Big Data and Hillary Clinton: the myth of exact election forecasting

Dirk Ziems

Many people have been taken in by the image of a managed campaign. Elan Kriegel, master data analyst of the Clinton campaign, was presented by the American media as “Hillary’s invisible guiding hand” and the campaign’s most important mastermind.

What states are decisive, which cities Hillary visited, where Facebook ads were placed, which doors campaign helpers should knock on – all this was predetermined by Kriegel’s Big Data algorithms.

The American polling industry has cultivated the image of completely analysed opinion trends and virtually failsafe predictability. In dozens of representative weekly opinion surveys by the cable networks, the Gallups and Rasmussens and the meta-polling systems like RealClearPolitics building on these, every turn in public opinion was apparently made transparent, identifying the state of mind and relevant attitude drivers of every target subgroup.

Then the election came, and two myths emerged badly battered: the myth of the exact election forecasts, and the myth of the exact market management by Big Data.

What can we learn from all this from the marketing point of view?

Opinion dynamics are not precisely calculable.

The image of exact measurement and management is based on fundamentally flawed assumptions. Voters (the same goes for purchasers) move in a field of dynamic tensions. An unconscious psychological interplay of forces determines whether they go to vote or not, and whether they regard one or the other candidate as the lesser of two evils, and whether they answer honestly about their opinions or give socially desirable answers. If voter research puts the “Sunday question” (“Which candidate will you vote for?”), it is only measuring a forecast of the final result of the psychological play of forces. However, the forecast is uncertain, because from one week to the next the unconscious dynamics of opinion and moods can lead to different results.

Segmentation pigeon-holes mislead.

Pollsters and Big Data marketers base their statements and strategy extensively on segmentation labels such right, left, disadvantaged lower class, globalisation winners, soccer moms, successful or Latinos. The voters, however, seem to contradict these labels – very much like the hybrid consumers, who can no longer be clearly pigeon-holed. Against all forecasts, Trump got more votes among Latinos than Mitt Romney did in 2012. The majority of the low-income population voted for Clinton, not for Trump. Trump owed his electoral success to a white middle class group with relatively good incomes, who had benefited on balance from the globalisation of trade that Trump was thundering against.

Moods trump facts,

and Trump’s campaign was known to have little interest in facts. In this, it adopted an established principle of product marketing: “Facts don’t matter, perception is king”. Product marketers have always known how to explain complex products in particular with simple images. Household cleaner advertising leaves out any mention of tensides and chemical processes. Instead, little men in the lavatory cleaner are shown chiselling away hard water deposits. The customer does not want to know what the urea in the catalytic converter does. The diesel engine has the touch of blue motion and clean energy.

Trump has transferred the advertising fairy story mechanism consistently to the political arena. “We’ll build a wall and Mexico will pay for it!” Trump presents simple solutions, implying feasibility and ability to act – completely divorced from the complexity of real situations. Moods trump facts,

Pull beats push.

Clinton’s well-oiled campaign machine ultimately failed. Rolling out superstars like Beyoncé and Jay-Z, collecting addresses of concert goers and contacting them on election day with a reminder to vote – that seems to be a clever strategy. Unfortunately, the Millennials only came to the concert for the stars, and not for Hillary, and Detroit did not mobilise enough votes to secure Michigan for Hillary. Once again, we see that pull is generally a better strategy than push in marketing.

Authenticity beats marketing.

Appearing genuine and credible creates real closeness to the voter. In her campaign, Clinton ultimately remained the uber-perfect performing robot politician at every moment. Her weakness of never admitting weaknesses and concealing them compulsively left her with the email scandal, made her hide her pneumonia, and ultimately cost her credibility and emotional appeal. By contrast, Trump was always Trump – in German terms, a figure like Dieter Bohlen with his salesman’s patter, running for Chancellor. When Trump discriminated against women in the notorious Access Hollywood interview, people were outraged and shocked. But in contrast to Hillary, where people always suspected another side hidden behind the campaign façade, Trump was always the undisguised genuine Trump. His edgy otherness, compared to the streamlined standard politician, meant that The Donald was always a spectacle with talk value. And talk value – whether positive or negative – is an essential success factor in our era of the twitter-attention economy.

An uncertain forecast:

what will happen when the campaign Trump takes over real responsibility as President Trump is uncertain. This uncertainty in turn makes markets uneasy and destabilises established political systems. Outcome: uncertain.

About the author: Dirk Ziems is Co-Founder and Shareholder at concept m, a market research and consulting institute which uses depth psychological methods to investigate consumer motivations. One key activity is global research into automotive markets and brands in Europe, China and the USA.

The article appeared on 10 November 2016 in absatzwirtschaft.

If you have questions, feel free to contact Rochus.winkler@test.local

Dirk Ziems
Dirk Ziems ist Experte für tiefenpsychologisches Marketing und berät auf Basis von Markt-, Medien- und Kulturforschung weltweit Unternehmen und Konzerne in zahlreichen Branchen und Ländern. Als Mitbegründer der Global Research Boutique Concept M und der Marketingberatung Flying Elephant begleitet er Themen wie die Adaption von Erfolgsprodukten in neuen kulturellen Kontexten, das tiefe Verständnis neuer Konsumgenerationen in China und USA, die Transformation der Werbekommunikation in der neuen digitalen Medienwelt oder die Neuorientierung der Brands in Post-Corona-Zeiten. Dirk Ziems ist auch als Gastdozent an verschiedenen Universitäten tätig.

Kontaktieren Sie uns!

envelopephone-handsetConsent Management Platform von Real Cookie Banner