to vote.

We added other data things. We knew if you had moved. We had actually gone out and asked you. You guys got all those calls—actually, in New York you didn't get many calls at all. Not a swing state. But especially in the most important battleground states, we would ask tons of questions. We would call people. We would email. We started learning all this information about people, and then a bunch of geniuses got together, people with Ph.D.s as well as a healthy number of professional poker players—I kid you not. Really, don't play poker with them—

JULIA TAYLOR KENNEDY: I've been told many times that I don't have a poker face, so I'm not going to try.

DAVID OSBORNE: I don't have a poker game. It could be worse.

They basically were able to build a model that identified the key characteristics that would describe your propensity to vote at all and whether or not you were supporting the president. Those are the two things we cared about.

So we used all this data, built this great model, and had this huge operation, doing so, that consisted of everything, like I said—a literal room (we called it a cave) full of geniuses, plus hundreds of thousands of volunteers across the country making phone calls, reaching out to people, and saying, "Hey, Julia, I know you're a four/four voter. I know you're registered as a Democrat. We asked you in 2008 if you liked the president. You said yes and you said you were going to vote for him." We would periodically ask you again, "Are you still going to vote for him? Are you still going to vote for him?" And you're, like, "Yeah, yeah. Stop asking me. Why do you keep asking me?"

So we had lots of information about people at the start. The big innovation in 2012 was that we were able to use that information across channels to have a really concerted effort to reach the people that mattered the most—in this case, swing voters in swing states.

JULIA TAYLOR KENNEDY: I'm going to stop you there, because we're about to do a little bit of a demonstration for you.

So, Leenda, I would like to invite you to the stage.

DAVID OSBORNE: Leenda and I met a while ago at the cocktail reception. So she has been prepped, but only a little bit.

What I just described, data-wise, was that across the country we had all this data. We knew all these things about people. So we knew that Leenda was a four-of-four voter, which meant she was very probable to vote. We had asked her lots of times, "Do you support the president?" And she was, like, "Yeah, of course." "Awesome." Those are the nice conversations.

The next thing we did was, we tried to convert people to be volunteers. We had a volunteer army. We were really proud of that. That's how we were able to speak to more people. One person can only make so many phone calls in a night, 100 to 200, depending on how quickly they dial and talk. But 10 people—do the math—can do a lot more. So we were trying to build our numbers.

I'm going to do a sample phone call that will try to illustrate some key concepts that we used. This is sort of the next piece of data. We had this huge data operation, but then we utilized some sort of proven tactics. Here we go. We're on the phone.

Ring, ring, ring, ring.

LEENDA: Hello?

DAVID OSBORNE: Hi. Is Leenda there, please?

LEENDA: Speaking.

DAVID OSBORNE: Hey, Leenda. This is David. I'm just calling. I'm a volunteer with the Obama campaign. I'm over here at headquarters in New Rochelle, over on Main Street. How are you doing tonight?

LEENDA: I'm doing well, thank you. How are you doing?

DAVID OSBORNE: I'm so good. I know you're really busy, but I know you've been a volunteer in the past and you're the kind of person that really steps up when things matter the most. We just need to win upstate New York for Obama, because it's nip-and-tuck up here. We need you. We know that you're the kind of person who really steps up when it matters the most. You've done that in the past. I want to thank you for that.

But I'm actually wondering if—we have two events coming up, where we're going to be making phone calls. I'm actually going to be there, too. Really excited. This coming Tuesday and Thursday at 5:00. We're going to go for a few hours.

I'm just wondering, which day would work for you, Tuesday or Thursday, to come and make phone calls?

LEENDA: Well, David, I'm going to have to check my schedule, but I'm sure I'll be available for one of those days.

DAVID OSBORNE: Which one are you thinking? Check your schedule. You've got your phone.

LEENDA: Let me go check my phone. Hold on.

DAVID OSBORNE: You could look at the Obama app, but don't do that now. So Thursday maybe?

LEENDA: Actually, Tuesday is better.

DAVID OSBORNE: Tuesday, awesome. So Tuesday we're going to see you here at about 5:00. I'm just wondering, are you coming right from work or are you going to come from somewhere else?

LEENDA: Actually, no. I'll be coming from work.

DAVID OSBORNE: Oh, from work, great. Do you take the train or do you drive? Do you take a bike?

LEENDA: I'm taking the good old mass transit.

DAVID OSBORNE: Mass transit. Me, too. I take the train all the time. And so you're going to go to the train. Are you going to make any stops on the way or are you just going to come right to—

LEENDA: Actually, I'll take the train to the bus and—

DAVID OSBORNE: Train to the bus, great. Look out for that transfer. It's going to be cold. Make sure to bundle up.

LEENDA: Okay. Are you going to be there, too?

DAVID OSBORNE: Oh, yeah, I'm going to be here. I do this all the time. I'm really committed.

So that's great. So we're going to see you next Tuesday, 5:00. Really excited. All our neighbors, we're all getting out to do this, because we care so much. It sounds like you're going to take the train to the bus and come on down.

LEENDA: That's what we do as New Yorkers.

DAVID OSBORNE: All right. Well, the president appreciates it. I appreciate it. Thank you very much.

LEENDA: Thank you.

DAVID OSBORNE: All right. Give her a round of applause. [Applause]

Thanks, Leenda.

JULIA TAYLOR KENNEDY: Thank you, Leenda.

You probably saw that these words were chosen carefully.

DAVID OSBORNE: Very carefully.

JULIA TAYLOR KENNEDY: Pretty much every one. So why don't you kind of dissect what you just did there.

DAVID OSBORNE: All right. As backdrop, I wasn't actually just ad libbing. I mean, I clearly was, but we spent a lot of time between 2008 and 2012 studying the most effective ways to move people on the phone. We did the same thing with email. We did all media. I spent more time when I was in Richmond, Virginia, leading this team doing things on the phone, so I got really involved in this. We found things that worked to move people.

As you know, I started by saying, "Hey, I'm also in New Rochelle. I live around the corner," or whatever. Personal connection is really important. If people feel like they are like you, they are more likely to be influenced by you. So that's the first thing.

The second thing was, I said, "Hey, you're the kind of person who volunteers. You volunteered before and I want to thank you. I hope you can do that again."

People don't like this term, but we called it typecasting. It turns out that when you set someone in a particular frame of mind and you actually frame their identity in a certain way, they respond; they act that way. I mean, there are limitations.

Julia, you're the kind of person that puts out fires. You've put out fires before. You put on a suit and you act like a—no, that's ludicrous. In this case it happened to be true. Leenda had volunteered for us before. How did I know that? Data. I had lots of data on her, which I know makes you kind of uncomfortable, but it's true.

So we had this data on the campaign. I know what kinds of activities they had done, and I brought that up. So that's the other thing.

The next thing—and this to me was the most surprising thing and, I would argue, the most effective—we called it "make-a-plan language." I said, "How are you going to get here?" You're, like, "I'm going to get there, don't worry about it." "Are you going to take the train? Are you going to take a bike, pogo stick? What are you going to do?" "Well, I'll take the train, and I'm actually transferring on the bus."

We would do this all the time for everyone. It became really seminal for us and critical for us when we were trying to get people out to vote. Voting—"It's happening all day; of course I'll go." And then what happens? Your kid gets sick. Your car breaks down. You go to the polling place and there's a huge line. There are lots of reasons not to vote, not the least of which is that it happens on a workday, and at bizarre hours. But that's a whole other diatribe.

So make-a-plan language: You actually get it in your mind. You walk yourself through it, which makes it much easier when you do it. It's like you have already done it before, because you have walked through that in your mind.

Those are some of the different tactics we use. Once again, they are all data-driven in that people actually did studies. The Analyst Institute was one of the foremost. We got trained by those folks. A bunch of other political scientists have done this kind of research.

The one I didn't mention, and it's also the most interesting, is the peer influence—all your neighbors are doing it. Everyone's doing it. In some cases they even did studies where they sent out lists: "Hey, these are all the neighbors who voted in the last election. Make sure you're part of that list in the future." It made some people very uncomfortable, but once again this is public record.

Those are some things we did.

JULIA TAYLOR KENNEDY: What if you were in a situation where you called someone, they said, "Well, I'm not really sure if I'm even voting for the president this year," and it seemed like you really needed to connect with them on a deeper level? How would you handle that?

DAVID OSBORNE: The first thing is, when we pulled a list, when we had a list of people we were calling, we knew more or less whom we were speaking with, because we had data on them. The truth is, most people are—and this was based on, did they say they were going to vote for us? Even if we hadn't talked to them in a while, we knew once again, we had this model that these super-bright guys and gals had pulled up, so we knew which way—because there were trends. Things shifted.

A great example of that is, we started to lose white women. That was a big thing on the campaign, especially in our battleground states. So we knew that when we were calling Julia in Virginia, you were in that demographic. Even if you said you were with us, you suddenly got moved to a new list. You were no longer on our Obama list to volunteer. You were on our persuasion list.

The persuasion script was pretty different. One of the things we added there was a story of self. We tried to talk more to your values. Basically, that was more about juxtaposing us as people with having values conversations that were similar to the president, which is a really effective thing to do because it tends not to digress to "you're wrong, you're bad, this is right," instead of "oh, I really appreciate that you came down here from New York and quit your job and you're working in Virginia. You think this is something that really matters, and the president's like that, too." "Oh, cool."

It's hard to be, like, "get out of my face" when people are just talking about their own—I mean, it's easy. It happens a lot. But it's a different kind of conversation. That persuasion script was pretty different than what we called the turnout script: We want our folks to turn out and vote; we want to persuade the other folks.

JULIA TAYLOR KENNEDY: Could you give me a quick example of what a persuasion script might be like?

DAVID OSBORNE: Sure. What's a good persuasion script? If I were calling someone, I already knew sort of the values they had, especially with someone we were kind of losing. In our scripts that we wrote—we had a physical thing that we wrote out—I would get into something like—I would sort of amend my story. Say we were talking about health care, and I knew this was something this person—especially if a woman—cared about.

I'd say, "Hey, my name's David. I just want to talk to you a little today about the president's position on health care. I want to let you know, my mother actually passed away from cancer. Thankfully, she had coverage. But something I feel very strongly about—and it happens that the president does, too, because he also lost his mother to cancer—is that being a woman should not be a preexisting condition. I know lots of people have feelings that the health care law could be this and that, but if it were for no other thing, this is the reason I support the health care and this is the reason I support the president.

"I just wanted to share my story. What are your feelings on this?"

And we would have a conversation. We would always start with your story of self.

It's harder if someone would be, like, "I don't believe you," or, "I don't care about that," or, "I actually want to argue the policy." It tends to get them out of that space of argument and into the space of, "Huh, wow. What's your story?" So you sort of level that.

JULIA TAYLOR KENNEDY: Right. And you immediately build a connection with someone when you're that authentic.

DAVID OSBORNE: Absolutely, yes.

JULIA TAYLOR KENNEDY: Thank you for sharing that.

I think this primes us all for our first exercise, which is to turn to a person next to you and pair up. I'm going to give you five minutes.

You have worksheets in front of you. Learning styles are different. If you want to take some time and think about how you might persuade someone in your community, in your line of work, whether you work for an NGO and you need to raise development dollars or you're trying to convince your sister to go to a party with you—whatever the scenario. You can choose.

DAVID OSBORNE: We hope the first scenario is harder than the second.

JULIA TAYLOR KENNEDY: For me, number two is pretty hard. It depends on the situation.

Anyway, if you can pair up and practice, at the end of five minutes, we're going to ask you to raise your hands if you're willing to share with the group. So be prepared for that.

[Break for simulated phone interviews]

JULIA TAYLOR KENNEDY: Who is the brave pair that wants to practice their phone call in front of the group?

This is Jerry and Michael. If you guys could share what context you're doing this in?

MICHAEL: Sure. We had actually chatted for about five minutes before this at the reception. He got out of me that I'm an opera singer, which is maybe not what you would expect to find at this thing. But that's what I do.

Given that background and the way it's trained to sort of—do you want to take it? He approached me with this, and it worked very well—for real.

JERRY: I felt like the first thing was to assess overlapping points of identification. From opera singer, I intuited that there's a level of exemplary performance and standards.

I am currently an advisor to a documentary that is exploring New York City's standardized testing to get into Bronx Science and Stuyvesant, and looking at very much the debate. I went to Science a millennium ago. We want to explore what it looks like right now, because minorities are now severely underrepresented.

So I asked Michael as to whether or not he had an opinion about meritocratic behavior and let him speak to find out the balance.

(To Michael) As an opera singer, I imagine that you care a lot about training. What do you think about meritocratic stuff? And you replied . . .

MICHAEL: I said that, on the one hand, in the field I'm in, there's nothing to be gained by pretending that a certain level of standard doesn't exist. You have to hit it. I believe in, basically, strongly meritocratic standards for admission to certain—whether it's training programs or whatever level, schools, programs—because when you get into the professional sphere, that's all that it is. Either somebody thinks you hit a standard or there are 300 other people waiting to also sing.

That being said, I don't at all think that ability to pay for training should influence your ability to meet those standards, just personally. So that's sort of how I started into it.

JERRY: Once we got to the point of figuring out that access was significant, that was my way in. I said, this film is about access. We are tracking diverse communities throughout the New York City area.

Here's where we are. We had a successful Kickstarter and we're in mid-production. The students that took the test in the fall will be receiving their results in February and March.

I'd like to know, is this of interest to you, to begin with? We're going to have events in Chicago, in New York. I don't know if you'll be touring, but maybe you can come around. And I can show you the trailer right now if you'd like.

MICHAEL: And I said yes.

JULIA TAYLOR KENNEDY: Let's go through the elements. There was definitely a story there. There was a connection based on shared interests and what you knew about Michael. You knew he was an opera singer. There was the meritocracy—

DAVID OSBORNE: I would actually pause. I think you did that as a documentary film person, which is great, because you are. If you couldn't do it that way, we would have a problem. I would challenge you to put on a different hat and do the same thing. The goal there was to get him to come. You did it from the perspective of, "I'm really curious about what you do."

That's not actually what we do on political campaigns. We make people feel that way, but we actually take a different approach.

What you did was what happened in the lobby. What you could do is, as soon as you figured that out, you would do something more like, "I also care about standards. I went to a really prestigious school. Now I'm kind of questioning that. Do you ever question things?"

A more direct bridge would be more typical of what we do on a campaign. Even in real life sometimes, to get to that, "Do you ever put your money where your mouth is," you tell your story and tell more about yourself. In a documentary—except if you're like Werner Herzog, who is really into it—you're pretty outside of it. In campaigns you're right in the middle of the conversation. There's no preamble. You're not, like, "Now, actually, what I'm doing." And that does happen in documentaries.

What would be something you could say to put yourself out there a little more?

JERRY: Well, I could just personalize it. I could say that when I went to Bronx Science, I had three or four weeks to take this test. That was my heads-up, and that was it. Now kids actually are studying for two years, and it privileges people that have money and have an inheritance culturally.

So what do you think about that? What do you think, in the context—

DAVID OSBORNE: Pause on "what you think about that." Save the "what do you think about it" for the filming.

You want him to have an emotional reaction to something. I would actually encourage you to retell your story in a way that gets in your stomach, a little away from your head and into your stomach. Students are studying for two years. I'm studying for two weeks. Analytically, that's an unfair balance. I can do that math.

JERRY: You are raising a good point. What I realized at the end was that here was a wonderful character, a student, and she actually bailed from taking the test a week in advance. It's about carrying the weight of being the first person in her community. To me, we need to tell her story, even though she's not taking the test, and why.

DAVID OSBORNE: That's awesome. That's a meaningful story.

I think you should talk about your own—there are these degree things. What you just told is a good story about someone else, which is great. That's better than where you started, which was very much in the head. If you can get closer to your stomach, then when you ask Michael to come—and you're going to make a concrete ask, "Hey, we're going to be in New York on February 24th. I'd love to see you there. Can I have your email?" That's what we call a hard ask.

"We're going to be in Chicago and New York sometime. It would be cool—maybe if you're on tour, you can maybe come." Right? "Hey, are you going to be here" —it was uncomfortable when I did it with Leenda. The more uncomfortable you can be, the better this stuff sells.

JERRY: It's a question of style. I go soft and then I go in for the kill.

JULIA TAYLOR KENNEDY: I know this about Jerry, yes.

DAVID OSBORNE: Julia can attest that you kill.

But what I'm telling you is what the data says. It's funny to be having a conversation like this. You are all coming with your your stats packages. But these conversations—data has said that working this way delivers results. Making a hard ask delivers results. Telling your story of self as it relates to other people and connecting with other people delivers results.

Typecasting—"I bet you're the kind of person who steps up for worthy causes. You know in the arts it's tough getting by. I'd love for you to be able to support our work as a patron. I bet you're the kind of person that does that. Am I wrong, Michael?"

Then make a plan: "You're going to give me your email. I'm going to email you. There's going to be an RSVP to log in."

By the way, if you haven't figured it out, I do this all the time. The only person it's not successful on is my brother, because he just knows it's coming.

"You're going to click the RSVP. I'm going to follow up with you. It's probably going to be a nice event, so you want to dress up. But I assume you have a tux. I'm sure you look great in a tux."

All these details—and it sounds silly, but this stuff works. And it works on voters because voters are people. So the application here is wide.

So I think all those elements—once again, you don't have to change who you are, but you have a goal, and you can totally sell that.

JULIA TAYLOR KENNEDY: Great. I actually think that's a really good segue to what we want to talk about next, which is team dynamics. Tell me the story of how you started working in Richmond and what your goals were when you got there.

DAVID OSBORNE: Sure. Richmond, Virginia—a fascinating place. I learned a lot about it.

I was working in Chicago, working on this software project that I can tell you guys about. But at the end of the election, it really comes down to turning out voters. All these ideas, all this data don't matter if people don't go to the polls at the right time, with their IDs in places like Virginia, and actually cast a ballot for your guy or gal. I had had experience doing this on a few other campaigns. I really wanted to just get back involved with people. I like people. I also like numbers, but I like people, too.

So I went down and I took over a team in Richmond, Virginia, that had been pretty troubled. They had had three previous managers in the past two months. There had been a lot of turnover. Some folks had quit. It was the kind of situation where they were so behind on their goals that they weren't really keeping track anymore, which on a campaign is like heresy. Especially in what we call field work, it's all about hitting your numbers, hitting your metrics.

It's a tough situation when you come in there. There are a number of things that are awry that you have to fix. One is that people don't have self confidence. It's hot in Richmond. There are some tough neighborhoods in Richmond, and while folks generally support the president, that isn't necessarily their top concern—high unemployment, things like that. If you're spending eight hours a day on a Saturday going through some tough neighborhoods, door to door, it's hot, doing this every day, not really seeing any change—at that time, the president's numbers were dropping. There was a real morale issue.

I actually used data to try to reengage folks. In addition to having what we call one-on-ones and just reconnecting with people and understanding their emotions and telling them that they seemed like the kind of people that would get up in the morning and go to bed late and kick butt all day—

JULIA TAYLOR KENNEDY: That's typecasting, by the way.

DAVID OSBORNE: That's typecasting.

I also had a really frank conversation with them about goals. I said, "We have to be delivering. I believe in measuring. Our leadership here—everyone across the country is doing the same program. What do you guys think the goals should be?"

We had a really open discussion. The goal was actually to engage them. I just wanted to get them engaged in the idea of "let's get some metrics we care about." The only thing worse than metrics you cannot hit and everyone knows you can't hit are metrics people don't believe matter.

So we settled on some things we agreed on: how many phone calls you made, how many doors you knocked on, and how many volunteers you recruited. So we're going to go out and get that.

I said, "All right, what do you guys think you can do? I'll go to bat for you. If the thing you think you can do is not what the expectation is, I'll say we need some time. We'll get up to that standard. But right now we just need to be successful."

So we had those. I said, "That's great. You can own these goals, and you can also own the method. You can own the strategy to get them. I'm here as a resource if you need it. I've done this a few times. But if you want to just go and get it, awesome. They're yours.

"I am going to hold you accountable to those, because I'm held accountable by my leadership. So every day we're going to look"—we were recording all this data in a web program—"we're going to look in the app and we're going to see what work you've done and if you're on track or off track. If you're on track, awesome. High fives. We're going to win the election. If you're off track, that's okay. The reason we're checking every day is so we can make an intervention. We should do something new. We don't want to do the same thing and not have it be successful every time. Maybe that means we adjust the goals, maybe that we adjust your methods. And that's what I'm here for."

That was the plan. It was interesting to watch people sort of turn around. I remember it was a funny first week. There were three toilets in the office. They all broke twice. And it wasn't a particularly good-smelling office to start. But we did better on our goals than we did on the toilets.

JULIA TAYLOR KENNEDY: You weren't measuring the toilets.

DAVID OSBORNE: We weren't measuring the toilets, thank goodness.

JULIA TAYLOR KENNEDY: You talked about how you set the measurements, which was driven by the team. How did you make sure you were measuring the right things?

DAVID OSBORNE: In something like a campaign, a lot of this is pretty thought-out for you. It's just a question of understanding the emphasis.

JULIA TAYLOR KENNEDY: Maybe we can step outside of it for a second, then. A lot of times teams don't know what to measure.

DAVID OSBORNE: I think there are inputs and there are outputs, and then there are results. You should measure as much of it as you can. I think the key to measurement, though, is to keep it simple—one or two, maybe three things. You do want to keep it simple.

And you want to understand that it takes time. You have to wrestle with it a little bit. What's really important here? Are we trying to measure if the toilets are fixed? I mean, it matters that the toilets work. Otherwise, we're not going to be able to achieve our goal of having lots of people come in and make phone calls.

But is that really the main thing we're going to measure? No. Ultimately, we have to win the election. Working back from that, what are the biggest levers that get us there?

The reality about goals is that it's not a step-by-step dictation of what you should do. Once again, I firmly believe that people need to own their strategies. It is not the job of management to own the strategies; it's to advise. And if people aren't performing, you need to get new people. You should help them. You should do all you can. But especially in a situation like we were in in the campaign that was really exigent—this is not for everyone. Sometimes I thought it wasn't for me.

So that's one thing. But if you've got the right people, they need to own their strategies. You should separate a little that goal conversation from the strategy conversation. It's when you have goals that you can both see if you're on track or off track, and if you're off track, you can make corrections. If you don't have that goal framework, I don't know how you don't get into a he-said/she-said kind of thing, where it's more based on feelings rather than objective performance.

JULIA TAYLOR KENNEDY: Say you have objective performance goals, like talking about the strategy or the process to meet those goals. We study ethics and business ethics all the time here at the Carnegie Council. There are so many examples of companies that have set goals and said, "Use whatever strategy you need to meet them." Those strategies could get you into legal trouble or not and still be unethical. How do you think about that?

DAVID OSBORNE: That is really tough. I think that's a little bit of a personal thing, but I think it's also organizational. When you're trying to elect a president, you tend to have some fairly lofty ambitions. I think that's the kind of situation where you want to bring out the best in people. Everyone is taking a pay cut. Everyone is working for nothing. When you say, "Why am I getting up this morning?" it's not for a dollar. It's not because someone else told you to do it. You have a lot of gut-check moments, where you're, like, "Yes, I'm doing this for the right reasons. I'm doing this because I want to see the country move in the right direction, and I think this matters."

JULIA TAYLOR KENNEDY: I think it's an interesting thing you just brought up, which is mission orientation. A lot of companies are talking about this now, that they need to move their people based on something outside of profitability of the company.

I talked to someone recently who used to work at Altria/Philip Morris during the tobacco wars. What he said was, "The way I kept my team together, because people were dealing with challenges right and left, was (a) we knew that we were great people, and so we had the culture of being colleagues, and (b) we knew we were seeking to be regulated and that other tobacco companies were not. So what we rallied around"—and he even said, false or not—"was that we were going to set a high bar for the industry, and if we weren't in existence, that high bar would not be set."

DAVID OSBORNE: I think that's interesting.

JULIA TAYLOR KENNEDY: I thought it was fascinating.

DAVID OSBORNE: Yes. My experience has always been that one big thing about leadership is that you look to your leadership to see how you should act, and you also look to your leadership to interpret motivation. I think in an organization how people execute things and the way they do things really matters.

The sort of ethical conundrums that we could have gotten into on the campaign would be: "Do you fudge your numbers to look better? Do you take shortcuts? Do you have a conversation with a voter or a prospective volunteer and then turn around and be, like, 'That person's crazy. I can't believe they said this'?"

That stuff comes out. In a campaign, we're working 14 hours a day, seven days a week. In that kind of pressure cooker, there's nothing to hide. If you talk to my team, they know me. I wasn't hiding anything from them.

So I do think that comes out and your true self comes out. I think there's a different conversation about how do you try to inject morality into an immoral situation. Then how do you live your life in what could either be values-neutral—how do you sort of make that honorable, and how do you do it the right way? That's sort of what I have learned.

But it's tough. There are opportunities on a thing like a campaign, which is very legalistic and also very challenging, where you can go astray. Oftentimes what I have found is that when the leadership was really strong, like it was on our campaign, it made it easy to do the right thing. It was rewarded and it was understood.

JULIA TAYLOR KENNEDY: We have another exercise for you that I would love to give you a chance to do. That is, based on what we have just talked about, about metrics—

DAVID OSBORNE: I just forgot one thing, before we think about metrics.

When we were originally talking, I talked about the Obama campaign, rooms full of Ph.D.s, masses of volunteers going out and collecting data. On my team, we were making our own data. We were obviously doing our work that was built on this foundation that was a very rich data set, but we made our own data. I think sometimes people miss the opportunities to make your own data, which can just be binary. It's about accountability and following up.

Some of the things we were reporting on were how many house parties you had, for example. For a while, that was a great lever to get more volunteers. The experts weren't telling us how many house parties we had. They did help us project how many we should have, which was interesting. But tracking that data—we made it up ourselves.

When it comes to the day-to-day nuts and bolts, that was a really big lesson to learn, that you can do a lot, sort of homegrown stuff.

JULIA TAYLOR KENNEDY: So the idea here is to help you come up with some of your own homegrown data for your team. Just think of a metric and then discuss with each other why it may or may not work.

[Break for metrics exercise]

JULIA TAYLOR KENNEDY: Mark, could you please stand up and share with the group.

MARK: Our goal was website traffic. I kind of played a backseat for a little bit and then told people that I do digital marketing. It was a little bit downhill from there.

We decided to set certain metrics for the company, which are just visits to the site, and then hold people accountable, depending on team. So if it's a development team, they need to be tagging the site a certain way and building it a certain way so that when Google is looking for whatever we want people searching for, we show up in the top 10, hit that first page of Google, and hold them accountable to that number of hits.

The PR team has to be out in the media driving links back to our site. So we hold them accountable for a certain number of articles.

By the way, I do all of this, so I'm one person that's split into three, and probably not doing any of them very well.

Then you talk to the marketing team and say, "Okay, there are all these channels you can market with, social media or general blogs or whatever it might be." Email marketing is something that came up. Can we drive traffic back to the site with all those? How do we trace all of those numbers back so that we know we're hitting our million visitors a month and selling whatever widget we want to sell?

DAVID OSBORNE: Thank you very much for your feedback.

It's funny to talk to someone who is in technology on this, because I feel like technology is in a totally different category than all the other stuff, in that a lot of this is readily quantifiable. It's almost like cheating, to be honest. It is. And I'm going to actually talk for a minute about my experience with software, which felt like—I don't want to say an enlightened state. But for those of you who don't work with software teams or really know about this, it's a totally different way to work, which I feel is somewhat evolved compared to the more hierarchical and staid and maybe traditional work I've done in places like the New York City Department of Education, the Obama campaign, larger organizations.

All that stuff you said is awesome and great. I think there's an interesting thing that happens when you try to translate some of that to the work we do. Whatever job you do, you probably can't measure through Google the effectiveness of your work. All those things are 100 percent industry standard to get you to ROI [return on investment], which you can actually calculate and you can actually assign value to your traffic and all this.

JULIA TAYLOR KENNEDY: So a gold star, I guess.

DAVID OSBORNE: Yes. Welcome to the top of the mountain, folks.

MARK: I'm waiting for the "but."

DAVID OSBORNE: No, there's no "but." This is great. Awesome job.

If someone else could share, someone who has a different experience—I think that's something that I'm going to talk about a little bit, about how that translates to the experience of software and how those lessons can be shared back and forth.

PARTICIPANT: We're going from the top of the mountain to the bottom of the mountain now. I think it will be helpful, actually, because it's much more qualitative over here.

DAVID OSBORNE: Cool.

PARTICIPANT: If I understood correctly—because we were talking about the work that these women I just met are doing, so this is very on-the-fly—they are actually measuring the ethical activities in companies that are not their own. They're going in and trying to engage people in a company or in companies all over the world to understand how they do their work.

When we talked about how you could measure that, there were the simple things, like, oh, we can count how many phone calls we have and we can count how many meetings we have. But ultimately the question of whether or not it would be—are they useful measures? Probably not, because you're really not talking about engagement in that sense. You're talking about something much more qualitative.

So how now would you measure the effect of a meeting, the effect of an email, and so on and so forth? Perhaps you can do some external measures looking at their ethical behavior, from an outsider perspective looking in. But that's very challenging, because you really need something dramatic to happen to really know what's happening in the day-to-day of an organization.

So I think it's a much more qualitative challenge in this group.

DAVID OSBORNE: I want to offer something that's interesting about both of these things. Both of what these two sterling, awesome groups pointed out were impersonal measures of success. You went in that direction.

There's another whole set of work that's around what does a person's success look like on a day-to-day, completing their job. You got into it a little. The marketing director has to do the marketing.

You have a company that is trying to come up with—and that happens in academia, those kinds of questions—not to diss academia. We learn a lot from them. How do we evaluate this at some other place?

What I'm interested in is, what does the end of a successful day look like for you at your company? Did you have a good day today or not a good day? If you're supposed to come up with these measures and you don't have them, I don't know if you had a good day or not.

But say you do something else. Say you contribute in some other way. I guess what I'm suggesting is that that's measurable, too. Maybe it's quantifiable. Maybe it's that every day I have to to make 20 phone calls, because, really, I'm trying to get more companies to sign on and participate in this effort. Maybe it's not. Maybe you're measured by hounding the scientists or the academics who are actually doing this work. Great. How many new measures did you come up with today, Jorge, the lead scientist, clearly, there?

So I think that's really interesting, if you are able to work backwards. I know I've done jobs where some of the jobs I have been least satisfied with, at the end of the day—I knew that if I had had an awesome day or not a great day, no one knew, because there was no rigorous measurement of what I was doing.

My favorite thing about working on the campaign was, every day I knew. Every day I knew if I was killing or not. Partly that's because all the people I was working with were super-high-quality folks, really strong folks, really smart. But the other thing is, we had set up a system that—my job was largely to make sure other people hit their marks, as a manager, but I also was contributing in my own way.

I think those kinds of comprehensive systems are really when organizations start to move, and you're using data in really smart, different ways.

I don't know if that's a good transition to the last piece, talking about my experience with the software team.

JULIA TAYLOR KENNEDY: Let's talk a little bit about it. You said, for software folks, the metrics are obvious, right?

DAVID OSBORNE: Yes.

JULIA TAYLOR KENNEDY: How do you build in the other piece, which is the qualitative piece?

DAVID OSBORNE: Once again, the back story. When I joined this group called Dashboard—and what we were building was—if you went to barakobama.com, and you were, like, "Oh, I need to be convinced," great, you had an experience and that was the barakobama.com experience.

If you said, like Leenda, "I want to get more involved. I love the president. I love the first lady. I love everything," great. You did not know it, but you were actually passing through to a separate web property that we in-house called Dashboard. It was branded the same way. You had a whole different set of experiences there. The goals there were to further engage you as a volunteer and get you actually doing the most important work, because we had an understanding that there were tiers of work—

JULIA TAYLOR KENNEDY: How did you know that people were like Leenda?

DAVID OSBORNE: There were a few things we could do. One is, if you came in, you registered on the site, so you became a registered user. The same way Google can track you when you're on their properties, you were on our property. You were registered. You were Julia—remember I said in 2008 we built the big data system? We knew you were Julia, you were a four-of-four voter, you were living in a battleground state, and you cared about the environment. We also knew your fundraising history. We knew that you had subscribed to these emails, if you had told us your Twitter. We knew all that stuff. We didn't do a lot about it.

In 2012, we did something about it. When you logged onto Dashboard, we knew all this about you, so we could actually encourage you to just pass through—you're a four-of-four voter. We're not going to send you a message to go vote. We want you to volunteer. We know you've already done this one volunteer activity. We want you to do more. We know you love Twitter. Why don't you tweet about it, to start, and then get to some bigger, meatier work? We know where you live. We use your zip code. We know. You're registered in the voter file.

There's actually a group in your neighborhood. There's an event coming up on Tuesday. Sound familiar?

JULIA TAYLOR KENNEDY: Make a plan.

DAVID OSBORNE: Yes. We didn't actually say, "How are you going to get there?" "I'm going to take a ferry, a yacht." Not a lot of yachts.

In any case, we could create those kinds of things.

But all that aside, for me the really challenging thing was—because I was really a newbie to this—how do I even understand what is a good idea or not a good idea? I have all these engineers who have all this experience. They had been in all these great organizations—one of the founders of Twitter, one of the first engineers on YouTube. These are impressive guys.

I don't know if their ideas are good or not. How am I supposed to evaluate those?

So I used some tools to do that. That was my first challenge. One thing I learned is, if it has worked before, it's a good idea to try that. The more similar and applicable the example, the more likely it will work—i.e., we're trying to find political technology. If you can't find that, just regular technology. If you can't find that, it's a new idea—cool, but it's less likely to work.

The second thing is consensus. If everyone thinks it's a good idea, it's probably a good idea.

The third thing is, do you have some data that it works? Which usually you can collect through doing some of the things you were talking about with your website. Are you driving traffic there? Their performance evaluation is great, but sometimes you can't get that until the idea is already launched.

What I really learned, though, is how the decisions are made within a tech group and how that work is tracked. It takes a lot to build a website. We used an agile methodology, which is one methodology for building software. Rather than starting with a long list of things we wanted to do and then doing them in that exact order almost like robots, we said, "Hey, we're going to revisit this list all the time. We can only get something done in one or two weeks, which we call a sprint. But we'll figure out how much we can do in that time and then we'll reconvene and see where we're at. We'll probably learn something. We have the idea of these things coming up, but we want to be able to exercise some judgment in picking that up or putting that down."

The "we" here is really important. I was not the boss of the engineers. They made this very clear to me—god love these folks—which was wonderful. They had a lead who was their engineering lead, and I was responsible for communicating, largely, and helping keep folks on track. But I had a separate job, and it was not management. It was very different than what I did in Richmond. I was not going to fire anybody, thank god, because that would have been a tough conversation.

Really, when we decided, it was very democratic. That meant we kept the group small. If the group is too big, you can't get a thing done. It meant you had to have some format. We did things like daily check-ins to see if things were on or off track. People had their work that was chunked out into what are called stories that had points. So you see, in a very soft way, you have built in some internal metrics.

How many points is the team going? We checked velocity week over week. How much work are you getting done?

Every day we had something called a stand-up meeting. Has anyone else participated in a stand-up meeting? What's the best thing about a stand-up meeting?

JULIA TAYLOR KENNEDY: You don't sit.

DAVID OSBORNE: Well, that's the truest thing about a stand-up meeting.

You want to move it faster. If we had all been standing, you guys would not be asleep and you would have been, like, "Okay, let's get to the point." So it helps move people along.

One of the key principles of agile methodology is a useless meeting is useless. Everything has a purpose. It's in and out. You try to be a little bit more directed about that kind of stuff.

Then at the end of the week, we have a retrospective. We see what went well, what didn't, how many points we got done, how many we left on the table. Are we on track or off track, big-picture?

So this was all really helpful for me, because once the thing is going—"Hey, I don't know anything about anything, but we committed to 25 points and we got one done this week. Team, I think we need to do something different."

When you don't have a leg to stand on, that's when you need to rely on these tools. In fact, it was liberating. There's a saying that a good product manager is like a ship in the night. That's often antithetical to the traditional understanding of a manager, who has a big personality.

This I just wanted to facilitate: "Are you blocked on something, Julia? Let me figure out who can help you. I can't. I can't code. Someone else can. I know someone on another team who is really good at this. I'm going to get you guys together."

That was a really humbling and really good experience. I know it's a lot of different stuff in there. It just felt like a very flat structure. It felt super-productive. It felt like it was designed for everyone to do the best they could and be really creative.

There were some hard things about that. But the hardest thing for me was just not being able to contribute as much on the code side. But that also gave me that experience of being sort of the least informed person, which, for me personally, was really important.

JULIA TAYLOR KENNEDY: And you could concentrate on your role as facilitator.

DAVID OSBORNE: Yes.

JULIA TAYLOR KENNEDY: Obviously we do not have time for interaction on this point. But we do have a homework assignment for you, which is on your worksheet. We have distilled some of what David just shared into a list of questions. The way we would like you to complete this homework assignment is to exchange contact information with your buddy and keep each other accountable to thinking about this and how you can apply it with your team.

DAVID OSBORNE: Can we review the questions before?

JULIA TAYLOR KENNEDY: Yes, sure.

DAVID OSBORNE: The kinds of things we want you to think about:

• How agile is your team? If you don't know what agile is in the software sense, you should look it up. If everyone goes to pilates, that's great, but that's actually not what we're measuring here.
  
  
• How about your organization? Are you an exception in your organization or are you the norm? Are you doing it the way everyone else does?
  
  
• How could data help to achieve a compromise? This is something we didn't get into as much. When you're talking about things like managing time to time and you're talking about week over week, and maybe you don't know what the right answer is, sometimes having a data point can really help move things along.
  
  

For instance, Julia really wants the website to be red. I really think it should be green. Say, "Great. Julia, let's build it red and let's build it green, and let's see which performs better using data to settle that."

I found that that was a really helpful thing, to just get right to performance, get right to data, rather than spending two years arguing about Christmas colors.

So that would be a quick example of how you could do that kind of thing, to just move it along.

Questions

QUESTION: Travis, Carnegie New Leader. Thank you so much for coming and joining us.

Ethically, how do you think about using—you and your colleagues, so we can think about it more broadly—how did you guys think about using big data to make these emotional appeals to people? I know you kept talking about, in the case study, that you might be a little uncomfortable that you know all this information for a voter, etc. Did you feel that discomfort from an ethical perspective?

DAVID OSBORNE: I think there are two things. One is, we believed that President Obama being elected was the right thing and that it would have cascading benefits, which I think we have enjoyed. So reverse-engineering that, talking about the benefits that we currently enjoy and that would enjoy seemed quite ethical to me.

The discomfort you feel is not an ethical discomfort. It's an interpersonal discomfort. Putting yourself out there, it turns out, is a great exercise. You guys should do it. I don't want to get into a conversation about emotional vulnerability and your relationships. But this is one of those true things. And speaking of truth, a lot of what we were doing was uncovering true things. These things work, and we thought putting them to our means was ethical. We thought it was even good.

QUESTION: Thank you so much for your presentation. Very interesting. My name is Peter.

I just have a very specific question. Is there ever a point with the making-a-plan tactic when that's kind of counterproductive; it's too detailed; it seems creepy, like Big Brother or something like that? Then people would just shut down—"Whoa, whoa, whoa, you have all this information about me. Now you want to know if I'm stopping on the way to the fundraising event?"

Generally, if you could speak to that. Does it work to a certain point and then it's counterproductive? Or does it always just work?

DAVID OSBORNE: Great question. All of these things are templates, and you have to make them your own. Earlier when you were talking about your work on the documentary, you were saying, "I like to start soft and build up to it." Yes, you do have to make this your own. You have to try it on, until it becomes natural.

What I was doing with Leenda was a little bit hammed up, but not much. You do these things, and sometimes the other person reciprocates that discomfort. But in our context, if they didn't hang up, it was a win. So you have to find those limits. You just feel it out. Hanging up—we knew when we had gone too far. "What kind of coffee are you getting?" [Imitating dial tone] "Oh, that's a good one."

What we would do is we would literally role-play. We would experiment with it. And you'd be, like, "Was that too much?" "No. It made me a little uncomfortable, but it seemed like a reasonable question."

For example, when you're talking to a voter—"Hey, I know you're going to vote. That's awesome. You voted before. I'm so excited. I'm going to be voting. All your neighbors are going to be voting. The lines are going to be really long. If you can get there early, that's good. What time are you going to get there?" "I'm going to get there at 6:30 a.m." "6:30 a.m., that's awesome. So you're going to wake up early. Are you going to have breakfast? I always eat breakfast in the morning."

They're going to be, like, "You eat breakfast in the morning. That's normal that you just said that." (sarcastically)

The whole thing is weird. Why am I calling you to talk about voting? That's weird. For me, I was, like, "I already took this step. I might as well do it right."

So you play with that. Adding that detail about yourself may make it less weird.

But, really, once again, it's about your goal. You'll experiment. But I really do encourage you to try it out at work, whatever you're doing. You'll screw it up sometimes, but you don't do things better unless you try.

JULIA TAYLOR KENNEDY: Great. Thank you guys for those great questions.

It was a wonderful conversation. Thank you so much for joining us.