Many around the world held their breath on November 7th as Americans voted for their next president.
On November 8th the sun still rose and the day arrived. For some, the US presidential election result created a Christmas-morning-like magic; for others it felt like a first day back-to-school ‘tragic’. But aside from the result, there is another story that has emerged.
At the center stage of this 2012 election were two players - an old, familiar one and a fresh, new face.
I’m not talking about the candidates named ‘Barack’ or ‘Mitt’, but rather two ‘spirits’ - one being the familiar, enduring American spirit of democracy (did I mention I’m an American?) and a new ‘spirited’ player, the ‘data-mining dynamo’.
Data mining - a reliable running mate
To fully understand bias and choice, perceptions and on-message triggers, preferences and awareness, market researchers have long endorsed the value of mining data. What we researchers have always known is the value data plays in quantifying the impact and size of an opportunity. Most of all, garnering insights from data and translating them into actionable strategies help our clients achieve their goals. Now the data-mining darling is proving her worth to the political playing field.
‘In politics, the era of big data has arrived’
In TIME magazine’s article on ‘Inside the Secret World of Quants and Data Crunchers Who Helped Obama Win’, Michael Scherer sheds light on the techies behind the campaign. Even though Obama became the first re-elected president in more than 100 years whose margin of the vote shrunk from his initial win, he secured the win by leveraging data in an unprecedented way!
Number crunchers take center stage!
Campaign manager, Jim Messina had promised a ‘new’ campaign. By hiring a huge analytics department (five times larger than in 2008) and an official chief data guru, Rayid Ghani, they set out determined to measure everything in the campaign. The effect unleashed a data powerhouse that helped secure a win in an election that produced a near tie, despite $6billion dollars of campaign spending!
Three things Obama’s campaign did effectively by harnessing their data:
1. Put insight to use
2. Traded many databases for a single, mega-smart one
3. Tested to find the gems.
George Clooney and West Coast females aged 40-49 - what’s the connection?
Obama’s number crunchers noticed that there was a strong correlation between a particular demographic group and their willingness to part ways with their money when it came to a chance to dine with George Clooney, while contributing to the fundraising efforts.
The 40-49 year-old west coast females were significantly swept away by this notion. By putting that insight to use, the campaign targeted other celebrities and the “Dinner with Barack” contest was born.
Too much data and no story to tell
Data has no value unless powerful decisions result from it. That’s true of election campaigns or storytelling consumer research. Individual sources of data only offer one perspective.
Harmonizing data together across numerous sources adds depth, texture and accuracy to the story of voters or purchasers. Democrats didn’t have a shortage of data in 2008. In fact, they exploited technology and social media in unprecedented ways. The problem of 2012 was that they had too many databases … and none of them talked to each other! Phone lists, ‘get-out-the-vote’ rolls and fundraising data were never harmonized.
Messina’s team set to work creating a mega-database where the information could ‘talk’ to each other for the first time. This allowed the techies to do predictive modelling to learn who would give, which messages were most effective with which audiences, and how fundraising would be most efficient overall.
Use data to test more data … whatever you do, find the gems
Using the ‘metafile’ to focus its fundraising efforts, the data technicians realized that cash raised online could be done through ‘intricate, metric-driven messaging’ campaigns.
Many emails sent to potential contributors were variations of different tests. Results got measured and tweaks were made. Adjusted messages were again sent and tested to learn which combinations would raise the most money. For instance, Michelle Obama’s emails performed best in the spring.
In many cases, these tested and targeted messages raised 10 times more than the ‘underperforming’ email campaigns.
In retrospect, it was an analyst’s election!
Top image : Google