(FORTUNE Magazine) – For the past year or so, running a broadcast network has meant fighting a losing battle against declining ratings, rising costs, and robust competition from cable. And now, broadcasters face a new threat: sophisticated media-buying software programs called optimizers that are suddenly the rage on Madison Avenue.

Optimizers crunch ratings, viewing patterns, and commercial prices by using newly available A.C. Nielsen data, which now track both individual viewing patterns and aggregate audiences. The optimizers then generate media plans that, insiders say, tend to favor cost-efficient cable and syndicated programs over the pricier fare on ABC, CBS, NBC, and Fox. For example: Instead of paying $400,000 for a spot on Friends, a media buyer can reach as many unduplicated viewers--and save as much as $100,000--by scattering spots across a dozen or more cable networks or syndicated shows.

"Optimizers threaten to reduce the power and influence the broadcast networks have over the national TV marketplace," says Jack Myers, who publishes The Myers Report, an industry newsletter. That's a worry for the Big Four, which collectively hope to sell at least $6 billion in prime-time advertising in the up-front buying season that begins in May.

Fundamentally, the optimizers are the best way of comparing thousands of media-buying options that are created as audiences are dispersed across half a dozen broadcast networks and 35 cable networks tracked by Nielsen. The simplest optimizers can be bought off the shelf for as much as $300,000. "Some clients want tonnage. This will drive them more into cable and syndication," says Page Thompson, president of Optimum Media, a DDB Needham unit that buys television time for such clients as McDonald's, General Mills, and Universal Studios.

But most big agencies, including Optimum, have developed their own proprietary optimizers that factor in other variables. TN Media, a division of Bozell, recently developed a secret formula to account for a program's holding power (how much of a show people watch), its viewer loyalty (how often they watch), and its attentiveness factor (how closely they watch). Media That Works, a Cincinnati-based firm, has an optimizer that tracks ad clutter, subtracting points for shows loaded with commercials.

By using such refined data, media buyers want to uncover hidden programming gems. "What's happening for the first time is buyers are starting to analyze television using different benchmarks from the networks," says veteran ad buyer Steve Sternberg, director of broadcast research for TN Media. "That should give us an edge."

But the more nuanced optimizers also should tilt the balance back toward the biggest broadcast hits that viewers follow each week, like ER and The X-Files. The losers will be overpriced shows in the bottom half of the Nielsens. "Ultimately, the top-tier shows will become more expensive and the bottom-tier shows will become more commoditized," says Jon Nesvig, Fox's president of ad sales.

While no one knows for sure whether optimizers will have a major impact this spring, everyone agrees they're here to stay. The next generation of optimizers, for example, will seek to match viewing information with purchasing data. "Is the purchaser of canned tuna more or less likely to watch Peter Jennings on World News Tonight?" asks Susan Benzinger, chief strategic officer at Media That Works, whose clients include Starkist Tuna. Eventually, buyers would like to determine with some exactitude just how their spending on commercials translates into sales. For two industries--TV and advertising--used to operating on gut instinct, that's a radical idea.