I certainly did learn a lot from “Learning From Obama.” The 2009 epolitics report was a useful, clearly written handbook for anyone looking to replicate the results of Obama’s 2008 campaign.
One of the sections I found most intriguing and helpful was the testing of campaign initiatives, particularity emails.New media team director Joe Rospars regards measuring and testing as an integral aspect of a campaign. So much so that he preferred Google AdWords over display ads purely because of the ability to measure success. The report read, “The point: what’s the use of doing something you can’t test? If you can’t test it, you don’t know how much good it’s doing you, and your money might be better spent elsewhere.” Very true!
The way that the Obama campaign tested mass email communication was to use varied emails to their listservs and gauge the responses. The email recipients were sometimes grouped randomly and sometimes divided by demographics/age/location. The emails sent out tested responses dependent on who the emails were sent from, message, subject lines, imagery, links and other adjustments. The responses, which included information like messages opened, actions taken, donations made, allowed them to determine how to change the emails for next time.
Now in the case of the Obama campaign they were working with lists of millions of names. For a campaign that you or I would be working on we would most likely have significantly fewer people to test communication on. If you are planning blanket emails to your entire list I think that dividing test groups randomly is probably the best for return data. This will give “average” results representative of your whole audience. If you have a segmented campaign that targets people dependent on location or age then demographic division would be best.
I am not sure exactly how the Obama campaign did it but I think that only one element of the email should be changed across your test groups per email distribution. This allows all other constants to remain the same and you will really know if a high response rate is due to your funny subject line or not. Too many varied elements will only muddle the outcome.
The other section of this report that stood our to me was how to respond to negative opposition. I thought it was very surprising, but very smart, to address the Obama-is-a-Muslim-like remarks and arm supporters with information to combat these ill-informed opposers.
The campaign also launched web pages and online action groups to fight the underground, e-mail whisper campaigns and robo-calls that surfaced in battleground states. In one effort, the campaign urged supporters to send out counterviral e-mails responding to false rumors about Obama’s personal background and tax policies.
“Propelled by Internet, Barack Obama Wins Presidency,” Sarah Lai Stirland, Wired.com, 11/4/2008
This meant teaching Obama supporters how to respond in favorable, non-hash ways when they are included on forward email chains. Such as responding to those group emails with a reply all that clearly outlines the truth and links back to credible sources. Building off the we-listen-to-peers-more discussion in the report, I think that this counter argument would have a better impact on the people who have been exposed to the negative message because it is coming from a real person, rather than a news article (though the articles are used as evidence).
Negative opposition can also act as a catalyst because supporters are more energized to prove the opposition wrong. This was the case after Sarah Palin’s speech that mocked community organizers, which “functionally lashed out at everyone on the Obama list who’d embraced the campaign’s organizing model.” The report said that Palin’s speech was followed by Obama’s biggest day of political fundraising ever. Turning negative into positive!