For many retailers and marketers, getting the type of actionable insight about customers that the Obama presidential campaign gained about citizens via data analysis – then used to sway voters – would be the equivalent of sales nirvana.
Staffers for the president’s campaign have historically closely guarded the “secret sauce” that helped them march to victory in the first big data election.
However, two former staffers of the Obama for America campaign have provided an inside look into the type of data analysis that helped Barack Obama secure a second term as president.
Evan Zasoski, the Obama campaign’s deputy director for data production, and Michelangelo D’Agostino, the campaign’s senior analyst for digital analytics, offer some tips for how businesses can use data analysis, and predictive analytics to unearth patterns and predict future customer behavior.
One of the team’s best practices revolves around isolating one “very quantifiable notion of optimization before going forward,” says Zasoski. The campaign has used this technique often to test how much money it should solicit from donors using historical giving data about donors paired with their patterns of giving.
“The campaign examined donors’ highest previous donation amounts and then did tests asking them for various percentages of those highest previous donation amounts,” the article notes. “They found that all versions of those requests did better than simply plucking a random number like $25 and asking donors for that set amount.”
Additional insights that the campaign staffers share include:
- In email marketing, don’t combine subject lines and content that perform well when they’re used alone and assume combining them will also have a significant impact.
- Creating models based on past responses to email campaigns can be used to more efficiently target future campaigns. This can be done by removing recipients who are unlikely to respond to specific kinds of request like volunteering but keeping them on lists where they would be more likely to respond.
- Give the people who are sending out emails direct access to the output of analytical models through Web forms so they can avoid emailing around lists that could be constantly in flux.
The type of experiments the Obama campaign staffers point out can be used by businesses to detect if a preconceived pattern is real, notes Darden School of Business Professor Robert Carraway.
“The key to a good experimental design is to ‘stack the deck’ against your pattern being real,” he adds. “If it emerges as still present in the results of your experiment, you are far more likely to have discovered something of potential value. By keeping experiments ‘small,’ in terms of time and resources, we can explore many more possibilities much more quickly. The net result is an organization that is in constant innovation mode, rapidly discarding ephemeral leads and doubling down on ones whose fundamental assumptions persist.”
Carraway advises firms to experiment on subsets of the big data they’ve used. Then if a pattern holds true for the subsequent sample, companies can further test it by designing an experiment to gather new, “live” data.
“The enhanced ability that exists today to spot patterns and identify potentially exploitable relationships must be accompanied by the ability to do some good, old-fashioned fact checking in the form of small experiments to confirm assumptions and hypotheses,” according to Carraway. “The million dollar question is, ‘Is it real?’”
- Subscribe to our blog to stay up to date on the latest insights and trends in data analysis and predictive analytics.