In the days proceeding the Iowa caucus, as the media fussed about the unpredictability of the horserace, confidence quietly swelled within the Cruz campaign. “Data,” said a high-level Cruz campaign operative on election night, “is huge.”

In fact, data was critical to Cruz’s win in the first vote of the 2016 campaign.

For months, Iowans have been swamped with paper mailers and phone calls. Superficially, these direct mail and robocalls are old-school, tried-and-true get-out-the-vote campaign tactics. But today, these traditional methods of voter persuasion are powered by cutting-edge data analysis tactics. Big data tells campaigns what to mail, who to call, when to call, and what to say.

SEE: Is tech breaking the 2016 election?

The Cruz campaign used personality-based data modeling, said the operative, and “machine learning to scale outreach.” Data-driven behavior modeling–predicting action based on social media activity–and microtargeting–personal and direct communication in the form of phone calls and in-person house visits–allowed Cruz to activate an army of volunteers, and lock in caucus-goer loyalty.


In his book The Victory Lab, reporter Sasha Issenberg explained how analytics help inform old-school methods of voter targeting. “The scientific revolution in American electoral politics has relied on lots of technology, particularly to assemble and sift through large databases,” wrote Issenberg, “but its most lasting impact may be a resurgence in lo-fi tactics. The genius alchemists behind microtargeting spend their days deciding where candidates should send postcards.”

Internal data analysis is frequently more cost-effective, and more accurate. “Polling still matters,” said another Cruz operative, “but we trust [campaign] internal data more than we trust the polls.”

A shift towards internal data analysis is occurring in every presidential campaign. In 2014, data news site FiveThirtyEight reported, “there are fewer high-quality polls than there used to be,” and tied the decline in polling to disruption also occurring in the business of news. “The cost to commission one can run well into five figures … under budgetary pressure, many news organizations have understandably preferred to trim their polling budgets rather than lay off newsroom staff.”

The Cruz campaign also learned a few tricks from the 2012 Obama campaign’s mastery of big data. On question and answer site Quora, Luke Riley, 2012 Virginia Deputy Data Director for Obama for America, said about the marriage between modern data analysis and get-out-the-vote tactics:

“My favorite experiment was when we were building a model to find the most persuadable people in our universe. That is to say, the people that take the fewest number of calls or door knocks to convince them to vote for the president. Basically, a round of robodials [sic] went out to see where on a scale of 1-5 their support for the president was. Then we did live calls from volunteers out to those people to try and persuade them. Then another round of robodials made the same 1-5 assessment as before. Then all of that data was used to see what type of person was most effected [sic] by the live call from a volunteer. This may all sound simple, but remember that this is in the middle of a giant campaign happening.”

The Cruz campaign built a small team of data scientists internally. The data team relied heavily on open source data packages, and built Python apps in the cloud using Amazon Web Services to create personality models based on demographics and psychographics.

SEE: Election Tech 2016: The 4 technologies that will decide the next US president

As they dissected voter information, the Cruz data team worked in tandem with the communication team to target voter mailboxes, phone numbers, and social media feeds with pitch-perfect messages.

The Cruz data machine gobbled up, and refined signals from both mobile application users, and from social media. Speaking of mobile apps, the Cruz team built one that gamified the election process and fed the campaign even more valuable data.


The Ted Cruz 2016 app for iPhone and Android is arguably the most sophisticated app deployed by the current field of presidential candidates. Like many mobile apps, the Cruz application connects with Facebook to send the campaign basic data about demographics, social graph, and user location. Users were awarded points based on participation, donations, and recruitment, then ranked publicly on a leaderboard. “Mobile gamification adds a layer of competition, and brings people back to the app,” said a Cruz spokesperson.

Social Media

While @ realDonaldTrump is a powerful presence on social media, so too is @TedCruz. According to data aggregator, on Iowa caucus day Trump’s social media posts reached more than 13 million Twitter and Instagram users. Cruz reached over 10 million users, and @MarcoRubio finished in third with just over 5 million users reached. However, where Rubio and Trump’s follower activity peaked after the Caucus, Cruz’s social engagements continued to rise. According to, Cruz was also mentioned on Facebook and YouTube slightly more often than his GOP rivals.

All of the top three GOP candidates as well as the top two Democratic candidates performed well on Twitter on caucus night. Rubio’s surprising performance in Iowa was predicted by a steady uptick of Twitter followers leading up to the caucus. This could also suggest a social media tailwind propelling Rubio through the New Hampshire primary next week.

SEE: Election Tech 2016: How social media and big data changed everything, a Q&A with Joe Trippi

While @ HillaryClinton won a squeaker victory over @BernieSanders in Iowa, Twitter follower data still suggests that Sanders’ enthusiasm online remains unabated.

TechRepublic’s Election Tech 2016 coverage

To better understand the relationship between big data, mobile, and social media on the 2016 campaign, TechRepublic gathered and and analyzed publicly available data in the weeks preceding, and during the Iowa caucus. We sliced some data in-house, and also used to aggregate and interpret caucus and engagments such as Likes, Tweets and Retweets, and new Followers on Twitter and Instagram.

TechRepublic data and charts were produced with assistance from political scientist William P. Stodden, using Microsoft Excel.

Over the course of the campaign we will continue to perform simple data analysis. In the future we hope to correlate sentiment with follower actions like retweets and likes. We hope to uncover additional and unique insights.

If you’re a data scientist, social media professional, or inquisitive TechRepublic reader we’d love your ideas on how to inspect campaign social media data. Please leave a comment below, or ping us on Twitter @ TechRepublic.

Read more: