The numbers game is up

How accurate is the research into what we watch and listen to? Even the industry isn't completely sure.

There are lies, damned lies and there are statistics: there is always concrete proof that your product or channel is really No 1. The media is obsessed with research. Television and radio audiences, circulation and sales figures directly influence revenue. Which is why the music business is reeling after details emerged last week of one of the biggest chart-rigging scams in recent years.

The record industry's trade body, the British Phonographic Industry, fined two of its members. Edel (UK) Records and Castle Communications, along with a third company - Love This Records, not a BPI member - are accused of "hyping" seven records by buying up large numbers to boost their positions.

The revelations surprise those who believed that computerised systems made the charts too difficult to rig. The last time anyone was caught and fined was five years ago. Says one industry executive: "Consumers see the charts as a buying guide, retailers use it to know what to stock. This undermines everything."

But it's not only deceit that can shatter trust. Recent changes in the way radio listening is measured have dramatically affected nearly all UK radio stations. With some stations suffering an apparent fall of up to 700,000 listeners over just four months, confidence in the new system has been rocked. Meanwhile, there is dissent among television companies, which claim existing viewing data has failed to keep pace with the expanding broadcast market.

1. The Charts

Alan Freed, the Fifties American DJ credited with coining the phrase "rock 'n' roll", was tried for receiving bribes to play records; the practice was called "payola". The last proven British case was in l992, involving London Records and a band called the High. The new computerised system makes rigging harder to achieve and quicker to detect - or so the chart compilers claim.

The two most important UK charts are the Radio 1 weekly Official Top 40 and its commercial rival, the Network Chart. The "Official" is the older of the two and is based solely on record sales. Based on a combination of sales and commercial radio airplay, the upstart Network Chart has become a valuable weapon for commercial radio stations fighting back in the battle for Sunday evening listening.

After long battles, an apparent truce ensued three years ago, when both charts were taken under the wing of a single body, Chart Information Network. CIN, owned by the record industry trade body BPI and the publisher of Music Week, commissions and polices research for each. The move reduced contradictions over the weekly No 1, so that today both charts share the same Top 10 - based only on record sales. Outside the Top 10, the Network Chart takes into account airplay. Yet this apparent compromise has done little to lay old scores to rest.

"The Radio 1 chart remains the official record industry measure," one record business source claims. But Chris Dickens, group head of programmes at Capital Radio, producer of the Network Chart Show, counters: "We're the people's choice - the true reflection of what the public likes."

Pop charts are perceived to be most open to abuse. Not so, claims Catherine Pusey, chart director of CIN. "New technology protects their validity," she says. "Both use sales data from 2,200 record shops that electronically register barcodes of all products sold."

While it is impossible to top the charts by buying up particular singles, it can take sales of just 10,000 to slip into the Top 30. Rigging is high risk - sales blips are soon spotted, says Pusey. Suspect stores are excluded and singles can be suspended from the chart, as happened to "Santa Maria" by Tatjana last year.

But there is nothing to stop record companies paying inducements to staff to promote singles, selling records to stores at discounts and promoting heavily in-store.

2. Rajar - Radio

Every year 150,000 people are recruited to monitor their radio listening for Rajar - the largest piece of regularly conducted media research in the UK. It costs pounds 2.5m a year, paid for - and jointly owned by - the BBC and commercial radio stations. Quarterly listening data has been published since Rajar was launched three years ago.

Then last autumn the unthinkable happened. "We modified our research method to cope better with the growing number of stations," explains Rajar's executive director, Roger Gane. "Pilot tests on the new system indicated close conformity to previous data. But then virtually all services began to show a drop in weekly reach." Some stations' monthly audiences plummeted by 40 per cent, and smaller audiences mean lower advertising revenue.

Rajar involves up to 50,000 households. Participants fill in a listening diary. The old diaries carried pre-printed lists of all stations; the new ones are blank for respondents to compile themselves, using names provided on stickers.

While the numbers responding remained the same, weekly declared listening has declined. This, stations claim, is because of the stickers. They say the system means listeners who only occasionally tune into a station are less likely to remember and to include it in the survey. "It is certainly not true to say these figures are right and those from Rajar [and its predecessor, Jicrar] over the past 20 years were wrong," says Classic FM's sales director, Nigel Reeve.

As a result, publication of the latest figures is not expected until mid-March - six weeks late. Rajar is now analysing all stations' occasional listeners and comparing figures before and after the change. It will also run a "parallel study", using the old method.

3. Barb TV charts

Each week the top 30 TV programmes are published by Barb (Broadcaster's Audience Research Board). Barb monitors who watches what and when, surveying viewing habits in 4,435 households using a set-top "black box".

"The public is not getting the whole picture," one industry expert observes. "The top 30s are dominated by shows like EastEnders and Coronation Street that are on all year round, whose ratings include repeat showings and video replay."

Moreover, he says, rivals do not want competitors to see which of their programme formats worked best or worst; the repeats and video replays conceal the true figures. And advertisers, who indirectly pay for the bulk of this research, are increasingly interested only in broader data: "They want to know about channel share rather than individual programme performance."

Barb is co-funded by the major broadcasters. Designed for a three-channel TV market, it must now accommodate more than 25 channels. Cable and satellite broadcasters are dissatisfied.

Only 800 Barb households receive more than four terrestrial channels and only 180 of these are cable homes. "We're the growth part of the industry, but we've been left behind by Barb," says David Brennan, vice president of research at United Artists Programming, whose channels include Bravo and Discovery.

Increasing the number of cable homes surveyed is prohibitively expensive, says Steve Wilcox, a director at the research company RSMB. Even so, at the end of January RSMB launched cable TV research in 1,000 homes - the first step towards establishing a credible, regular survey. It is, of course, being funded by the cable TV industry.