Lessons from the Obama Campaign
Lauren Bacon | Monday, December 1st, 2008Unsurprisingly, we’re getting a lot of questions these days from our clients about how they can apply the online tools that helped Obama win the U.S. election. Typically clients come to us excited about one tool they thought Obama’s team used particularly well, whether that’s video, social media, email marketing, graphic design, or some other branch of the Obama web communications plan. But the tools are only a small part of the story; the real source of the Obama web campaign’s success, in my opinion, was its thorough, consistent strategy; its investment in a brilliant team of experienced staff; and the unprecedented breadth, depth and scope of its database.
Let’s start with the latter and work our way backwards…
Collecting and Mining the Data
Okay, the fundraisers and marketers out there will instantly grasp the value of a good database, but for the rest of you who are wondering what the heck is so interesting about crunching data, here are a couple of examples of things the Obama campaign was able to do because of its database:
- The campaign had instant access to over 10 million people via email. (Sidebar from the Washington Post story that reported that number: “The list is considered so valuable that the Obama camp briefly offered it as collateral during a cash-flow crunch late in the campaign.”) The Nation puts the list’s size (which they estimate at 11 million) in context: “It enables direct communication at a remarkable scale. The next President can instantly address 16 percent of his national supporters, based on the popular vote. To put it another way, the list dwarfs the audience of all the nightly cable news shows combined.”
- Volunteers who signed up to do phone canvassing could easily access lists of potential supporters, or non-voters, in their area by signing up to my.BarackObama.com, Obama’s social networking site, and entering their zip code. As a result the commitment threshold for volunteers was very low — they could sign up to make a half-hour’s worth of phone calls to neighbours if that’s all the time they could afford.
- On election day, one of the key tasks volunteers perform is to contact supporters (in this case, registered Democrats), both by phone and in person, to get out the vote. In the past, these volunteers would be provided with a full list of supporters, and they had no idea who had already voted; this time out, the Obama campaign tasked their poll checkers with a new responsibility: As one volunteer describes it, “We took the real-time results of who actually showed up at the polls and fed it back to the campaign so that they could adjust their GOTV [Get Out The Vote] calls and canvassing as the day wore on. Every time someone came in to vote, their names were entered into a computer system and their names disappeared or escaped, Houdini-like, from the call and walk lists.” The project was dubbed “Project Houdini,” and one can only imagine how much more efficient volunteers’ efforts were as a result.
So that’s what a good database can help you do. But the key to establishing a useful database is to set up the infrastructure well ahead of time, so that when the time comes to start adding names and profiles, you’ve got a framework that will stand up to rigorous road testing.
When it comes to political organizing, most parties and candidates start out with a pre-existing database of some kind, and often it’s a Frankenstein-like beast that’s been cobbled together over a series of elections. It’s critical to preserve the data that already exists, but usually that data is in need of a serious clean-up, and the database structure itself (from determining what data gets collected, to the kinds of queries that can be run on it) is typically a good candidate for a makeover. But more often than not, in the high-pressure, fast-paced campaign environment, data cleanup and database restructuring are tossed by the wayside.
Not so for Obama’s campaign database:
Howard Dean’s DNC brought the party in the 21st century. Under his direction, the party was able to create what the Republican Party already had — a single national voter file interface. The DNC struck data-sharing agreements with state parties, got everything uniform (more or less.) It spent tens of millions of dollars in 2005 and 2006 building this, much to the consternation of the incoming chief of staff (Rahm Emanuel) of the man who benefited the most from that money (Barack Obama.)
[Emphasis mine.]
Here’s what stopped me dead in my tracks when I read that: The work was done BEFORE election campaigning began. Someone with a vision (maybe Dean himself, given the groundbreaking nature of his 2004 online campaign, but more probably a group of people) made their database a priority during the post-election lull, when probably very few people were actually using the data — and when in fact, very few people were even thinking about it. That’s some solid long-term planning.
See this blog post from The Atlantic (from whence the above quote comes) for more details on the Obama campaign’s back-end setup; the breakdown is fascinating for those of us who gasp at the monumental task of organizing the amount of data the Obama campaign had at its disposal. And there’s another interesting piece about the daunting database cleanup process here.
Staffing Up
Let’s talk about staff resources for a moment. When our clients ask us about the costs of various technologies — whether we’re talking about video, podcasting, Facebook groups, or Twitter feeds — they frequently overlook the ongoing cost of the staff resources they will require to produce and deliver content through these various channels. The Obama website, email marketing, and associated social media channels were very well-managed; so what did they cost, staff-wise? We don’t have a dollar figure, but we can tell you that the Obama campaign employed 95 staff for its internet campaign.
So who was behind the campaign?
Answer #1: Some of the most brilliant and experienced minds available.
Blue State Digital, a company founded by veterans of the Howard Dean campaign (and which has honed its chops on a ton of Democratic political campaigns), provided the major back-end framework for the website, and two of the company’s co-founders key staff*, Macon Phillips and Joe Rospars, left Blue State to work full-time on the Obama campaign. (* Corrected @ 11:46 am, 1 Dec 08 - Rospars was indeed a co-founder but Phillips was not.)
The Obama website’s social networking platform, my.BarackObama.com (AKA MyBO), was set up under the guidance of no less than Chris Hughes, one of the co-founders of Facebook.
That’s just a start to the list of great minds who came together for this project; there’s a full list of key Obama staffers here for those interested in more details. Some fun stuff there for the real political geeks — you can start seeing the strands that connect various people and campaigns. (It’s also here that I discovered the name of Obama’s chief speechwriter, who by the way turned 27 this year.)
Answer #2: An absolutely killer volunteer base. Obama’s volunteers were well-organized both on- and off-line, but let’s stay focused on the online effort here. Volunteers were given an enormous amount of leeway to self-organize on Obama’s campaign, which is consistent with the President-Elect’s repeated claim throughout the campaign that the campaign was not about him, but rather about the people who were electing him.
Have a look at this article at CIOZone (actually page 3 of an in-depth and well-researched 6-page article) describing the self-organizing tools Obama’s website provided to supporters; it indicates that during the primaries, staff support for volunteers was minimal for states where Hillary Clinton was expected to win, and furthermore, “in places like Pennsylvania, Indiana, and Oregon, whose primaries might not have mattered as much if the race had been decided quickly, volunteers who found each other through the Web site organized their own freelance pro-Obama efforts long before the first campaign staffer came to town.”
In other words, a great deal of trust was placed in volunteer organizers, who held events, made phone calls, knocked on doors, raised money, all in the name of the Obama campaign and with minimal oversight from staff. The CIOZone article linked above goes into some depth about the relationship between staff and volunteers; I’m impressed (and a little suprised) by how few problems the campaign seems to have had with their volunteer base, given the size of the group they were working with. I would have expected more kooks in the mix, to be perfectly honest, but it sounds like they didn’t require much policing.
One description of the volunteer-management system goes like this:
[T]he system works like a pyramid, with state officials given access to a lot of functionality, and growing numbers of people below them, down to the volunteer level, allowed fewer and fewer functions, depending on what they need and how well they are known or trusted. The Obama system learns as it goes along, allowing volunteers to feed information gleaned from their work back into the database via their web browsers. Campaign staffers at the local, state, and national levels can see which volunteers do the most work and get the best results, making the organization more efficient over time. Nationwide, MyBarackObama.com has more than 1 million individual user accounts and has been used to promote over 75,000 campaign events.
If that’s accurate, then it sounds like the Obama team struck an appropriate balance between monitoring and managing their community, and allowing supporters enough leeway to do their own thing without too much interference from above.
Strategy First, Then Execution
A solid database is an essential campaign tool; great staffing is another. But tools are only tools — without a smart strategy, they can only get you so far. The Obama campaign seems to me (from an admittedly outside perspective) to have had a brilliant and highly consistent strategy from start to finish — and that strategy continues even beyond election day.
Here’s my off-the-cuff, incomplete summary of that strategy: Create an inclusive, forward-thinking movement that embodies hope and positive change. Empower self-starters and problem-solvers to effect that change. Set up mechanisms that allow supporters to connect (to each other and to the campaign), contribute their ideas, and elect Barack Obama.
That strategy informed decisions across the board, from the messaging that came through in speeches, interviews and debates, through policy statements, graphic design, and the selection of online tools. It was also consistently carried through in all online campaigning. Emails sent to supporters carried through the messages of hope, change, and the need for entrepreneurial ideas and hands-on election support. The website drove a million people to sign up for my.BarackObama.com and help elect the president they believed in. And perhaps most importantly, supporters were invited to make the campaign their own — to reach out to their neighbours, throw a fundraising party, post a note on their Facebook or MySpace profile, whatever the action that fit best with their personality, strengths, and social network.
Beyond the election, we’ve seen the launch of Change.gov, which promises to carry through the election campaign strategy into the transition period (and possibly beyond, though there are questions flying about how the election database can make its way to the White House without becoming the property of the U.S. government in perpetuity). There’s talk of how the White House site may evolve when the President-Elect becomes the President, but much remains to be seen there. We do know that some of the key staff members from the campaign have been hired on as part of the transition team.
One of the big questions that is on the minds of a lot of people right now is how the grassroots movement that helped elect Obama will interact with him as President. There’s an interesting piece here on that subject; here’s an excerpt:
[Organizing guru Marshall] Ganz makes three really important points: The first is that we’ve never had a president enter office with an organizing social movement attached to him, and there’s no precedent for thinking about how the participants in that movement have a voice in his presidency. The second is that this movement isn’t going away, and the critical question isn’t “who’s going to get the list” but how will this movement govern itself. The third, which is somewhat of an open secret, is that there is a group of organizers meeting in Chicago right now trying to figure this out, and Ganz believes that their deliberations should be more open. “I think it’s important to create the public space for this kind of discussion,” he told me.
I could go on and on about this campaign and its lessons, but I think I’ll stop there for now. If you have thoughts to share, please leave them in the comments — I’d love to hear them.