MEDIA: Psst! So you want a guaranteed ratings hit? Just call us

Television's new gurus of market research claim that success is a formula. Meg Carter wonders about, ahem, creativity

Recipe for television success. Take one household name. Add a proven programme format. Blend in a popular idea. Commission research to prove viewers will watch. Serve up your programme proposal. Garnish with an audience guarantee. Easy.

So easy, indeed, that Angela Harbutt, research and development director of the TV airtime sales house TSMS, claims it could become standard practice for the industry within a year. "Whether or not a programme is commissioned will increasingly depend on audience guarantees," she says. Moreover, Ms Harbutt and her number-crunchers believe they know how to get those guarantees: she says research can accurately predict potential public interest in a topic, programme genre or even star.

TSMS already does this for a number of broadcasters, including Meridian and Anglia, assessing programmes' likely ratings performances by subject, genre and talent. It also evaluates scripts and can arrange "focus groups" - rounding up and interviewing members of the public. "Just as we study programmes and can identify common factors for success, we can take one step back and look at a script for strong characters and plot," Ms Harbutt says. "Even at script stage it's possible to identify danger zones: tired themes and the like. Research offers a solid programme-making tool, as increasingly channels commission by presenter or star, and the marketing of a programme hangs on that choice, too. Take Cadfael. Research persuaded the makers to keep the glamour and mystery of the medieval setting and play down the rats and rotting vegetables."

She cites The Knock and Turnaround as two recent programmes commissioned to deliver a particular audience that was not watching ITV drama. Schedulers regularly use research to predict the type of audience that might be interested in a show, she says. From this month, programme-makers can do the same, as TSMS makes its services available to them for the first time.

TSMS has the support of Pact, the independent producers' association. Until now, independents have been at a disadvantage, Pact says. They are unable to gain full access to Barb (Broadcasters' Audience Research Board) data that include Audience Indices, a measure of viewer appreciation available only to Barb's subscribers.

Broadcasters fiercely guard their research findings, fearful of giving rivals a competitive edge. But there is evidence that pre-broadcast research is employed comprehensively, if quietly. "ITV, Channel 4, the BBC - we're all doing it. The battle is to ensure its influence does not get out of hand," says one ITV source. John Bishop, controller of entertainment at Carlton UK, agrees: "We are much more research-oriented than we used to be, and this is reflected across all our programming." But, he adds quickly, research is employed "merely as a tool" to identify target audiences.

But does such research work? "It's absolute nonsense," says Paul Knight, a leading drama producer whose credits include London's Burning and The Knock. Mr Knight eschews all forms of pre-broadcast research. "Making a TV show is - and has to be - down to what you feel is instinctively right." London's Burning consistently achieves top audience-appreciation indices - "a fact I find out only after the programmes are filmed and aired".

In the US, broadcasters have been pre-testing concepts for decades. There, film studios regularly shoot different endings, which are screened in front of audiences before the final cut (though readjustedmovies that test "through the roof" still go out on release and flop). And TV companies rank stars by a system known as TVQs, under which Doctor Quinn, Medicine Woman's Jane Seymour outranks Glenn Close.

"Everyone wants to evaluate their risk in advance," explains Michael Clark, director of TV at the advertising agency Leo Burnett. But this has led to "lowest common denominator" TV. He says: "One of the main reasons why there is a lot of ordinary television in the US is because people rely on numbers, with only limited creative input."

Mr Knight fears that creative input would be the first casualty of adopting research wholesale: "When used as a tool, research is usually about procrastination - when a broadcaster can't make up its mind." There is some value in research after a show has been made, he adds. "But even then you can never believe everything these people say. Their sphere of influence is small, their questions selective and loaded."

Even Ms Harbutt admits research can be misused. "There is a danger people will use it to excess," she says. "And it certainly won't make a bad idea good." Also, it can fail to spot a gem. "Qualitative research of whether Robbie Coltrane would work as Cracker would almost certainly have come out negative before production. Research groups are notoriously unimaginative."

A Sunday night peak-time drama based around the exploits of a bunch of hookers hardly sounds like a recipe for ratings success, but it worked and already a second series is planned. Luckily, ITV followed gut instinct with Band of Gold. Time will tell whether research works as well.