What Obama’s 2012 Election Can Tell Us About Using Data to Build Communities
The 2012 Obama campaign is often held up as the example par excellence of disruptive impact of data analytics. Plenty has been written about the campaign’s use of data in voter targeting to identify and engage online audiences in a way that far exceeded the Republican effort.
But it wasn’t just on the web where the campaign had an advantage. Obama for America’s tech team, comprised mostly of political newcomers, also used data analytics to rethink the way television ads were bought, and where its army of volunteers were sent. One of the big breakthroughs, says Carol Davidsen, director of integration and media targeting for the Obama campaign, was in using data to navigate the twisted world of local television, and finding audiences in unsuspecting — and often far less expensive — places.
In the lead-up to Street Fight’s Local Data Summit in Denver on February 25th, we’re taking a deep dive into the world of local information, speaking with some of the sharpest minds in the industry about using local data to make businesses more efficient, and to make experiences richer for consumer. Street Fight recently caught up with Davidsen, now-founder of Cir.cl, a peer-to-peer marketplace, to discuss the key learnings from the campaign and trials and tribulations in bringing those techniques to the private sector. (Davidsen will be a keynote speaker at the Local Data Summit.)
One of the areas where you’ve the Obama campaign thrived was the use of “big data” analytics to identify audiences. Talk a bit about your experience on the campaign, and the way in which analytics changed how you approached political operations.
For me, analytics companies are more like law firms. Today, a lot of analytics companies want to be more like a software company where they build products where a user is simply going to press the button that answers every customer’s questions. The customer puts in their data, they press the button, and they have their answer. But here’s the thing most people don’t realize: data is only as good as the question you ask. There’s this huge push to data and analytics and people sometimes think that means that their gut or their instinct disappears and I don’t think that’s true because you still need to ask the data questions.
You still need to have a theory or speculate about something and then you arrange the data so you can ask a question and find something out. For instance, when I looked at historical campaign reporting and how we’d spent money in the last election, I saw that there was huge waste in the way we bought television ads. The goal of older school advertising was to target people by doing content to find relevancy. So if people watch the news, the idea is that they’re into politics, or they’re engaged in their communities. But the local news knows what’s going on, so during elections they jack up the prices like crazy.
In the local market, the ability to determine a device’s location algorithmically has separated content from targeting in a similar sense. In a data-rich environment like mobile, what’s the role of content?
When television first started, brands would sponsor entirely television shows — or, in the case of soap operas, actually bank the full production. While things have changed, that focus on content has left its legacy. People still think that they need to be in certain programs, and I think that’s actually very judgmental of your audiences. In political campaigns, for instance, people will quickly brand these 40 percent of people that don’t watch the local news as low-informed voter. That’s actually a huge judgment on the people that you’re trying to convince to do something. I don’t watch the local news and I’m clearly not a low-informed voter.
Instead, it should be — even though I’m the President of the United States, I don’t care what the content is. I care where the audience is or what it is that people are watching. If that’s “Honey Boo Boo,” I’ll put it there. If it’s reruns on TV Land, I’ll put it there. I don’t care, I just want to reach the people. At the end of the day, I think a lot of people try to push these content restrictions instead of allowing actual math to tell you answers.
Proximity is just a reality. It’s like designing for gravity.
There’s obviously been a lot of focus on the various impacts of ‘big data.” In the political realm, how did new data and analytics tools allow your team to create a new approach to campaigning?
David Seamus, who was the director of messaging on the campaign, a very hard-core politics guy, started working on super-local elections. The way you run them is very different than how you traditionally ran a national campaign. In a local election, you would get your biggest supporters in a room, you all sit around the table, and you go through your list of voters in that area, and everyone’s like, “I know that guy. You know this guy. Or, so and so knows that guy.” But in national politics, typically that sort of personalization couldn’t happen.
I always like to think that while technology like planes and trains brought us globalization, the Internet allows us to brings back the personalization that was endemic in a more local world. For the first time, at a national scale, data allows us to actually think about each individual again as each individual because I understand their behavior. I know who they are.
This is where there’s fundamental change. Where it used to be all about customer relationship management (CRM) and corporations managing their relationships with their customers, now because of the way the Internet it’s more about people managing my own relationships with brands. It’s becoming less about the CRM, and more about trying to build a model where consumers can take charge of these relationships.
Presidential campaigns have effectively unlimited financial and human resources, making them more the exception than the rule. Given those circumstances, how transferable are these tactics to the private market?
I really just don’t think you’ll never be able to do it in the private market. That’s why I’m not building another targeting company because I don’t know how to scale it down to the level where it actually make it work. Everyone’s really excited because they’re like, “Oh, it’s a 70-billion-dollar-a-year industry and I want to take some piece of it. But the reality is that its not like another 10 billion dollars is just going to show up. It’s just going to be money that we take out of our pocket that we put into another pocket.
Access to information and the ability to act on it are two very different problems. How did the campaign put those data into action?
The question for the campaign, and the one we’re working on at Circl, is how do you help people spend their social capital better. It’s coming towards Election Day, and I’m a huge supporter of Barack Obama. I could spam all of my friends with posts about the need to get out to vote, but a lot of my friends have stopped listening to me, or simply don’t care anymore. What we want to be able to say is: “Carol, you’ve got 638 Facebook friends but there’s only two that you should be talking to today because early voting started in the United States and they haven’t voted yet. Ping these two people directly.”
So what we learned is how to better use that first line of supporters, which any quality product or campaign will have. And with technology you can do this in an automated way, at scale. On the campaign, both in 2008 and 2012, we we’re very focused on making our volunteers more effective. If you give someone a list of people to call and every single person hangs up on them, that volunteer in general is less motivated to come back next time because they didn’t have a great experience. So a big part of it is making sure that they’re knocking on doors of people and places that they know best. They’re using their social capital effectively.
Talk a bit about the intersection of personalization and local.
Local is an interesting word to me because it can quickly mean a lot of things. And I believe that often, geography is too limiting a constraint in defining local. In our work on Cir.cl, we’re finding that lot of companies that have built products in the sharing economy are super focused on geography and I think a lot of that is because people are building mobile apps, and they become obsessed with the geo-location capabilities. In some cases, yeah, I’m buying a couch. And I want to buy a couch somewhere near my house.
But if we’re talking about a somewhat smaller audience, like my local community of people who are into a particular type of comic book — that’s really just all the people that care about that comic book whether they’re near you or across the country. In that sense, that’s my local community. The people who are local to an idea or interest.
Proximity is just a reality. It’s like designing for gravity.
Steven Jacobs is Street Fight’s Deputy Editor.
Find out more about how big data can be used in local context at Street Fight’s Local Data Summit, taking place on February 25th, in Denver. Learn from and network with some of the top local data experts in the country. Reserve your ticket today!