Shaken & Stirred - Influential Brand Profiling and Positioning

Big data to predict 4th Oscar winner in a row? … Exponential

The company which successfully predicted the Best Picture Oscar winner in 2013, 2014 and 2015 – by analysing the web surfing habits of people with similar demographics to Oscar voters – today released its prediction ahead of Sunday’s 88th Academy Awards.

Digital advertising specialist Exponential predicts The Big Short will win the Best Picture Oscar, narrowly beating SpotlightThe Revenant and Room are the next most likely.

Oscar winner rankings

How’s it done?
Exponential successfully predicts Oscar winners using a technique called “look-a-like modelling” which analyses the behaviour of 32,000 people working in the Los Angeles Film industry with similar characteristics to Oscar voters. They’re from a very narrow demographic group say the Los Angeles Times – 94% Caucasian, 77% male, average age 62.

“We know the typical Oscar voter is a frequent traveller, invests heavily in home theatre systems, follows tennis and baseball, is concerned about privacy and Social Security, buys expensive watches, and drives a European luxury car,” explains Bryan Melmed, Exponential’s VP of Insights. “Thus, the film interests of people with similar interests give us a strong clue as to where votes would go.”

Why the others came up short
These were the main areas in which fans of these films differed to the typical Oscar voter:

  • Brooklyn: not big film fans, prefer TV – 80x more likely than average to watch Downtown Abbey
  • Bridge of Spies: fan base too conservative and strong interest in armed forces, the defense industry and the FBI
  • Mad Max and The Martian: fans are too young and diverse
  • Room: audience doesn’t show a strong male skew, women make up a large portion
  • The Revenant: fan base is too diverse, conservative and religious

Melmed says: “This left The Big Short and Spotlight but it was incredibly difficult to predict the winner as these two have the strongest overlapping audiences, sharing 95% of the 10,000+ behavioural indicators we identified. However, The Big Short edged it on mirroring the key interests of Oscar voters.”

The lessons for marketers
“Whilst this is a bit of fun and mainly of benefit to people who fancy a wager, it provides marketers with an important illustration of the power of look-a-like modelling to predict consumer behaviour,” concludes Melmed.

“It identifies groups of people with shared interests which can be modelled against a brand’s customers to find new people who are most likely to convert. This improved targeting minimises wasted ad budgets and serves people more relevant ads, which reduces a motivation to block them.”