Big Data Case Study - How Big Data Help Obama win re election


Traditional punditry is nothing new during an election process. And when it’s the process of choosing the president of the most powerful nation the United States, then punditry, analysis, predictions, etc. seems breaking all the records to prove their claim authentic yet realistic.

But, wait a minute. Have you ever thought how the collection of massive data has been processed to predict Barak Obama’s re-win? In simple words how big data helped Obama win re-elections? Yes, you heard it right, it was the magic of big data analytics that helped the president of the strongest nation wave his magic wand again and create the aura for the re-electoral win.

While traditional pundits and conventional news media stressed on the eve of the United States Presidential election a close contest, there were many predictions out there indicating that it’s gonna be like clean-sweep or a one-sided contest. But, one thing that came out of this poll is that not all polls are made equal. The polls garnering the potential of latest data processing and analysis techniques tended to have accurate predictions and success rates. Those who showed patience and watched it for long 22 months found out that it’s not just the war between ideologies and campaign issues, but between old media gurus and latest media analysts as well.

If we go a few years back when US presidential elections were held and dig in then we’ll find that their dissimilar polling aggregators used different methods in order to foretell the results of the elections. In June 2012, Drew Linzer posted that the election would be won by Obama with 332 votes and Romney would get 206 votes. Over the months, this prediction didn’t change and when new data came in, the counting was- 332 votes for Obama and 206 for Mitt Romney.

Did you find any remarkable stuff here? Well, it’s not just that tech savvy pollsters are making use of prophetic analytics to conclude results, but just how exact and accurate these predictions have proved as compared to old methodologies. So, these results indicate that the world has now a better predictive instrument than what we used a few years back, and these predictions are influencing the outcome too. Companies are also using the power of Big Data and they are giving Big Data Hadoop Training to their staff for better analysis of data.

If we talk about President Obama’s election campaign then we would find that it ran under utterly sophisticated yet persistent digital operation that kicked out the conventional method and didn’t take the assumptions for granted.

Their advertising campaign was clever enough to draw the attention of the masses. Their campaign was instigated by data analysts who did real work and didn’t depend on chance. They refurbished their virtual campaign centre named and influenced supporters to share their personal details, comments, share photos/ videos and even donate funds. This was their straight weapon, which they used cleverly. What happened afterwards is even more interesting. It was a multi-pronged engagement strategy with the audience and the webmasters used this supporters’ content to attract others and boost traffic to other campaign sites including social media and networking sites.

While Obama’s campaigning flourished, the Romney campaigned struggled to outshine them in this digital race and they remained connected to the old-fashioned channels to attract headlines and influence people, however, they just attracted 23,700 subscribers along with 26 million page views.

This signifies that data mining is the game player and the campaign of Obama targeted even the micro supporters. The whole volunteer system of this campaign along with an inconsiderable number of folks was thoroughly analyzed and call lists assigned based on the likelihood to match. In order of persuadability, call lists ranked names- 75\% of data included fundamentals like age, gender, address and voting record, however just 25\% of data helped them to foresee who’s going to make donations online, do did it by email, and who would become a helper.

When people get aware of the campaign it had created enough momentum with inciting supporters, it accelerated fundraising and then ended up with large voter turnout. If we dig out a little more then we’ll find that every single night, the campaigning team track 66,000 computer simulations over how the elections might turn out so as to find the maximum breakdown. Such practices panned them onto avenues they had not taken ever, for example, Reddit- a social news site. Also, it was the very first time that cloud computing had been introduced in a campaign. The team Obama ran their data particularly in AWS (Amazon Web Services) and made use of open source software & Amazon services to economically write or modify them their own programmes.

The experience of team Obama will tell you his campaign wasn’t super expensive, but realistic micro-targeting messages tailored to the needs of every reader. The messages were directed in such a way that they appeared differently to different people.

This is something that you as being the corporate marketer or CIO can repeat effortlessly. The same campaign can be run with little changes to drive more consumers and analyse their behavior all through the campaign.

Hence, it can’t be ruled out that big data played a pivotal role in Obama’s win in re-elections.

Get Weekly Free Articles

on latest technology from Madrid Software Training