From...

January 6, 1999
Web posted at: 2:47 p.m. EDT (1447 GMT)

by Stewart Deck

(IDG) -- In 1997, MovieFone Inc. started to construct a data warehouse so that it could follow the historical calling patterns of people who phoned in to its free movie-listing service.

The company wanted to make sure it had plenty of phone lines set up to cover peak calling times, and it wanted to dig into some of the data about calls. Pretty standard stuff.

MORE COMPUTING INTELLIGENCE
	IDG.net home page
Computerworld's home page
Computerworld "Emmerce"
Get Media Grok and The Industry Standard Intelligencer delivered for free
Reviews & in-depth info at IDG.net
IDG.net's personal news page
Questions about computers? Let IDG.net's editors help you
Search IDG.net in 12 languages
Subscribe to IDG.net's free daily newsletter for IT leaders
News Radio
	Computerworld Minute
	Fusion audio primers

But when the system went live 11 months ago, researchers at the New York-based firm realized that the data they were seeing was even more useful. Data from MovieFone and its sister Web site MovieLink could help them quickly advise theater owners on how many screens a blockbuster movie should play on and even let them advise theater developers on picking hot locations.

Seem far-fetched? MovieFone covers more than 15,000 theaters in 38 cities. This year it will handle approximately 60 million calls and 110 million Web site user sessions. Each of those calls and site visits requests data about theaters and films in particular area codes, and sometimes they even conclude with a ticket sale.

"We built a warehouse because we needed to access this information dynamically and drill down into specific markets, specific theaters and specific movies," said Irwin Miller, vice president of computer systems development.

The warehouse was built on a Sybase Inc. database and business intelligence tools from Business Objects S.A. It uses Sybase IQ for query acceleration.

In addition to providing information about current movies, MovieFone also sells ads in the form of audio movie previews. It uses the warehouse to demonstrate to distributors how effectively those previews turn into ticket sales. Its researchers use the data to map historical patterns and model caller behavior.

Miller said MovieFone is working out how to properly add the Web site surfer behavior data from MovieLink to the warehouse. The researchers hope to have their hands on this data in 1999.

"Retail generally is a good user of data warehousing because so much of what they do is marketing-intensive," said Mitch Kramer, an analyst at Patricia Seybold Group in Boston. Predictive marketing is a pot of gold for retailers, he said.

But Miller said his group has just begun to exploit the warehouse's potential.

"Since we know how many people are seeking information and buying tickets to particular movies, we can now provide exhibitors and distributors with accurate data the day a movie opens," Miller said. If the calling pattern fits the "blockbuster" model, MovieFone can use the data (for a fee) to help exhibitors decide how many screenings to add.

The data may also be used in the future to figure out the best location for new theaters. "If someone wanted to open an art-house theater, they could come to us and ask for a demographic breakdown by ZIP code of people who've requested art films. We can give them ticket sales by ZIP code; they can then choose where to build that new theater for the best results," Miller said.

But one industry observer wondered how much more information distributors and exhibitors need. Distributors already know the good neighborhoods, and moviemakers have extensive demographic audience data from test screenings and focus groups, said Paul Marsh, an analyst at S.G. Cowen & Co. in New York.

"It seems like every company that does [data collection] always figures they can sell the data. If I were a theater company, I don't know if I'd be interested in buying his information," Marsh said.

Deck is a Computerworld senior writer. His E-mail address is stewart_deck@cw.com.