Five years ago, Barack Obama harnessed the power of social media to transform himself from a relative unknown into presidential-candidate 2.0, taking grassroots fundraising viral and defeating a field of more prominent candidates to become president. This year, his campaign staff put him in an even more technologically savvy roll: pivoter-in-chief.
Obama’s reelection staff didn’t just pivot once, twice, or 10 times over the course of his campaign. By accident or design, they followed the “lean startup” methodology pioneered by Eric Ries, whereby a company launches a rudimentary product, sees if people want it, then fine-tunes its strategy if it’s not popular, or stays the course if it is. Only the Obama campaign did it not with products, but with ads and online strategies, constantly measuring their effectiveness and making changes whenever reality didn’t jibe with their assumptions.
Here are five ways that the Obama campaign was run like a lean startup. If you know of others, let us know in the comments.
Top campaign staff developed a Nate Silver-like appetite for the data they needed to make decisions. “This is the part I’m excited about: This campaign has to be metric driven,” gushed campaign manager Jim Messina in an April 2011 YouTube video to supporters. “If something’s working, we’re going to go do a bunch of it. If it’s not working, we’re going to go throw it out.”
“If we just run that same (2008) campaign, we stand a good chance of losing,” Messina went on. “We’ve got to run a new campaign.”
That new campaign included a massive data-organization effort. Instead of a bevy of databases, some for fundraising, some for voter-drives, some online, some accessible only in campaign offices, as was the case in 2008, they developed a single, comprehensive database. Then they hired the quants. According to Time, their analytics department was five times the size of the previous election.
The Obama campaign organizers performed “split tests,” as Ries calls them, sending out fundraising emails from different figureheads to relatively small sets of supporters to see which one would be more effective. When First Lady Michelle Obama tested more popular than Vice President Joe Biden, her name would appear in emails sent out to larger lists, according to the Time article, published shortly after the election, which detailed the campaign’s statistical apparatus.
The campaign “micro-listened,” Obama campaign staff told Bloomberg News and Slate. They used the data–not just the basic age, sex, race, and neighborhood–but also consumer information, to figure out not just that, say, Joe Smith of Anaheim was likely to donate, but that he was most likely to donate the most online when sent an email by Michelle Obama offering him a chance to enter a raffle to eat dinner with the president, while his neighbor would likely give more after receiving a phone call from an in-state supporter. It was no longer a numbers game, a hope that their supporters would know how to best approach people. The campaign had, scientifically, figured out the best way to ask people for money and votes.
After introducing QuickDonate, a kind of one-click donation program that didn’t require supporters to reenter their credit card number, they analyzed the results. What they discovered was that QuickDonate subscribers donated four times that of average contributors, according to Time. Then they expanded the program.
The campaign pivoted away from television ads around local news to less traditional programming, like Sons of Anarchy and The Walking Dead, when they tested and found out they could be 14% more effective that way, Time said. They also figured out that women in their 40s were the group most likely to bid to meet celebrities, so they organized a fundraiser around dining with an actress that would appeal to those women: Sex and the City‘s Sarah Jessica Parker. And when they learned “turnout targets” used Reddit, Obama went there too, answering voters’ questions.
Don’t let the hundreds of micro-pivots that were an integral part of the campaign obscure Obama’s macro-pivot: the move away from the mass viral strategy that defined his 2008 campaign to a focus on granularity. In the earlier campaign, his grassroots supporters ran their own drives, figured out who to ask for votes, and who to ask for money or time. One supporter would get others to participate, and those supporters got others, and on and on. Without a doubt, it was effective. Obama raised a then-record $750 million, and 1.5 million volunteers registered on My.BarackObama, creating 27,000 individual groups to raise money and encourage people to vote.
Like all successful pivots, the concept for a new model just makes the old one seem outdated. “In 2012, we have the opportunity to make 2008 look prehistoric,” Messina said in 2011. The campaign not only raised $1 billion this time around, it won a rather stunning victory. After a mid-term when Democrats took a “shellacking,” in Obama’s words, losing eight governors seats, and control of the House, they won all but one major battleground state in the presidential race in 2012.
A year and a half later, it’s fair to say Messina was right: the 2008 strategy certainly does look antiquated.