Movie exhibitors, particularly theater companies, face similar problems as delicatessen owners in managing “shelf space” for movies — making room for new products, removing old goods before they spoil and balancing demand to reduce the amount of waste or lost income. Yet theater owners are overwhelmed by the number of films released each year, pressure from their distributors, a limited number of screens on which to display them and the uncertain, quickly changing tastes of moviegoers. Complex agreements dictate which films may be shown on specific dates for what periods of time and which must be withdrawn. A more optimal schedule might be arranged by using a mathematical model called SilverScreener, argues Jehoshua Eliashberg of the Wharton School at the University of Pennsylvania, creator of the model with Sanjiv Swami and Charles Weinberg of the University of British Columbia.

Hollywood’s major studios turn out more than 450 feature films a year. Although the U.S. and Canada have 27,000 theater screens, many films go out of rotation before they reach their entire audience. Many films do not recoup their production costs even after accounting for overseas markets, cable and television sales or video rentals. During peak months — May through September — major studios planned 59 films in 1997, up from 49 the previous year.

“Today the studios recognize the scarcity of shelf space but individual companies are finding it harder to find a screen for the films they want,” Eliashberg said. “Exhibitors have to make a master plan three or four months before a movie is even available. At that time, they can’t predict how long it will run. The industry is just too dynamic and complex.”

One reason is the agreements studios require for exhibitors that agree to show their films. That commitment period ranges from two to 10 weeks, depending on the theater owner, its bargaining power and the marketability of a particular film. Some contracts require a minimum run even when there is no consumer demand. Another unusual feature is the split of box office income, which is divided in ratios such as 70/30 in favor of studios during the first few weeks of release, but gradually shift greater income to the exhibitors with ratios reversed to 35/65 by the end of a run. Although attendance typically declines the longer a movie plays, the theater gets a larger share of the box office gross. Complicating that equation is the loyalty that can be rewarded when studios rely on commitments from national exhibitors.

“The entire industry is built on the relationship between studios and exhibitors,” he says. “Every Monday, exhibitors have to decide which films will be shown on the following Friday. And central offices or regional managers may not know about local preferences, likes and dislikes.”

That can lead to a less than ideal rotation of films. The SilverScreener model takes experience from earlier ticket sales then compares that history with attributes of upcoming films: genre, rating by the Motion Picture Association of America, sequel, stars and distributor. Those features often can forecast a film’s revenues. In an example that analyzed a six-screen theater in New York City for summer 1995, SilverScreener’s model would yield a 37.7% improvement in profit, when the exhibitor is limited to the movies it had already scheduled. The same theater could have seen a 121% increase in profit if it were able to schedule movies in rotation at that time, but not scheduled in the theater studied.

Instead of trying to please all the studios by showing as many films and withdrawing them quickly, Eliashberg said, theater chains need to be more selective in scheduling and run the movies for longer periods. It is possible that experienced theater managers, who know their area and the tastes of their audiences are already making the optimal choices, Eliashberg admits. But the increasing trend toward corporate ownership of theater chains makes this largely a business-to-business issue akin to the purchasing habits of retail chains like Wal-Mart, which must make contracts with large suppliers such as Procter & Gamble or Coca-Cola.

Another feature the model can evaluate is the cost of honoring or breaking agreements with distributors. Would it be more profitable to close a longer running film, in favor of a more popular, shorter run of a title from a rival distributor? One limitation of the research is that with new theaters with more than a dozen screens, it can be more difficult to determine the theater=s capacity. Theaters can switch films from one set of screens to another, and he plans a future study of Amicro-scheduling issues such as the effects of different screens on different days when audience sizes vary widely. Finally, with theater owners constructing screens at a rapid rate, the supply-demand imbalance may shift in the near future.