Hollywood studios rely on tracking firms like Greek generals once enlisted oracles — to shine a light through the fog and predict an unknowable future.
To some, the most sophisticated data sets and intricate algorithms have proved to be about as scientific this month as the soothsayers of yore in determining who will emerge victorious at the box office. One by one, the weekends have ticked by only to see initial estimates of how a particular film will perform exposed as wildly over-inflated or overly conservative.
“Research is expensive and it’s wildly inaccurate,” griped a senior level distribution executive.
August kicked off with “Guardians of the Galaxy,” which had many projecting a debut in the $70 million range. Instead it put up $94.3 million. That was followed up by “Teenage Mutant Ninja Turtles,” which the most bullish firms had topping $50 million. It did $65.5 million of business.
Last weekend, both “The Expendables 3″ and “Let’s Be Cops” were looking at openings of more than $20 million, but both fell short with $15.8 million and $17.8 million, respectively.
This weekend’s indignities include “If I Stay,” which pulled in $16.4 million, but was expected to hit $20 million or more, and “Sin City: A Dame to Kill For,” which was pegged for a $15 million to $17 million debut, but came in at $6.5 million.
“You’ve got to listen to the conversation that’s going on and what’s being said about the movie,” said Phil Contrino, vice president and chief analyst at BoxOffice.com.
To their credit, tracking firms began peeling back their estimates on “Sin City: A Dame to Kill For” as the weekend approached and it became clear that fanboys were going to give this comicbook adaptation a pass. But it was too late.
Old habits die hard. Contrino argues that firms are often too eager to look for a historical comparison, matching up a current film with another from the same genre or release date to get their numbers.
“The old factors that came into play with traditional tracking, they don’t work any more,” said Contrino. “Things pop at the last minute now. The way people make decisions in terms of what they’re going to see is more in the moment.”
Part of the explosion in interest is being driven by Facebook, Twitter and other social media outlets — platforms that tracking firms often turn to for data, with varying degrees of success.
In the case of “The Expendables 3″ and “Sin City: A Dame to Kill For,” both films may have benefited from a high level of awareness, since audiences were familiar with their predecessors, but that knowledge didn’t translate into ticket buying. Dreadful reviews didn’t help, nor did concepts that appeared rickety by their second and third go-rounds.
There are also more films entering the marketplace this month than at the height of summer — three or four new wide releases each weekend as opposed to one or two. That invariably means that one film or more is going to eat into the box office of its competitors.
For studios, eager to massage the results of their films so they appear in line with or ahead of projections, this presents an unwelcome challenge. Even a film that does well, such as the low-budgeted “If I Stay,” can look like a disappointment when measured against expectations.
“The tracking has been remarkably unreliable,” said the distribution executive. “It certainly puts a lot of pressure on the studios and unfairly on the filmmakers when they’re this far off.”
Another studio executive said that traditional tracking firms such as NRG and Marketcast should only be viewed as a part of the picture, and that companies should also be pulling data from IMDB, Flixster, Twitter and Fandango’s pre-sales in assembling their predictions.
“If you’re still looking at just tracking, you’re going to be surprised most Sunday mornings,” said the executive.
That’s to say nothing of the tricky business of defining the audience for a particular film. Take the case of films such as “The Fault in Our Stars” and “Heaven is for Real,” two recent examples that proved difficult to peg down when it came to box office predictions because they appealed to certain religious and teenage subsets.
“Under-represented groups are not being accounted for accurately,” said Josh Lynn, president of Piedmont Media Research, a statistical analysis firm. “If there’s a passionate minority, whatever they may be, that really feels strongly about a film, it may not be adequately represented in the general population.”
Tracking firms say that surveying audiences is more about gauging the effectiveness of a marketing campaign than it is stating definitively how a film will perform.
“Tracking should point to predictiveness, but it’s not inherently a predictive model,” said Ben Carlson, president and co-creator of Fizziology, a tracking firm that relies on social media responses. “Tracking, in a lot of ways, is about a moment in time in terms of what people are thinking.”
Of course, media outlets seize on tracking as a measuring stick for a film, an opportunity to peel back the curtain on tightly guarded projects to discover which films will flop and which will soar, all backed up by numbers. But the numerical nature of the data can be deceiving.
As Paul Dergarabedian, senior media analyst at Rentrak, puts it, “Like astrology, tracking is hugely entertaining, and sometimes it’s right and sometimes it’s not.”