Turning Big Data Into Smart Data: 5 Lessons For Marketers From The Obama Campaign

Joe Rospars, founding partner of Blue State Digital, talks about how marketers should not be seeking more data, but better relationships.

If nature abhors a vacuum, the business world adores a buzzword. And for the past few years, data—specifically big data—has been among the most buzzy. That’s been especially true since the 2012 election in which President Obama’s campaign made waves for its surgical use of data in winning a second term.

Given its stature in the marketing world, big data isn’t very well understood. The one thing most people know is that it’s, well, big. An IBM report on big data states, “Every day, we create 2.5 quintillion bytes of data, so much that 90% of the data in the world today has been created in the last two years alone.” But how marketers should be using this data, and how it does or doesn’t integrate with the creative marketing process, is less clear. How do you make sense of a quintillion bytes of data?

Joe Rospars, co-founder and CEO of Blue State Digital, the agency behind both of Obama’s campaigns, says harnessing the power of big data is not about simply analyzing antiseptic information, it’s about using whatever information is at your disposal to understand the people behind it all. “Big data is about having an understanding of what your relationship is with the people who are most important to you and an awareness of the potential in that relationship," says Rospars.

On the campaign trail, that meant understanding how Obama’s supporters were connected to people outside of the organization’s reach. “If you have customers or members that you are in touch with, bringing the whole picture of your interactions with them to bear on the organizational side allows you to ask them to do particular things,” Rospars says. The 2012 campaign upped the data ante through a program dubbed Narwhal—a platform that unified all of the campaign’s data, such as voter, donor, or volunteer information—which allowed organizers to get a picture of the person behind the bits.

Once individuals emerged from the data, for example, the campaign realized that they only had phone numbers for 50% of potential supporters. Given how important phone-based campaigning is, that was a problem. However, the data showed that for the 50% of people they didn’t have numbers for, 85% of those people were one degree of separation on Facebook from the people they did have numbers for. “We used this information to email our supporters and ask them to reach out to their friends. We said, ‘You know these people, please give them a call because you might be the only call they get.’”

For Rospars, the power of big data is not its size, but what you do with specific segments of that data. While a political campaign is a different beast than a brand marketing operation, there are good lessons that can be learned from politics which, after all, also revolves around influencing individual behavior at the most singular level. Here Rospars takes us through five key things marketers should remember when wrestling big data.

Understand there’s a person behind the data.

I think the biggest misconception about big data is that volume somehow equals sophistication. It doesn’t, necessarily. You can have a ton of data, but if you’re not using it intelligently, there’s no point. Better to have a program that is ultimately about people and the data is a reflection of your potential relationship with those people. When we speak with companies or organizations, we say it’s really about understanding how your organization relates to people. On a campaign that could mean talking to anyone from a first-time voter to a regular contributor. Understanding that all of the data we have is fundamentally about a relationship with people is a very key insight that very often gets left out.

It’s not about big data; it’s about smart data used at scale.

“There’s a tendency with big data to say, ‘Mmm… I just want more data.’ But basically everyone already has too much data and the question is how to use it and integrate it within your organization. It’s about using it intelligently in terms of structuring your conversations with people around the best data you have. In our case, it was pretty boring in that our data was publicly available voter files, but it was how we interacted with people and tracked it and measured it and built models around what we thought would make people donate or volunteer or vote. So being smart about the data we had all along and integrating the systems on which people are connecting with people was a big deal.

Smart use of data has the potential to make all of the rest of your decisions smarter. So whether it’s buying traditional TV or print or individual direct marketing, and the overall strategy, the smart use of data has the potential to unlock value and insight across everything. It’s not a silo; it’s a sandbox.

Tools are just that…tools.

All of the tools we used or created were answers to challenges posed by our original mission and strategy: to go and have relationships with a whole bunch of people on the president’s behalf, get them involved, and use relationships in their communities. If your strategy relies on getting a whole bunch to do something, having the notion of a person in your database is very important. You can start by saying we have this great tool, but you have to know what you can do with it.

There will be different answers to the questions of what tool or which outreach method to use depending on what your overall strategy and mission might be. Whether you use a hammer or a nail gun or a saw, you need to have an idea of what kind of house to build. And that’s something that a lot of people want to skip because they’re hard questions. When you build a relationship with people, how are you going to serve them, how are you going to interact with them, how are you going to listen to them, and how are you going to give them ownership over your brand? These are the questions a lot of organizations don’t want to answer.

As tools become more ubiquitous, they become easier to use and integrate into your strategy so that people can feel as though they can take the risk to build that people-based strategy and actually have meaningful connections.

It’s not creativity or data; it’s both.

At its best, data and creativity combine and are harmonious and result in the best program possible. Many organizations tend to really be focused on and place a big bet on one or the other. But for us, it was about trying to hone both approaches. With fundraising campaigns, we’d have an opportunity to spend a week focused on trying to get people to give us money. A bunch of ideas would come from the creative process. They’d be honed from 27 ideas down to six that we loved. Then we’d build out multiple approaches for those six ideas and we’d then roll those out and test them. It’s about putting good stuff into the machine and then following the results. Over time, it starts to become a virtuous cycle and you learn. The idea that didn’t work, that doesn’t mean it wasn’t a good idea; it’s just that it didn’t work on that given day for a whole bunch of reasons.

For example, at one point we were being heavily outmatched in fundraising by the Romney campaign itself. We’d expected to be outspent by the Super PACs by about 2-1 or 3-1 and still win. But it was looking like 7-1, 8-1, 9-1 from the Romney campaign itself. It felt, in our gut, like something the voters needed to know about in terms of understanding the landscape of the election. We tried to communicate the problem several ways but none of it seemed to work—until the fifth time we tried it. That email we sent was the most successful one we had sent in the history of the Obama organization. There are lots of reasons why that worked on that particular day—things that were in the news, our copywriter’s copy. We could have said, 'Well, the data shows that people don’t care about it.' But we didn’t give up on it.

Data Up The Organization

Data, like digital, like tech, has to be something that a leader is comfortable with and native to and able to hold their internal folks accountable to. You can’t just say I’ve checked the data box. Leaders at the C-suite level and the people that sit on their boards really need to understand the data and what it says about their relationship with people, and then what their organization is doing in the digital space and traditional marketing to change that relationship, and then how are they are going to measure the success of those efforts. That’s a big thing, but it does ultimately boil down to leadership. Having the people who are digitally native have a seat at the table is something that every leader can do.

[Images: Flickr user Jared Tarbell ,Kristina Alexanderson, and Ginnerobot]

Add New Comment

3 Comments

  • Robert Jacobson

    ...And wouldn't it be nice if politicians like the President actually cared for their constituents rather than special interests (mostly banks, military contractors, telecom/media monopolies, and oil companies) beyond the duration of the campaign.  I realize this is not the responsibility of campaign operatives like Joe, whom I remember from the "(Howard) Dean for America" campaign had a passion for blogging also not shared by the candidate.  Obama is living off of resentment against his opponents, who so far have shown themselves even less able to connect wholeheartedly with the American people, but gradually he's losing his edge, becoming as ingenuous as they were.  By the end of this term he'll be regarded as just another Presidential Huckster.  Then it's time to share advice above loving up the electorate with the next posseur.  It must get old after awhile.

  • MADE BOARDS

    Rae, I agree with your insights into the the application of consumer data. We are a start up that uses performance tracking data to design products and provide insights into athletic performance. The truly remarkable aspect of what we are doing is using data as the primary design driver - and then turning those measurements into 3D printed sailboards, surfboards and stand up paddle boards. 

    The board sports industries are inundated with overlapping product selections - yet none meet the exact needs of the consumer. Our approach uses data collected through a mobile application that combines geographic, atmospheric and performance data to design the perfect board for the rider, that's then further customized by the rider and printed on-demand. A data-driven approach allows us to not just provide insights - but create usable goods that are the manifestation of behavior. To quote you:

    "You can have a ton of data, but if you’re not using it intelligently, there’s no point. Better to have a program that is ultimately about people and the data is a reflection of your potential relationship with those people. " 

    We're creating a relationship between board and rider, a more intimate interface, an extension of the human body by surfacing large-scale patterns that go unnoticed by the rider. These patterns can then be compared to the patterns of other riders and provide deeper insights into individual performance. The result? A better, and more fulfilling relationship with our brand. We're less concerned about selling more equipment and more concerned with making sure you have exactly what you need to have fun. I can't think of a better use of data.

    www.madeboards.com