But would you watch it?

Can TV programmers find out in advance whether a show will be a hit? Paul McCann looks at concept-testing, and, opposite, our journalists run their own test of the profiling software Convict

It would be hard to find a phrase less pleasing to the ears of an old-fashioned television auteur than "concept-testing". The words drip with the kind of Nineties marketing-speak that is identified with the worst Birtists at the BBC. From the late Dennis Potter to the scriptwriter Andrew Davies, creative types have condemned modern drama departments for churning out worn-out, formulaic "genre television" - the vet-who- is-a-chef-who-is-a-detective syndrome - and concept-testing gets the blame.

But few people, even in the industry, know for sure what it is. At its most straightforward, concept-testing is the use of research to find out whether a programme idea is likely to work. At ITV it is as old as the decade and coincides with the creation of the new ITV by the 1990 Broadcasting Act. Without it, half the mainstays of ITV's drama schedule would never have seen the light of day. Heartbeat, Band of Gold, Bramwell and Kavanagh QC were products of concept-testing.

Its first use in the UK was at Tyne Tees in 1990, where a researcher, David Brennan, was working on the broadcaster's franchise bid. Tyne Tees needed to know if any of the new programmes it was proposing to the Independent Television Commission would work. Brennan decided to ask members of the public if they would actually watch what was proposed.

The system seemed to work, and when he moved to run the whole of ITV's programme research at the Network Centre in 1993, Brennan took concept- testing with him. By then, ITV was becoming more sophisticated in its use of research. Rather than put a new programme idea to a member of the public, ITV created a false listings magazine with its proposed new programmes interspersed with totally fictitious ideas.

"If you present a whole new programme idea to someone in outline form the way a commissioning editor would, it is too conceptual. Instead, you give them the same information a viewer gets when a new programme is launched - listings information and a trailer if you're lucky," says Brennan. "Then you can ask why they chose it and what they like about it without forcing them to focus on it in an unnatural way."

For certain ideas, concept-testing is used to make sure the whole thing isn't a turn-off. ITV already had Band of Gold, its drama about a group of Northern prostitutes, in pre-production when it started to worry, not surprisingly, that the subject matter might prove a turn-off at the outset.

"About a year ago someone writing about Band of Gold used it as an example of how programme research would stifle new programmes," says Brennan. "They said no one would ever have predicted a programme about Northern prostitutes would be a success, so it proves you can't do research on these things. But the fact is that we did test it and we found the audience would accept it as long as the prostitution was in context."

The BBC began concept-testing in 1993, but uses it much less than ITV where every second of airtime has a commercial imperative to make money and bring in the right audiences for advertisers. The BBC tends to test only its bigger-budget, long-running dramas, such as Dangerfield and Silent Witness. With costs for location dramas hitting pounds 500,000 an hour it is hardly surprising that broadcasters test ideas before they spend.

"Entertainment shows are less well-suited to concept-testing," says Katherine Lannon, the BBC drama department's development analyst. "With those, you can tinker with the format once they are running. With a filmed drama, there's not much you can do." But even tinkering could not save ITV's Man O Man and The Baldy Man entertainment shows. They tested badly, were made anyway - but flopped even after adjustment.

Critics of concept-testing say it leads to bland programmes inevitably similar to those already on air, as only familiar concepts can be tested. The Fast Show is a case in point. "There's no way you could concept- test something like The Fast Show," says Brennan. "You probably couldn't even test a pilot, because it took three or four episodes for you to begin to understand the characters. But the fact remains that 95 per cent of what is on TV is nothing like as creative and innovative as The Fast Show. Most of it falls into standard genres that can be researched."

It is this that prompted one ITV programming department to try to create the perfect programme entirely by computer. All the elements of current shows were fed into a piece of software, including location, stars, characters and, of course, ratings. "What emerged was horrible," says Brennan. "It was a half-hour, ensemble sitcom about binmen set in Manchester. It contained every cliche in the book."

Katherine Lannon maintains that concept-testing is still on a "smallish scale" at the BBC, and even when it is used on its own it would not dictate whether something was made or not. "It is just one element of the whole commissioning debate. It's the bit that keeps us in touch with the audience." The corporation doesn't admit to testing the turkeys of its output, such as Rhodes or Nostromo, but said that both series showed that it was willing to risk failure in the pursuit of creativity.

David Brennan believes there is one basic reason why old-school television types needn't worry about concept-testing taking over. "If research could really tell you exactly whether everything would work, you wouldn't need lots of well-paid programming executives - and they are the ones in charge of concept-testing"n