Until this month, not many people knew that Viagra use was rising faster among men under 45 than among those over 45. Or that women in America buy more cars than men. Or that alcohol consumption has fallen faster over the past 40 years in France than anywhere else in the world.

Nor was it widely known that a million British couples live under separate roofs even though they claim to be in long-term monogamous relationships, or that fully 1 per cent of Californians aged 16 to 22 want to be snipers — trained killers in uniform — when they grow up.

But whatever it all means, the word is getting out. It is spreading because of a book by Hillary Clinton’s chief strategist, the pollster Mark Penn, that gives the clearest indication yet of how he plans to put the first woman in the White House.

Penn, a 52-year-old “numbers junkie” and multimillionaire, has been called “Schlumbo” for his dishevelled style, and the Democrats’ Karl Rove (in deference to President Bush’s former strategy guru) for identifying the so-called Soccer Moms who helped to win the 1996 US election for Mrs Clinton’s husband.

Most of Penn’s data is American, but much of his research applies to Britain, where he is a frequent visitor. Labour under Tony Blair paid him £503,000 to bombard voters with questions and slogans during the 2005 election. One result was the “Forward Not Back” poster campaign. Another, arguably, was Mr Blair’s third victory.

Will Gordon Brown be his next client? Penn was in London this week, declining to answer the question directly but holding meetings with Labour insiders that he said “have been going well so far”. And the Prime Minister is rumoured to have taken an advance copy of his book on holiday.

The book, called Microtrends, is a dizzying assemblage of 75 “new” demographic groups plucked from mountains of polling data, arranged by theme and given labels from which sober sociologists might recoil. Meet the Caffeine Crazies (if you dare — a can of their favourite energy drink contains eight times as much caffeine as one of Coke), the Jew-Lovers and the Sun-Haters, the New Luddites and the two million Late-Breaking Gays, who married then came out.

These groups can be called groups only because of the way that new technology has brought them together. But that technology also empowers them: Penn’s central claim is that each microtrend could, by itself, swing an election and change the world — if it hasn’t already.

This is not the “big idea” that politicians conventionally crave. It is an argument that only by using endless number-crunching to grasp hundreds of little ideas, many of them counter-intuitive, can any candidate hope to win a major election; society has become too complex and individualised to understand in any other way.

As Penn puts it: “America and the world are being pulled apart by an intricate maze of choices, accumulating in small, under-the-radar forces that can involve as little as 1 per cent of the population, but which are powerfully shaping our society . . . Small is the new big.”

The idea is not entirely new. Karl Rove famously targeted highly specific constituencies within the Republican base and helped to win two elections for President Bush by ensuring at all costs that these voters’ concerns were answered and that they made it to polling stations.

Nor is everyone convinced that Penn’s “small” was big enough for Labour in 2005. One backbencher, Peter Kilfoyle, complained after the election that “we lost a lot of votes and we lost a lot of seats . . . If the bottom line is payment by results, I have to ask if he was worth it”.

But Penn retains the respect of his clients and his peers, and his book, already a bestseller in the US, is eagerly awaited here. “What’s not knowable is how many votes [Labour] would have won if they hadn’t had his help,” says Andrew Cooper, director of Populus and pollster for The Times. “I imagine it would have had many fewer. I’ll certainly read the book with interest.”

Even as Microtrends falls out of briefcases across Washington, it has been criticised for Penn’s refusal to tease out any broad conclusions from his disparate research. If he has any, he is presumably keeping them up his sleeve for Mrs Clinton.

Either way, this much is clear: she will be trying harder than any candidate in history to be all things to all people.

Conventional wisdom, Penn likes to point out, is often simply wrong — but only diligent data-mining and deep respect for numbers will reveal when and how it is wrong.

So what are the demographic groups that Penn says the politicians must woo?

New Luddites
You thought that internet use was heading in an endless upward spiral? It may be — but not one that is driving non-users to extinction. Rejecters of the internet and all it stands for, as distinct from those who had never used it at all, numbered 12 million in 2000 in the US and 15 million three years later. The New Luddites may be more cynical than avid web-users. They are probably lonelier, too.

New Geeks
New Geeks may use the internet to meet people online but they also meet them in real life. They are twice as likely as reluctant users of new technology to regard “hitting the town” as enjoyable entertainment and 60 per cent of them describe themselves as extrovert, compared with 49 per cent of US adults.

Militant Illegals
The first of two microtrends that Penn thinks could swing the US election, Militant Illegals are ineligible to vote but are related to the seven million legal Hispanic voters who might turn out for the first time for a Democratic candidate next year to express their anger at the failure of President Bush’s immigration reforms.

Older New Dads
The new Soccer Moms? They include those fathering children when aged 40-44, who have trebled in number since 1975. The number having vasectomy reversals has shot up 40 per cent since 1999. As for the Soccer Moms, they’re ten years older than when Penn first made them famous. Many of them have taken up unusual sports. They are now Archery Moms.

Living-Apart-Togethers
Penn gives microtrend status to the million British couples living under separate roofs even though they are in long-term monogamous relationships. They account for three in 20 people aged 16 to 59.

Sun-haters
These first emerged in Australia — but the phenomenon of militant mothers demanding rules to enforce the use of sunscreen on their children at school is spreading worldwide. In a vintage example of his method, Penn adds together America’s dermatologists, their families, the families of victims of skin cancer, current skin cancer sufferers and an existing demographic known as “America’s Generally Cautious” to produce a two-million-strong constituency of “Sun-Hating, fedora-in-August-wearing Americans”. Look out for Mrs Clinton bearing sunscreen in Florida next year.

The Upscale Tattooed
The hunt for microtrends is not constrained by categories. It just expands them or creates new ones as new data spikes appear. Thus, under the “Looks and Fashion” heading, we find wealthy clients of America’s exploding profusion tattoo parlours — up from 300 to up to 1,500 in 20 years, with the biggest customer group by income earning over $75,000 a year. And seven million of those with tattoos are Republicans. Swing voters, perhaps?

Moderate Muslims
Under “Race and Religion”, Penn lists one microtrend on which global security already depend. Adding together non-mosque-goers, mosque-goers with a “flexible” attitude to their religious practice and “swing” Muslims still deciding between flexibility and a more conservative approach, he counts at least three million moderate Muslims. But will they stay that way or retreat into radicalism like many of their coreligionists in Europe? And are they among the four in ten Americans he labels Jew-Lovers for being very or somewhat interested in dating or marrying a Jewish person?

The book is fascinating and Penn’s questions cannot be ignored by his clients or their rivals. He may even be right that catering to myriad distinct constituencies is the only scientific way to win elections. But whether anyone can fight a microtrend campaign and then lead effectively is another matter. “If you simply defer to the numbers, you end up in a mad world,” Andrew Cooper says. “Unless there is a unifying theme, the individual messages won’t work.”

Times Recommends

• Hollywood stars in black and white
• Aga sales fall victim to economic worries
• Ageing actors are at a premium