Political parties hook voters, map dissent with Big Data

Some political parties also use Klout—an app that uses social media analytics to generate a number between 1 and 100 that represents a person’s social media influence.
Some political parties also use Klout—an app that uses social media analytics to generate a number between 1 and 100 that represents a person’s social media influence.

Mumbai: When the Indian Navy submarine INS Sindhuratna caught fire off Mumbai on 26 February, the Bharatiya Janata Party (BJP) found heightened activity on social networks, but analysed that this was restricted to the metros and cities such as Pune, Chandigarh and Dehradun that house a large population of ex-servicemen.

This information was one of the inputs the party’s media cell used to decide what its response should be, who should deliver it, and who should appear on behalf of the party on TV debates, claims the BJP information technology (IT) cell’s national co-convenor Vinit Goenka.

Taking a leaf from US President Barack Obama’s 2012 campaign that took recourse to analytics to garner votes, Indian political parties are moving beyond just having a presence on Facebook, Twitter, Google Hangouts or YouTube to using Big Data (read: analysing large amounts of data to cull insight) to influence voters ahead of the general elections scheduled for April and May.

Besides holding workshops to educate candidates about social media practices, parties such as the Congress, the BJP and the Aam Aadmi Party (AAP) are employing sophisticated social listening, reputation management and analysis tools to identify and nurture social media influencers—like those Twitter users who have specific party leanings and thousands of followers.

They are using digital management platforms to crunch and analyse the data, intelligent SMS tools for targeted messaging (to a specific state or body of people, for instance) and algorithms that can determine the age, demographics and location of the targeted voters.

For instance, around the time that Parliament was taking up the contentious Telangana Bill, Goenka and his team mapped elevated social media activity from cities such as Pune, Bangalore, Gurgaon and Noida and concluded that techies from Andhra Pradesh living in these cities were participating in discussions on the Bill.

Like in the case of the submarine accident, this data became an input in the party’s response.

The BJP also uses WhatsApp for internal communication, said Goenka.

Central and state media cells of the party articulate its position on various issues, and communicate this to communication groups in each parliamentary constituency.

“This helps us to have a uniform response on issues and our experience is that once the content is shared, it is accessed by our activists within five minutes,” said Goenka.

Even regional parties such as the Nationalist Congress Party (NCP), whose support base is largely rural, are not far behind in adopting such techniques to gain an edge.

“We have dedicated a team of eight volunteers who track trends on social media and after tracking the trends, in consultation with senior party leaders, we prepare our response and our activists from across the state who are technology-savvy are communicated the party official line and requested to join discussions on social media,” said NCP media and IT cell coordinator Ravikant Varpe.

Helping hand

Executives at advertising agencies say political parties have enlisted their services and also use tools such as Social Customer Relationship Management (Social CRM)—typically used by retailers, e-commerce firms, online travel sites and auto companies—to identify “influential voters” and engage with them.

They didn’t disclose the names of such parties because of confidentiality agreements.

Some political parties also use Klout—an app that uses social media analytics to generate a number between 1 and 100 that represents a person’s social media influence.

Goenka insisted that the BJP does not take the help of any professional agency for data analytics.

“We don’t believe in paid campaigns. Why should we employ professional agencies when we have a clear advantage over all our rivals in the tech and cyber world?,” he asked, adding that the BJP has three teams comprising about 70 professionals from top IT and consulting firms, who have taken a sabbatical from their jobs to help the party with data analytics.

The NCP acknowledged that it had enlisted Mumbai-based advertising agency Driving Minds to help in the task. Varpe declined comment on the fee paid to the agency.

Hareesh Tibrewala, founder and joint chief executive officer (CEO) of digital marketing firm SocialWavelength.com, said a social media listening tool such as Radian6 (owned by Salesforce.com and licensed out by SocialWavelength) could cost between Rs.5 lakh and Rs.50 lakh “depending on the functionality one wants”.

According to him, political parties customize such tools to “look at conversation density—in other words, where the chatter is—so that they can push content accordingly”.

Political parties seem to prefer using open-source software frameworks such as Hadoop for storage and large-scale processing of data.

According to Rajdeep Choudhary, consulting lead at analytics solutions provider Teradata India Pvt. Ltd, the use of analytics by political parties is all about increasing engagement with voters by tapping into the math of “online and social media conversations, among other things”.

Big spending

“For an Obama type of campaign, one will need a massively parallel processing system,” he said. “Today, we have such systems that can process billions of events (conversations, etc) in minutes when earlier these used to take days.”

Political parties are willing to pay the price for these new digital tools.

According to Pitch Madison Media Advertising Outlook 2014 estimates, released on 19 February, the Indian advertising market is forecast to grow 16.8% in 2014 to Rs.37,216 crore, mostly because of spending on general elections, which will be held in nine phases starting on 7 April.

The digital medium, it said, is forecast to contribute Rs.3,950 crore in 2014.

But what about the returns?

According to an 8 October report by the Internet and Mobile Association of India (Iamai) and market research firm IMRB, there could be a vote swing of 3-4% in 24 states where Internet users are sizeable, and the vote swingers could be predominantly young men and non-working women whose affinity for social media is high.

India has an Internet user base of 200 million, (a majority of whom access the Internet on their feature phones and have a presence on Facebook, Twitter and LinkedIn).

“This is an incredibly large population in terms of engagement, interaction and response when compared to traditional media sources,” said Kapil Gupta, founder and CEO of OMLogic Consulting.

According to the Iamai report, political parties have set apart around 2-5% of their election budgets for spending on social media. Recent media reports have put the media spending by the Congress and the BJP during elections at about Rs.500 crore and Rs.400 crore, respectively.

The spending makes sense. According to Obama’s 2012 campaign manager Jim Messina, it was Big Data and social media technologies that provided the incumbent’s campaign an edge against the Mitt Romney camp in the election.

On 7 June 2013, Messina told Datanami—a site owned by Tabor Communications Inc.—that the 2012 campaign spent $1 billion “amassing and using a Big Data-driven analytical system that was also hooked into the campaign’s social media activities. Over the course of 14 months, the campaign ran a total of 62,000 simulations of likely voter behaviour based on that data”.

Such techniques can also help parties get to know voters better.

Chandrabhanu Pattajoshi, business head at digital marketing data provider Precision Match, said that companies and political parties are analysing the searching, surfing and clicking habits of users to classify them into segments such as tech, auto or political enthusiasts even as “some overlap in data segmentation is unavoidable”.