His pitch was simple: Take corporate America's love affair with learning everything it can about its customers, and its obsession with carving up the country into smaller and smaller clusters of like-minded consumers, and turn those trends into a political strategy. The Bush majority would be made up of thousands of groups of like-minded voters whom the campaign could reach with precisely the right message on the issues they considered most important.

At first, Rove and campaign manager Ken Mehlman had doubts about the potential of micro targeting, according to Bush pollster Matthew Dowd.

"I had to really sell Karl on it, and Ken to a degree," said Dowd, who said the skepticism was rooted in whether the investment in databases and computer modeling would yield better results than the traditional precinct-by-precinct targeting of likely supporters. "I told them it was going to a major expense on the front end to save money on the back end."

As a test, Gage was asked to produce targeted messages in several Pennsylvania judicial races in the fall of 2003. Why? The state offered a diverse mix of geography and ethnicity, and it almost certainly would be a battleground for both parties in 2004.

When the election was over, the Republican National Committee commissioned a poll to figure out whether Gage's suppositions about why people voted were accurate. Gage's models predicted voters' tendencies with 90 percent accuracy, according to Dowd, and Gage was hired to microtarget the 16 or so battleground states in the 2004 election.

It wasn't long before this new, more sophisticated form of data mining became part of the mythology surrounding Rove and his role as "the architect" of Bush's reelection. Its use in Ohio, in particular, was credited with unearthing Bush supporters and delivering the state and the election to him.

Now Gage is working for another Republican presidential candidate entranced by the possibilities of micro targeting  — Mitt Romney. A Harvard Business School graduate who went on to head Bain Capital, Romney has made a point of adapting modern business techniques to politics, and it was in his successful 2002 campaign to be governor of Massachusetts that Gage's methods were first tried.

"The governor believes in accountability, benchmarks and metrics," said Beth Myers, Romney's campaign manager, explaining his interest in micro targeting. "He believes in using data when it comes to making decisions."

Describing what he does, Gage, 57, sounds part marketer, part political strategist — and more than a little Big Brother. "Micro targeting  is trying to unravel your political DNA," he said. "The more information I have about you, the better."

The more information he has, the better he can group people into "target clusters" with names such as "Flag and Family Republicans" or "Tax and Terrorism Moderates." Once a person is defined, finding the right message from the campaign becomes fairly simple.

"'Flag and Family Republicans' might receive literature on a flag-burning amendment from its sponsor, while 'Tax and Terrorism Moderates' get an automated call from former New York mayor Rudy Giuliani talking about the war on terror, even if they lived right next door to one another," Alex Lundry, the senior research director of TargetPoint — the firm Gage founded in 2003 — wrote recently in Winning Campaigns magazine.

Some people are not convinced. Skeptics think that splicing the electorate into small subgroups does not tell a campaign anything it can't learn from a traditional poll.

"It's harder and harder to reach voters these days, so the desire to cut corners is understandable," said Steve Murphy, a Democratic media consultant and campaign manager for former congressman Richard A. Gephardt (D-Mo.) during the 2004 presidential campaign. "But it still comes down to shoe leather. I have NASCAR's Hot Pass on DirecTV, and I read the New York Post. What micro targeting  category does that put me in?"

And in a presidential primary, in which voters are far more homogenous than in a general election, can micro targeting find meaningful distinctions between groups? Gage and Romney are convinced that it can.

Using consumer data to predict buying behavior is nothing new in the business world. Bruce I. Newman, a professor at DePaul University and editor of the Journal of Political Marketing, said the term "micro targeting" began popping up in marketing textbooks in the 1960s, when the field of consumer behavior began gaining popularity.

Pat Caddell, pollster for Jimmy Carter, employed a rudimentary form of micro targeting during the 1976 presidential campaign when he set up a chart with issues on one axis and regions of the country on the other. Caddell used the chart to advise Carter on what issues to emphasize as he stumped across the nation.

Today, companies of every size use micro targeting on a "very regular basis" to make basic decisions about how to market and sell their products, Newman said. Also, whereas the political world has long copied the techniques of the business world, that dynamic is changing.

"What's beginning to happen now is the commercial side is looking at the political side," said Newman, asking such questions as "We would like to know what you did with George W. Bush in 2004."

Gage said when he pitched micro targeting to the Harvard MBAs advising Romney in his gubernatorial campaign, they were stunned the idea had never been used in politics. "You guys don't do this already?" they asked, according to Gage.

For Gage, using the same consumer information as corporate marketers to figure out voter behavior was a logical step. His career had been spent crunching numbers as a pollster, much of it with two pillars of the Republican survey research establishment — Robert M. Teeter and Fred Steeper.

By the 1990s, Gage was spending most of his time on corporate work. "I was pretty burned out" on politics, he acknowledged. But Gage had also begun to mull the rudimentary elements of political micro targeting.

Working with a few Michigan based operatives — direct-mail consultant Fred Wszolek; Michael Meyers, executive director of the state GOP; and Brent Seaborn, who is now director of strategy for Giuliani's presidential campaign — he came up with a methodology he called "super segmentation." Later, they borrowed the term "micro targeting."

Around that time, Michael Murphy, then Romney's campaign strategist, became intrigued by the high number of independent voters in Massachusetts, seeing them as the key to winning in a Democratic stronghold. He sought out Gage for help.

"I wanted to break the independent-voter file into target segments and Alex's approach was the best way to do it, so I reached out to Alex and we, along with Tagg Romney and Alex Dunn of the Romney staff, sort of invented micro targeting  in that race," Murphy said.

What did they find?

That a 32-year-old white Protestant woman with two children and a retired Catholic male engineer — while both independents — were driven by often contradictory issues, Murphy said. "Some independents are more base Republican — like, some are pure fiscal voters, some are focused on education," he added.

All of this seems somewhat straightforward — after all, anyone with even a passing interest in politics knows that a mother of two and a retired widower are probably motivated by different issues.

Wszolek, the Michigan-based direct-mail consultant, has known Gage since 1984 and worked closely with him to fine-tune a theory of political micro targeting. Wszolek acknowledged that "what you're doing is putting a very fine point on the obvious."

But, he added, the key insight of political micro targeting  is rather than simply determining whether married men are more likely than unmarried women to support a candidate, a campaign can identify segments within larger demographic groups and tailor messages down to the household level — an extraordinary amount of precision helps turn a guessing game into a series of targeted strikes. If television advertising is painting with broad brush strokes, micro targeting  is political pointillism.

The first step in doing this is conducting a large survey of voters. By matching up their political views with detailed information about their consumer habits, a model is established that can be applied to the population as a whole.

A campaign would then know which issues are important to an unmarried woman who subscribes to Outside magazine and is a frequent flier, and how they are different from issues important to an unmarried woman who has two grown children, uses corrective lenses and is an AARP member — even if they are next-door neighbors.

"A lot of people were skeptical that a big sample would tell you anything different than a small sample," Wszolek said. "What we found with large-sample research is you see something totally different. That was Alex's central revelation."

It took TargetPoint six months — and cost the Bush campaign $3.25 million — to conduct surveys, overlay them with thousands of data points and break down the electorate into unique segments.

To Mehlman, the information meant the campaign was fundamentally different from the one before. "In 2000, we very broadly talked to people on broad issues," he said. "In 2004, instead of talking about what we thought was most important, we talked about what the voters thought was most important."

In Ohio, the key battleground of the 2004 campaign, Gage's micro targeting  showed that black voters — who had traditionally not been drawn to the GOP — wanted to hear candidates talk about education and health care. As a result, they received a series of contacts — direct mail and phone calls, primarily — emphasizing Bush's accomplishments on just those two issues. It was a much different message than the president's broader attempt to cast the election as a choice between staying the course in Iraq and the anti-terrorism effort or switching teams in midstream.

It worked. Nationwide, Bush won 11 percent of the black vote, a two-point increase from 2000; in Ohio, he won 16 percent, an improvement of seven percentage points. Bush won Ohio by 118,601 votes, or approximately 2 percent of the more than 5.6 million votes cast for the two major-party nominees.

In New Mexico, Gage's micro targeting  discovered a segment of 19,000 lower- and middle-class, middle-aged Hispanic women whose children attended public schools. That group was strongly resistant to Republican candidates — just one in five said they would back a GOP candidate — but about half said they would back Bush. Why? Because 80 percent of the group were strongly supportive of his No Child Left Behind education legislation.

The Bush campaign made a targeted strike with a message focused on his push for testing and standards in public schools. It focused particularly on the 6,000 women in the group who were all but certain to vote. Again, the goal was not to win Hispanics or even Hispanic women but rather to minimize the Bush campaign's losses in this particular demographic.

On Election Day 2004, Bush carried New Mexico by 5,988 votes. It was the only state that he lost in 2000 and won four years later.

In response to Gage's success, Democrats have made their own attempts at micro targeting , and they think they have caught up in the technology, if not the organization, needed to apply it. Republicans worked to hone their micro targeting  techniques under the single roof of the RNC-Bush campaign, but Democrats have been experimenting with a patchwork of smaller, less centralized efforts, according to Ken Strasma, founder of Strategic Telemetry, a Democratic firm.

Gage doesn't sound worried. What he does is as much art as science, and he never stops tinkering with his models. "Part of the challenge is to constantly attack what you're doing and try to do it better," he said.

Eighteen months ago, Gage made the trip up to Boston to meet with Myers. At a Beacon Hill restaurant, the two old friends chatted about Romney's potential as a presidential candidate and micro targeting's ability to help deliver him the GOP nomination.

Over the next months, Gage and Myers talked from time to time about how micro targeting  might best be used to make a difference in a presidential primary. One Saturday last fall, Myers, Gage and Will Feltus, a member of National Media Inc., the company that handles Romney's advertising, gathered for a final bull session.

At issue was whether micro targeting  could find meaningful — and measurable — differences in a primary electorate that was Republican to begin with and similar in its demographic and ideological traits. After hashing out the details on maps and graphs, Myers and the rest of the Romney team reached a decision.

"The question was whether you could differentiate between the eight kinds of chocolate," she said. "I became convinced that the power of micro targeting  was enhanced by segregating a generally homogenous universe."

Myers's conversation, like that of her candidate, is more from the business world than from the political one. She likes to talk about the "seamless web" that allows the campaign not only to "see at any given time what the left hand is doing" but also to use the "right hand to tell us what impact it has."

But the Romney campaign is decidedly circumspect when it comes to divulging details of exactly what Gage and his team are doing, other than to say the process of interviewing individuals has begun in Iowa.

Romney communications director Matt Rhoades is only slightly more specific when asked about the campaign's plans for micro targeting. "Our micro targeting  strategy is tied to the calendar, and we have developed micro targeting  models in Iowa," he said.

Developing that strategy has placed Gage in a central role in the campaign. Myers describes Gage as its "strategic orchestra leader" — he oversees polling, media and online operations and works to ensure every part of the Romney operation is working in concert.