After the 2000 presidential campaign, strategists for US President George W. Bush came to a startling realization: Democrats watch more television than Republicans.
So by buying millions of dollars worth of television advertising time, Republicans were spending their money on audiences that tended to vote Democratic.
What to do? With the luxury of four years until the next election, the Bush team examined voters' television-viewing habits and cross-referenced them with surveys of voters' political and lifestyle preferences.
This led to an unusual step for a presidential campaign: It cut the proportion of money that it put into broadcast television and diverted more to niche cable channels and radio, where it could more precisely reach its target audience.
While advertisers of commercial products have been heading to cable for years, presidential campaigns have generally relied on the reach of broadcast television to try to influence the widest possible audience.
But the Bush team's micro-target strategy could change all that, making an enormous difference to cable channels and networks as they vie for the escalating amount of money in politics.
Candidates, political parties and independent groups at all levels of government spent at least US$1.6 billion on television commercials this year, more than double the US$771 million they spent in 2000, according to the nonpartisan Alliance for Better Campaigns. At least US$600 million of this year's total went to the presidential election alone, more than twice the amount spent in 2000.
Kenneth Goldstein, director of the Wisconsin Advertising Project, which analyzes political advertising, said that the Bush campaign's use of targeted data and its exploitation of cable was likely to be the beginning of a trend, particularly as advertisers pursue new ways of reaching consumers using technology like text-messaging and cellphones.
"We're talking very, very small effects, but we live in a time when small effects can be decisive," Goldstein said, citing Bush's 537-vote victory in Florida in 2000, which catapulted him into the Oval Office.
Donny Deutsch, the New York advertising expert, said that now, "the selling of a candidate is no different from the smart media buying for toothpaste and automobiles, especially as people fragment their media habits."
As the Bush team analyzed the data, stark differences between the viewing habits of Republicans and Democrats quickly emerged. The channels with the highest proportion of Democrats were Court TV and the Game Show Network; for Republicans, Speedvision and the Golf Channel.
During the week, Republicans switch off the tube earlier than Democrats do. (Republicans who stay up are more likely to tune in to Jay Leno, while Democrats flock to David Letterman.)
Such revelations convinced the Bush team to alter its media-buying strategy. In 2000, the campaign spent 95 percent of its media budget on network television; this year, that dropped to 70 percent.
The campaign spent no money on national cable channels in 2000; this year it spent US$20 million. It spent little on radio in 2000; this year it spent US$12 million, much of it going to religious, talk and country music programming.
"This year, we reached a wider audience of potential Republicans than we did in 2000," said Matthew Dowd, a top strategist for Bush's re-election campaign.
Democratic strategists working for Senator John Kerry's presidential campaign said they had much the same consumer data as the Bush team, but they stuck largely with broadcast television because that was where their viewers were.
"You're tying one hand behind your back if you're not using the most sophisticated tools possible," said Mark Mellman, a top polling and media strategist for Kerry.
The additional money in this election allowed the Republicans to experiment with a different media mix and to apply techniques used by advertisers of consumer products.
"Politics is a mass product: 50 percent of American adults `consume' the election," said Will Feltus, senior vice president for research and planning of National Media Inc, which bought media time for Bush's re-election campaign.
The most alarming realization for the Bush team, he said, was learning that Democrats watch more television.
"It's like Coke finding out that they can't get to their consumers on television as easily as Pepsi can," Feltus said.
The Republicans' data, compiled by Scarborough Research, a leading market research firm, showed that nationally, Democratic voters were 15 percent more likely on average to be watching television than Republican voters. The research did not investigate the reasons for the lopsided viewing, but some analysts surmised that it had to do with Republicans not trusting the broadcast networks and with more programming being aimed at women, who tend to vote Democratic.
Feltus said that the Bush campaign, which began analyzing the data shortly after Bush took office in 2001, ran test projects in 2002 in the Texas Senate race and in a Colorado congressional race. The data in Colorado revealed, among other things, which roads Republicans drove as they commuted to work, helping the Republicans determine where to place billboards.
This year, before the Democrats had even selected a presidential candidate, the Bush strategists were considering advertising in movie theaters and health clubs. The data showed that Democrats were more likely to go to the movies than Republicans, so they dropped that idea. But it also showed that health clubs were a good way to reach Republicans and swing voters ages 18 to 34. So the campaign bought time on a cable channel that goes into health clubs across the country. It had reams of data that were not of immediate practical value but that helped the campaign understand its voters: for example, Porsche owners were more likely to be Republican; Volvo owners, Democratic.
Evan Tracey, who analyzes political television advertising for the Campaign Media Analysis Group, said the Bush campaign helped solidify its base of Republicans early with targeted cable commercials. These commercials, he said, were filled with "images of people saying grace and talking about faith and being optimistic about America, but there was also a lot of negative on Kerry."
The Democrats said they used similar data, with help from a new group called Media Vote, based in Los Angeles, but came to different conclusions about how to use it. They focused strictly on the battleground states, buying local cable instead of national cable, but still relying mainly on local broadcast programs.
The Republicans mostly bought national cable channels instead of breaking down those purchases by market. This had unexpected benefits, like helping Bush in Hawaii, a reliably Democratic state that Republicans had not focused on.
In October, Bush was suddenly running strong there, a result of his presence on national cable, Democrats said. That forced the Democrats to buy advertising time in Hawaii and route party notables to the state to try to counter Bush's gains.
The data also yielded unexpected insights. One of the shows most popular with Republicans, especially Republican women ages 18 to 34, turned out to be Will & Grace, the sitcom about gay life in New York. As a result, while Bush was shoring up his conservative credentials by supporting a constitutional amendment against gay marriage, his advertising team was buying time on a program that celebrates gay culture.
The Bush team broadcast commercials 473 times on Will & Grace in markets across the country between Jan. 1 and Nov. 2, according to the Wisconsin project. (The Kerry campaign broadcast commercials 859 times on the show.)
Dowd said the campaign had not tailored its message to match the demographics of the Will & Grace audience or any other audience but rather wanted to reach more viewers who might vote Republican. Besides, he said, "people are interested in broad national messages."
Goldstein said Republicans did not customize their message because they had one basic point.
"If your message is `Kerry is bad,' you don't have to tailor it," he said.
The Democratic strategy was to focus on swing states and tailor the Kerry message to the market.
In Pennsylvania, for example, the campaign ran commercials in Pittsburgh with workers talking about their jobs; in more liberal Philadelphia, it ran commercials with Michael J. Fox, the actor who has Parkinson's disease, talking about expanding embryonic stem-cell research.
But such targeting did little to resolve the question of how much difference advertising makes in a presidential campaign, particularly when there is so much of it. The real force of political advertising may be felt when it is absent.
Bradley Perseke, a Democratic strategist who bought the television time for Kerry, said that Bush's get-out-the-vote effort probably made more of a difference in the election than his advertising, although if one candidate had not advertised at all, that candidate surely would have lost.
Dowd agreed up to a point.
"What is discussed in earned media is more important than what's on the paid media," he said of news versus advertising. "But if they are in concert and the message is consistent, it has a tremendous effect."