Turns out he wasn't that far off the mark. Because qualitative research isn't just being used to inform the curating of museums and galleries. And it's not even just a question of helping to shape artistic programmes. The people are having their say about what they want from music, television and literature.

For more than 20 years, RPM Research has been providing companies like EMI, Sony BMG, Warners and Universal with research that helps the creative juices flow ­ profitably. It develops strategies to ensure, for example, that an artist's sales potential is realised, that a retailer fully meets the needs of the customer, or that a website is performing to its full potential.

It may research an artist before they've been signed, or alternatively those signed but not yet launched. Sometimes it will choose a release song before even a note has been heard on the radio. "For one band that proved phenomenally successful we chose the name, the line up, the polaroids, all with the help of focus groups," says partner David Lewis.

The company uses a combination of qual, quant, and a quarterly panel study to monitor changing tastes in music, musicians and musical culture. "It's not a complete panacea but does a good job of seeing what is coming into fashion," says David. And it can help companies avoid expensive mistakes.

He cites one company where a group of 14 to 15-year-old girls slated the concept, saying: "Oh my God, this has been written by a sad 30 to 40-year-old marketing guy." The said marketing guy was by this stage chucking stuff at the two-way mirror saying: "What do they know?" After some discussion, the company extricated itself from the deal. Another record company picked it up and, without the benefit of research, spent £1m. The project bombed.

Research can't, says David, be used to predict innovation. But it can identify if music is wrong for the market it's aimed at and suggest ways of changing it to make it more successful.

In the world of drama, meanwhile, casting agents can command a pretty penny. But sometimes ­ as with reality shows ­ the public can choose the star character. For ITV's Miss Marple, Chorion ­ the producers ­ turned to qualitative research to work out which actress was right for the part. Several were put through their paces before Julia McKenzie was chosen.

And in children's literature, focus groups are being used to even greater effect. They are helping to identify themes and story lines that can be used to brief authors. London-based business Hothouse used Discovery Research to discuss story ideas with kids and get feedback from them on which characters, plots and ideas they liked best.

The upshot has been Tom Becker's Darkside, which won Waterstone's prize for children's fiction and Puffin Books" new series, Fright Night.

Qualitative research never predicted punk rock. And it panned The Office, which got the lowest ever score on the NBC focus group and on BBC2 (joint bottom with women's bowls). The Hothouse experience, though, shows that it's unwise to ignore its potential contribution to creating content. And with media fragmentation creating an unquenchable demand, that's well worth remembering.