Trump’s Election Victory at USA Presidential Election: Big Data vs the Job-To-Be-Done
The election of Donald Trump as 45th president of USA came as a huge surprise. Indeed, since Trump was an independent candidate for the Republican Party, up to the moment when he became the unexpected front-runner for the presidency against Hillary Clinton; polls were saying that he was going to lose. However, history has taken a different path and Donald Trump is the new president of U.S.A.
Why was everyone so surprised?
The most surreal aspect of the last presidential election was not Trump’s victory; but the fact that before the tabulation of votes, Hillary Clinton was absolutely confident of being the winner.
Moreover, the opinion leaders and the mass media agreed on the Hillary view; all their polls were supporting the idea that the Democratic candidate victory was a piece of cake. And so the 8th of November became a total debacle for the mass media and for the elites.
Indeed in the past two years more data has been created than in the entire previous history of the human race. (source: Jim Harris )
For this reason, it is common opinion that such astonishing quantity should allow data scientists to profile people in a way that it should be straightforward to forecast what everyone of us likes, wants, desires, does and votes. In this vision, the 58 million American citizens voting for Donald Trump were not contemplated.
The data-driven models that forecast the election outcomes, due to their nature, focus on the voters attributes. For example, according to these models, a black woman would vote for Hillary, whereas a white man would prefer Donald Trump. The data-driven models do not explore why someone prefers one candidate over another . This approach creates huge problems at the moment of describing the past with the aim of realizing a strategy for the future.
In fact, just after the Election Day, a competition started among political analysts in order to figure out who is the Trump voter. According to the data published in the most popular newspapers, the demographic groups formed by men, white people and white women voted in the same way. So the data-driven conclusion is that the Democratic Party lost the election because it did not have enough appeal among men and white people regardless of their gender. The drama is that this analysis could bring the Democratic Party to a larger disaster than the last election.
Indeed, this vision does not explain what has brought those demographic groups to choose Donald Trump as their future president. The raw data are giving us a misleading picture of the reality. A famous example of the professor Clayton Christensen will help to explain the distortion caused by data. The Harvard Business School professor often repeats that he is a 62 year-old educated man who everyday buys the New York Times. On the one hand, he fits perfectly the typical market segment targeted by New York Times. On the other hand, he does not buy the New York Times because he is a 62 year-old educated man; but since the NYT helps him stay informed, which is the reason why someone “hires” the newspaper: this is the so-called Job-To-Be-Done.
Avoiding to discover the Job-To-Be-Done of the NYT would lead to disastrous consequences for the newspaper.
First, the JTBD (helping stay informed) can move from one market segment to another, with the consequence of reducing the efficacy of NYT’s marketing efforts focusing only on a specific market segment and harming the future sales of NYT.
Second, data does not tell anything about people that want to stay informed, but that don’t buy the NYT. This no-consumption area is usually the best opportunity of growth and it is where the marketing actions have the highest returns on investment.
The recent history of NYTimes is a vivid proof of the JTBD’s power.
The newspaper of the Big Apple has been one of the Hillary Clinton’s biggest supporters, and it gave encouragement to the opinion that Donald Trump was unfit to be President. Moreover, the NYT incorrectly predicted the victory of Hillary Clinton. A typical linear logical view should have brought to the conclusion that NYT would have been strongly damaged in its business by Donald Trump’s victory, but it had not been the case. Unexpectedly, the New York Times subscription increased of tenfold, in comparison to the same period of 2015; a net gain of 132,000 new readers. This shocking growth is the effect of a business based on a JTBD. The unlikely election of Donald Trump, increases the need of being informed also among people who are not keen on politics (no-consumption area), and they automatically “hired” the NYT to accomplish the task.
In conclusion, the New York Times was wrong about the US election winner, but the fact that its business was build around a JTBD has transformed a threat into an opportunity.
What is the takeaway for Democratic Party?
In my humble opinion, the main takeaway for the DP is that it should not follow a specific demographic group before figuring out which is the right Job-To-Be-Done among American population.
For example, if it is taken for granted the hypothesis that Trump’s voters elected him in order to be heard by the “system”, a likely scenario is the one where Trump policies will be strongly in favour of white people working in the manufacturing sector. A probable consequence is that the disenfranchised and frustrated people will belong to totally different demographic groups in the next president election in 2020.
If the Democratic Party fights for winning particular “political” market segments, it will not obtain new votes whereas it will lose some. Instead, the DP should discover a deep and urgent Job-To-Be-Done required by USA citizens, and should integrate its political action to fulfil it. In this way, the Democratic Party will not run behind the voters, but the voters will come to the party.
Trump had clearly in his mind this point when he declared:
“my voters are so smart I could shoot somebody and I would not lose any voters”
Indeed, a company, an NGO or a political party built around a JTBD will obtain extremely faithful clients, supporters and voters.
To sum up, the lesson is that the data are misleading without good theories providing adequate lens through which see reality.