[an error occurred while processing this directive]
BBC News
watch One-Minute World News
Last Updated: Tuesday, 3 June, 2003, 09:30 GMT 10:30 UK
Rating the ratings
Nick Higham
By Nick Higham
BBC media correspondent

All television and radio audience research is a benign confidence-trick.

Since no-one can possibly know for certain what the entire nation has been viewing and listening to, the TV, radio and advertising industries use classic market research techniques instead to record the behaviour of a sample panel, and extrapolate from that to the behaviour of the population at large.

This benign conspiracy is conducted by two organisations called Barb, for television, and Rajar, for radio.

Radio dial
Rajar measures stations by using diaries
The system works perfectly well so long as everyone agrees to play along with it, and no-one thinks they get a raw deal.

It enables programme-makers and broadcasters to know how popular their programmes are and with whom, and allows hundreds of millions of pounds' worth of advertising airtime to be bought and sold on the basis of Barb and Rajar figures.

But now the system is threatened. Someone believes they are getting a raw deal.

The Wireless Group, run by Kelvin MacKenzie, believes Rajar figures - based on diaries compiled by a panel of listeners - seriously underestimate the audience for the group's national flagship station, Talksport, and that Talksport as a result is missing out on millions of pounds in advertising revenue.

Radio 4
Wristbands (GfK): 17.9m
Diaries (Rajar): 10m
Radio 2
Wristbands: 15.2m
Diaries: 13.2m
Radio 1
Wristbands: 12.7m
Diaries: 10.3m
Five Live
Wristbands: 10.7m
Diaries: 6.4m
Radio 3
Wristbands: 3.8m
Diaries: 2.1m
Sources: Rajar/RSL (Jan-Mar 2003) and GfK (10 March - 20 April 2003)
MacKenzie believes he has found a better system for measuring audiences and last week he produced the latest piece of evidence: the first in a projected series of monthly surveys of radio listening conducted by GfK, a leading research company which runs television audience measurement systems in six European countries, including Germany, France and the Netherlands.

Surprise, surprise, Talksport did rather well.

According to GfK it has 8.1 million listeners a week; according to Rajar it has just 2.2 million.

In fact on GfK's figures all radio stations seem to have larger audiences, but speech stations do particularly well: Radio 4, for instance, has 17.9 million listeners a week, almost eight million more than recorded by Rajar, and overtakes Radio 2 to become the most popular network.

The differences are down to two things. One is a different definition of what constitutes "listening": a minimum of four minutes, according to Rajar, four seconds according to GfK.


The other is in the technology. GfK uses a wristwatch-style meter which records everything the wearer hears, rather than those cumbersome diaries.

Kelvin Mackenzie
Kelvin MacKenzie says Talksport gets a raw deal from Rajar
Rajar, not surprisingly, has reacted rather huffily to MacKenzie's initiative.

It has been conducting its own trials of the GfK meter and a rival pager-style device marketed by the US company Arbitron and the results are expected shortly.

It wants answers to questions like whether the meters can distinguish between spoken word radio stations and ordinary conversation, and whether people will wear them first thing in the morning, even in the shower (vital when so much radio listening happens at breakfast).

Rajar suggests GfK's figures aren't necessarily more accurate, but may simply represent a different kind of inaccuracy.

But even if they are more accurate, it isn't simply a question of replacing one measurement system with another.

For one thing the meters are likely to be considerably more expensive. Rajar is already one of the most expensive market research operations in the world, because it has to measure audiences for hundreds of local radio stations, many with overlapping transmission areas.

Persuading the radio stations and advertisers to stump up yet more money for a system that may actively disadvantage some of the existing subscribers won't be easy.


Meanwhile MacKenzie's initiative has had two significant consequences, one potentially bad, the other potentially good.

On the debit side, the existence of two radically different sets of audience figures threatens to undermine confidence in the ratings generally - once the con trick is exposed it becomes harder to maintain.

Wristbands (GfK): 8.1m
Diaries (Rajar): 2.2m
Classic FM
Wristbands: 7.3m
Diaries: 6.8m
Virgin AM and FM
Wristbands: 3.9m
Diaries: 2.7m
Sources: Rajar/RSL (Jan-Mar 2003) and GfK (10 March - 20 April 2003)
Confidence is fragile: the Barb TV ratings are still recovering from the shock delivered last year when changing the entire panel of viewers surveyed produced some very peculiar results for a few months.

On the credit side, GfK has shown it might be possible to combine TV and radio audience measurement.

The wristwatch meters pick up television viewing as well: Sky News and Sky Sports 2 have much larger audiences according to GfK, for instance, because they are watched a great deal in pubs and offices, while Barb only measures viewing in the home.

In the long run, combining the two systems might be one way to pay the cost of measuring all radio listening with meters.

But with Barb only just embarked on a new five-year contract with its existing supplier, it's not going to happen overnight.

Radio 4 tops new survey
29 May 03  |  Entertainment

The BBC is not responsible for the content of external internet sites


News Front Page | Africa | Americas | Asia-Pacific | Europe | Middle East | South Asia
UK | Business | Entertainment | Science/Nature | Technology | Health
Have Your Say | In Pictures | Week at a Glance | Country Profiles | In Depth | Programmes
Americas Africa Europe Middle East South Asia Asia Pacific