In the last Tech Bytes we looked at big-picture takeaways from the 2012 elections. This time, let’s get down to nuts and bolts. That means learning from the best: an Obama operation that so stunned its opponents that they didn’t realize how badly they were being beaten until election night. What are some specific lessons we can learn from this digital juggernaut?

Data-Targeted TV Buys
According to a December 3, 2012 Ad Age article, data was key to Obama’s television advantage: “His camp matched its extensive voter data with data purchased from Rentrak and FourthWall to determine what potential voters were watching. This allowed the campaign to buy advertising based on model voters as opposed to traditional age- and gender-based ratings,” said Carol Davidsen, director of integration and media targeting for the Obama campaign. “As a result, the campaign bought networks and time slots it might not have previously considered, including daytime and overnight, and national buys on TV Land and Syfy. Traditionally, political campaigns make buys based on content relevancy, with a huge dependence on news programs.”

As the Washington Post noted on December 12, the Obama campaign spent its money far more prudently than its Republican rival—a surprising result considering Romney’s reputation as a shrewd businessman. In part, the disparity arose because of the Obama decision to buy ad inventory early: in the summer of 2012, besides running devastating battleground-state ads painting Romney as an out-of-touch plutocrat, his media team also banked ad space for the campaign’s final stretch while prices were still low.

But a big driver of Obama’s TV edge was detailed data on the voters his campaign needed to reach, including when they watched television and what they viewed. While Romney and Republican independent groups competed against other political advertisers for top-dollar ads on local news and network broadcasts, this data let Obama’s team choose different ground on which to fight. Guess who got more bang for their buck?

Data-Targeted Grassroots Outreach
Besides creating demographic models of the voting blocs they needed to reach, Obama’s data managers went granular: they assigned each potential voter a numerical “score” based on criteria including their past voting record and donation history, along with information from commercial databases. The higher the score (from 1 to 100), the more likely a particular citizen was to turn out at the polls and cast a ballot for Obama. Later in the campaign, staff could incorporate the results of individual phone interviews and grassroots canvassing into their models, building a system that let them understand the electorate with a remarkable degree of detail. In fact, in the weeks before the election, they could run simulations of different electoral scenarios daily to test their preparations—unlike the Romney team, who fooled themselves by skewing their own polls to match what they wanted to happen.

Most importantly, this individual-level data helped the Obama campaign identify the exact voters they needed to reach, focusing for instance on those with a high affinity for Obama but a relatively low chance of making it to the polls. Rather than literature drops or direct mail, these voters got direct visits from campaign volunteers who lived near them, often many visits over the months leading up to the election. In other words, data didn’t replace person-to-person contact; it made it more efficient and effective.

As Obama’s Virginia GOTV coordinator described at a recent presentation, when she canvassed targeted neighborhoods in Richmond the weekend before November 6th (including public housing projects whose residents voted rarely), she found that the people on whose doors she knocked had been visited again and again. As a result, most were registered and ready to turn out: instead of turning them of, the regular contact had gotten them excited about voting for Obama. As she described it that was the moment she realized just how much the campaign’s emphasis on data had actually worked. And on November 6th, Obama won Virginia again by a three-point margin racked up in the state’s major population centers—exactly the parts of the state the campaign needed.

Inside the Cave
Election post-mortems few fast in November and December, but none of them contained the level of detail within “Inside the Cave,” a slide presentation prepared by Patrick Ruffini and crew at Engage. Patrick’s hoping to show his Republican peers how to catch up with Obama and the Democrats, but we all get to benefit from this comprehensive digital-campaign overview, including an organization chart and lots of outreach-related numbers. I’m using it as a resource in an expansion of the 2009 “Learning from Obama” e-book to include the 2012 campaign, and you can download your own copy at:

Obama Campaign Pro Tip
A quick tip from Eric Reif and Nathaniel Lubin, from a presentation these members of Obama’s digital marketing and advertising teams made at the recent RootsCamp conference in DC: when you’re running persuasion ads online, don’t use a medium uncommitted voters can skip. For instance, those video ads that appear when you’re trying to read an article on a newspaper website that allow you to “skip to content.”

Guess what someone who’s not committed to your candidate is going to do? Often, they’ll skip ahead and ignore your content completely. Instead, use a pre-roll ad or some other placement that they can’t dodge (Note that when targeting your own supporters for GOTV, the reverse is true, since you don’t want to annoy an ally.).

Television ads’ unskippability goes a long way toward explaining the persistence of the tube as campaigns’ main outreach tool, at least for now. But with more people consuming shows through Netflix and other streaming services, or saving them on a DVR and fast-forwarding through the ads, how long can this dominance survive?

Colin Delany is founder and editor of the award-winning, a 15 year veteran of online politics and a perpetual skeptic. See something interesting? Send him a pitch at