WHEN COLUMBIA LAUNCHED “THE FIFTH ELEMENT” and Warner Bros. debuted “Fathers’ Day” on May 9, industry pundits expected a close weekend box office race. Tracking studies gave the sci-fi actioner the edge with estimates between $14 million and $15 million. The Robin Williams-Billy Crystal antic comedy was just behind in the $11 million to $12 million range.
The final tallies saw the futuristic thriller exceed expectations with $17.2 million and the comic remake underperform at $8.7 million. Charitably, the predictors were 15%-20% off the mark.
The prior week’s premiering trio of Paramount’s “Breakdown,” New Line’s “Austin Powers: International Man of Mystery” and MGM’s “Warriors of Virtue” were even more askew. While the ranking order was correct, each of the trio had greater grossing potencies of between 30% and 40%.
So, the question is once again being asked in industry circles: Why can’t tracking surveys more accurately predict the box office?
“Sometimes they get it right and sometimes not,” observes producer Mike Medavoy. “My guess is that inaccuracies crop up because there’s not a large enough sample being polled.”
If true, one naturally wonders why polling strategies — virtually unchanged in a decade — haven’t been rethought. The science of tracking public interest has been refined to between a 1%-3% margin of error in political polling, while the film industry is surprised on an almost weekly basis by a new picture’s grossing power — or lack of it — when contrasted with what the trackers had predicted.
A MARKET RESEARCHER from outside the film industry says the effectiveness of tracking public opinion boils down to two simple factors: the method of collecting data and the quality of the predictive model. He says arriving at the latter shouldn’t take more than two years using a simple regression analysis process. Simply put, one devises the survey, collects data and after a period of perhaps three months compares polling results with actuals. Then, systematically elements of the survey are taken out of the poll. Eventually, the elements with the best predictive qualities are retained.
“There is too much weight being put on a paper poll. People forget that tracking is a tool not an oracle,” said Kathy Jones, exec VP of marketing at Universal Pictures. “There are plenty of things that are wrong with the tracking studies.
“For instance, they include new pictures with films that haven’t opened and others that are virtually offscreen. It also cuts off at 9 p.m. Thursday, so the impact of reviews can’t be incorporated.”
Jones endorses tracking but is mindful of its limitations. She senses that there are no quick fixes to the process because respondents just have too many choices to consider.
In Hollywood, Market Research is dominated by the National Research Group headed by Joe Farrell. NRG has exclusive or semi-exclusive contracts with all the majors and most of the leading second-tier companies. He has stated that his margin of error is a very respectable plus-or-minus 5%, excluding client misreadings. However, he steadfastly refuses to talk to the media in depth, pointing to a confidentiality clause in his contracts.
“NRG is great. It really set the foundation for industry research,” producer Steve Reuther said. “There’s no question that a lot of data is misread. The closer you get to opening the picture, the more people’s expectations become inflated.
“You can convince yourself that your film will do well in all sorts of ways, even if it’s not in the research.”
Reuther’s complaint about these studies is their lack of propriety. He and Jones decry an industry tendency to embrace the data as gospel. He says the habits of moviegoers are changing faster than the industry can absorb and that the same set of statistics from five years ago translates differently today.
Another recurrent criticism is the composition of NRG’s polling sample. A changing set of moviegoers are called three times a week. However, no one under the age of 12 is queried and that demo is a significant slice of the audience, particularly during the summer and holiday seasons.
“It’s not just kids’ films. It applies to any film with a niche appeal,” New Line marketing/distribution president Mitch Goldman said.
“We have a lot of pictures that are targeted to ethnic audiences and they just don’t track on NRG.”
PRODUCER POLLY PLATT said she feels for market researchers because often information is transmitted to them poorly, or they receive conflicting objectives from the parties.
“Artists by nature are not scientific,” Platt said. “We cling to superstitions and quantifying what comes from the gut is just going to be very confusing and divisive for the show and the business.”