The Ted Cruz and Marco Rubio campaigns each rode to strong finishes in Iowa on the strength of well-organized and data-driven ground games. But the magic might not work as well in New Hampshire.
With four days until the vote, Donald Trump, by most polls, has a sizable lead over Cruz and Rubio. A new University of Massachusetts-Lowell poll of likely GOP voters has Trump leading with 36% of the vote. Rubio, after a strong finish in Iowa and a growing image as the GOP establishment candidate, came in second with 15% after recently leapfrogging Cruz, who got 14%.
The Trump ground operation is rudimentary compared to those of the Cruz and Rubio campaigns, which have invested millions of dollars in data and analytics. Trump has relied on his personal rally appearances and constant TV coverage to win supporters.
And it appears to have worked beautifully in New Hampshire, where Trump's anti-government messaging plays well with the many independents in the state. A strong voter targeting and get-out-the-vote operation can make all the difference in close races but isn't likely to produce a 10 percentage point swing, one campaign data officer told Fast Company.
Still, many believe the polls could be overestimating Trump's support in New Hampshire, as they did in Iowa. People who tell pollsters they support Trump don't always make it out to vote.
Even if one concedes victory to the Trump camp, there's a lot at stake for Rubio and Cruz. A second-place finish for Rubio would add a second surprisingly strong finish for the campaign. The Cruz camp is clearly hoping for a good enough showing in New Hampshire to keep the momentum going from its Iowa win.
The polling and voter data modeling operations in both camps are in high gear. That's partly because New Hampshire is a new ball game and the voter data models used in Iowa are of little use. New Hampshire turns out far more voters than Iowa, and the voters are less diverse, far less religious, more likely to be independent, and more likely to be interested in fiscal issues rather than social ones, as one campaign operative told me.
The state's non-partisan (no party affiliation) voters could be a huge wild card in the primary. In New Hampshire, voters must register, but they can register as a non-partisan and vote for candidates of any party.
"The state has a lot of independent voters who face low barriers to participate," said Mark Stephenson, founder of Red Oak Strategic and former chief data officer for Gov. Scott Walker. "They can easily request a ballot from either party so using data and analytics to know who might vote, mobilizing them, and targeting them with the right message is key."
Both campaigns have been busy polling in New Hampshire, searching for pools of undecided and independent voters who might be receptive to their candidate's positions, and who are likely to go out and vote.
It's a hard job. There are roughly 800,000 registered voters in the state, and around 260,000 registered Republicans who can vote in the presidential primary. There are 310,000 registered non-partisans. In 2008 roughly 240,000 people voted in the primary; In 2012 that number grew to 250,000.
"So it’s this process of narrowing down and it’s very hard to get to the target group of voters," says 0ptimus founder Scott Tranter. "It's like finding a needle in a stack of needles." Tranter runs the data operation for the Rubio campaign, reporting directly to campaign manager Terry Sullivan.
The task is even more difficult in the 2016 GOP race because of the high number of candidates still in the running, and because their positions on key issues aren't markedly different. All the candidates want lower taxes, smaller government, a stronger defense, a lower deficit, and an end to Obamacare. Their differences on those issues are a matter of degrees.
But those differences can matter a lot. With Trump somewhat of a wild card, the race between Cruz and Rubio (and possibly Bush) could be very, very close in New Hampshire.
"The results can be very close, and are often won on the margins," said Red Oak Strategic's Stephenson. "The job of analytics is capitalizing on those small differences in a strategic way."
"Mobilizing a small set of new participants, or persuading a few thousand voters in your candidate’s direction can change your position and ultimately, your momentum," Stephenson says. In Iowa, an additional 2,000 votes could have pushed Rubio into second place, one source said.
0ptimus's Tranter believes that in races where there are no "major fault lines" between the candidates' positions, the voter data modeling becomes even more important. The best-performing campaigns will be the ones that effectively identify and target people who can be counted on for a vote.