Pre-release audience tracking surveys, which have been honed into a precise science over decades, seem to be evolving into an inexact art, causing considerable consternation for Hollywood film marketers. Distributors use tracking surveys to decide how best to allocate the $3.5bn spent annually on paid media to advertise films in the US market. When this research is off, too much advertising may be directed at audience segments already sold on the film and not enough on consumers who need persuading.

Examples just keep piling up. In December, Alvin And The Chipmunks opened at $44.3m for its three-day weekend, about twice as much as tracking estimates. That could be excused since Alvin's family audience is likely to involve intermittent film-goers. I Am Legend did $77.2m, well above the mid-$50m expected, which was a narrower miss on a percentage basis though more of surprise given its youth/young-adult audiences tend to be monitored closely.

Earlier in December, the North American three-day opening of The Golden Compass weighed in at just $25.7m, which was below confidential tracking forecasts clustered between $30m-$40m. In October, The Heartbreak Kid mustered $14m for its opening weekend, short of forecasts of $15m-$20m.

Most misfires overestimate opening box office, but one of the widest misses was The Passion Of The Christ whose $83.8m premiere weekend in 2004 eclipsed tracking estimates of just $15m-$30m.

Still, distributors follow the survey results closely as they are mostly on target. But instances of misjudging the audience have been more common over the past few years.

'Consumers can tell you in a survey that they are 'definitely interested' in seeing a given movie and are being completely honest,' says Vincent Bruzzese, senior vice-president of the worldwide motion picture group at US-based OTX Research. 'But that doesn't mean they will actually go see the movie. Consumers have so many more entertainment options than years ago, with video games, multi-channel TV, watching DVDs on elaborate home-entertainment centres, DVR replay of recorded TV shows and surfing the web. It takes a lot more expertise to properly interpret tracking surveys today than 10 years ago when the only options might be two or three new movies.'

OTX is one of four research companies conducting movie tracking surveys on a year-round basis for US distributors. The data measures consumer awareness of a given film, the level of interest people have in seeing the film at the cinema, and how a film compares to other movies opening at the same time. A film is placed in surveys six weeks before premiere, when the marketing push is building and public awareness is hopefully growing. Just before opening, each survey issues a forecast for opening box office with a 15% margin of error.

Results are supposed to be confidential for the client film distributor, mainly the major studios. Indie distributors often lack the marketing budgets to be regular tracking subscribers and do not have the wide-release films for which tracking would be meaningful. Over the years, tracking data has been leaked regularly to exhibitors and the press, both of which can misread the incomplete snippets of data they receive, according to researchers.

For example, the surveys focus on heavy cinema-goers, since the inclusion of infrequent cinema-goers would only muddle results with the least important audience. So, it can be argued, the older, more infrequent movie-going audience that flocked to The Passion Of The Christ was never intended to be measured. Also, the closely measured youth audience that enjoys frothy comedies is not a good indicator for the entire genre, particularly off-beat comedies from the likes of Judd Apatow (Knocked Up and Talladega Nights: The Ballad Of Ricky Bobby), whose films follow a unique trajectory in tracking.

Break-Up breakdown

Adam Fogelson, Universal Pictures president of marketing and distribution, recalls that in 2006, some press stories asserted comedy The Break-Up was tracking poorly prior to release, and then leaped to a further conclusion this indicated female lead Jennifer Aniston could not carry a Hollywood film. 'Not only did those stories prove to be inaccurate after the fact, but they were also an incorrect reading of the basic tracking data when compared to both Vince (Vaughn) and Jennifer's prior successes,' says Fogelson. The Break-Up later grossed a blockbuster $118.9m in North America.

Fogelson, who says Hollywood still has faith in tracking data, is taking measures to improve reliability. Universal now subscribes to more than one survey, something that would not have been the case 10 years ago. Also, he is doubly careful in selecting past films for data to compare to new releases and is more alert to anecdotal signals in the marketplace, such as whether pre-release screenings draw packed audiences or are only half full.

Yet audiences still confound even the best-designed radar. With e-mail and social networking via the web, premiere-goers can recommend or slate a film to their friends almost as soon as they leave the theatre. So a film riding high or low on Friday night can have an abrupt reversal of fortune by Saturday morning. A favourable peer-to-peer blast is credited with helping Borat far exceed expectations from its second day of release.

Another new challenge in predicting the behaviour of consumers is simply being able to find them. For years, researchers only had to dial phone numbers randomly to assemble a representative panel, but call waiting, caller ID, call blocking, unlisted phone numbers and internet voice can now skew the audience samples. Teenagers are particularly elusive and unlikely to be found on the other end of any fixed-wire telephones.

'It's getting a lot harder to find the consumer,' says OTX Research CEO Shelley Zalis, whose firm uses the internet to locate and poll consumers. 'Mobile phones will soon become a strong alternative for reaching movie-goers on the go, especially with the latest innovations in mobile community technologies.'

[s19] Robert Marich is author of Marketing To Moviegoers: A Handbook (Focal Press, 2005).