Vote for Analytics! How Predictive Analytics Impacts Elections
Can political parties actually leverage Big Data tools and data analytics solutions to reach out to probable voters? They are already doing so! This white paper explores the role of Big Data analytics in elections across the world. Elections can change the destinies of nations & the very direction in which countries are headed in. It’s important that both the candidates and the voters alike understand the impact of predictive analytics and Big Data can have on election results.
The First Big Data Election
The first election to use Big Data analytics was in 2012, at the US Presidential elections. A total of $7 billion was spent on the many electoral strategies and about 10% of this amount was dedicated to the online advertising efforts by both parties. Both Mitt Romney and Barack Obama had their own strategies; and obviously the winner of the election executed the plan better. Mitt Romney had a high profile ORCA to help in electoral campaign. ORCA is a sophisticated plan that brought together 30,000 volunteers and their web apps to monitor the poll activity on voting day. The volunteers were meant to study every activity and use the analysis to reach strategic decisions. It was a very smart plan, but execution? Not so much.
On the day of the elections, there was a flood of information, pouring in from all over the country. ORCA crashed for more than an hour; when it powered back, there were problems in password recovery and the volunteers stated they were not briefed on the procedure, but were handed a 60-page document a day before it all was to take place! The idea was very simple; to use the Big Data activity of voters that used smart phones, and connect it to the election. As mentioned the ARS report, the whole plan was a 7-month rushed job and was very unstable. The fact is, to use sophisticated tools and technology for a plan such as this, the plan needs more time, effort and support from the leadership.
Compare this strategy and execution to that of Barack Obama and you would be greatly impressed. Team Obama fully understood the gravity of what Big Data activity could result to. They were prepared for the difficulties and challenges way ahead of the electoral day - 18 months to be precise. They also focused all their data into one single repository. The data metrics were churned thousands of times, every night. Moreover their analytics department had more skilled resources larger than that of 2008 campaign. Now, wasn't that a contrast to the Romney plan?
How Big Data Made All the Difference
The biggest decision the winning political party made was to move away from the traditional approach to voters; which is to send just one broadcast message to the whole country, blanketing everyone, regardless of age, gender, location or any other key differentiator. Using Big Data, the winning candidate was able to get closer to the target audience, improving engagement and talking about the exact things that was important to the voter. A smarter approach is to send messages that is relevant to people, not just banal updates that do nothing but just document the campaigns. Big Data analytics empowers the political parties to get to know their voters on a personal level. Superior technology and sophisticated tools help find & understand their needs and the different ways the voters can be approached. This is a huge deal for the world of politics.
Smarter Targeting of Voters
Traditionally, political parties relied on that minority that supported them wholeheartedly and depended on them to voice their love on social platforms to build more awareness. Now, with Big Data, a significant number of voters who really ‘don’t know’ can be targeted. Let us call these don’t-knows, floating voters. Machine-based learning can help with smarter targeting of such floating voters. These voters can also be segregated into many other categories such as the section that can be swayed into making a decision, another section that is unsure and many others. Therefore, converting floating voters can get in votes that otherwise were never possible to get.
Data analytics tools can help understand which sections are more likely to choose and support a party. Instead of using location as a parameter, Big Data analytics can churn out a lot of details such as newspaper preferences, age, and education background and look up keywords in Facebook and Twitter feeds as well.
Different Channels to Reach the Voters
With Big Data, machine learning and data mining, political parties have an edge over others who continue focusing just on traditional models for their campaigns. This edge is simply connecting to voters via social media, an important channel. Data mining from social media platforms such as Facebook and Twitter helps parties understand the concerns, issues and emotions voters have towards their party and how it could be diverse within the different regions of the country.
It was such data that was used by India’s political parties to build their strategy, effectively target voters, recruit volunteers and improve their different channels to reach the voter at the doorstep to personalizing messages on social media accounts. Before we go there, let’s understand what data mining is in the first place.
Data Mining is “Sifting through very large amounts of data for useful information. It usesartificial intelligencetechniques, neural networks, and advanced statistical tools (such as cluster analysis) to reveal trends, patterns, and relationships, which might otherwise have remained undetected. In contrast to an expert system (which drawsinferences from the given data on the basis of a set of rules) data mining attempts to discover hidden rules underlying the data. Also called data surfing, it enables users to discover patterns and insights from large-scale repositories.
Political Parties Moving Away from Traditional Models
Being inspired by successful campaigns of Obama’s Big Data in elections, even political parties in the UK have slowly moved away from the traditional models in recent years. Both the labour and the conservative parties have invested heavily into their digital mediums; they’ve also hired the same digital advisers who were part of the Obama team!
Data can be used research yes, but analyzing data from public reactions to political parties, their campaigns, policies and even responses in critical situations needs to be done in in real-time. Data analytics plays a very important role in changing the end result of any campaign. It is no wonder that India’s very own Narendra Modi is considered to be one of the most technology and social media-savvy politicians in the world! Prime Minister Modi has no less than 10k followers on Twitter, 32 million likes on Facebook & 440 million views on Google+!
Modi belongs to the BJP party which won the 2014 elections in India with the help of open-source digital tools that put them directly in touch with their voters. Most of the metrics to do that were achieved using data mining and data analytics to dig into the plethora of social media activities. Even mobile users were taken into consideration. However, there is no denying a major number of voters had to be reached the traditional way - direct human contact.
Regardless, BJP made smart moves to get in touch with their potential voters using a combination of digital and traditional channels to recruit volunteers - both online and offline. Although mass media was the dedicated channel for the voters sure of voting for BJP, to reach the floating voters and even negative voters, communication reached out at the micro levels on the Internet, mobile and social media - apart from the traditional street campaigns.
How BJP Used Social Media to Influence Voter Behaviour
BJP’s IT team monitored the social media chatter that surrounded discussions over the Internet. Even if relevant hash-tags weren't necessarily used, volunteers still were able to send messages personally and responded to potential voters in real-time. All the insights and patterns were then used to create the political tactics and comprehensive vision of the political party.
One effective illustration of this was when Narendra Modi, almost 7 months before the election day, made a controversial statement of “prioritizing toilets, before temples". This was a great opportunity for the BJP’s digital team to get into action and monitor all the social chatter pouring as an aftermath of such a statement. The insights were almost in BJP’s favour.More than 45% of respondents agreed with this statement and these respondents turned out to be part of the floating voter segment! These were the same sentiments of the floating voters and they appreciated a political party’s effort for such a vision for India.
Of course, the traditionalist totally misinterpreted the statement, understanding it as a choice between toilets and temples. The point was about hygiene and sanitation and half of the chatter agreed with that too. Seizing the opportunity, the communications team converted this controversial statement to declare “Swachh Bharat” which translates to Hygienic India. With increasing number of debates surrounding this subject, eventually there was a positive sentiment towards it. As a result, the initial 45% support shot up to 68% within just 48 hours!
The Other Side of Using Predictive Analytics
The temptation to view predictive analytics as the magnet to draw voters can be very strong,given how numerous businesses benefit from it.Companies like Amazon use Big Data, data mining and machine learning to predict what the customer would buy next, and keep the stock within shipping locations in anticipation of the potential buy. It can help decision making, increasing revenue and even process or product development.
However, Big Data is not a magic problem-solver for everything. For a company to use analytics to predict and anticipate trends, must have access to data in large-scale. This can be affected by human behaviour can be easily affected by the weather or by relationships. A model that worked in the past may not work again because of change in human behaviour. A thorough understanding, sophisticated tools and support from upper level management is crucial to the success of any strategy that decides to use technology for Big Data analytics.
The Future of Data Analytics in Elections
The fight to get voters through technology has the advantage of two indispensable trends - rise of the young population and the advances in technology. In India alone, over a 100 million new voterswere added to the mix in 2014. A sizeable portion of this population will engage the internet in general and social media heavily, creating data that will be extremely useful for data analysts.
The young India is all about being urban, educated and optimistic and is geared with the latest smart phones & high-speed Internet connectivity. This will lead to a massive data burst. It’s only a matter of converting all that data into useful and insightful information to change the course of any business, or in this case, electoral outcomes.
A new dawn has come upon the way elections are fought. During the 2014 Indian elections, the expenditure on advertising campaigns shot from the 2009 figure of $83 million to$300 million, with digital marketing being one of the reasons. Experienced politicians who used their guts and instincts to make decisions, will now move towards technology to make fact-based decisions. Educated engineers, programmers and data scientists will enter electoral scenes, and profitably so, what with the requirement for talented professionals skilled in data analytics!
Using predictive analytics is not just about taking technological advantage to win the electoral battles. It’s more than that. It’s about focusing political efforts to plan and build their strategies based on real public sentiments. Politicians can now really be part of people’s lives every day. Advances in technology can bring forth the issues that really matter to the people. Thus Big Data & predictive analytics can take elections beyond political campaigns to bring real change and win-win situations for whole nations.
About AIMS Institutes
In today’s data-driven world, AIMS Institutes let students hone essential skills that will enable them to enter and make a mark in the Business Analytics arena.
Devoted to delivering the best possible education to aspirants since 1994, AIMS Institutes (NAAC accredited with ‘A’ grade) are amongst the best B-schools in India, and also offer courses in a host of disciplines such as Hospitality & Tourism, IT, Arts, Humanities and Commerce.
The latest addition to their impressive portfolio is the Executive PGDM in Business Analytics – an industry-specific programme, tailormade in association with IBM for working professionals. This globally-recognized certification can open up excellent opportunities in the Analytics industry – a sector that’s slated for unprecedented growth in the foreseeable future, with almost every major segment looking to make the most of its benefits.
The relevant curriculum developed by industry experts and hands-on training on up-to-the-minute tools and technologies clubbed together with case studies and projects-based evaluation makes this programme an ideal stepping stone for Big Data & data analytics aspirants.
Learn more about the EPGDM in Business Analytics programme at AIMS Institutes.