±±¾©Èü³µpk10

Media buyers spent £640m on radio advertising in 2004, money that was placed with radio networks on the strength of audience figures supplied by the industry's measurement body Rajar.

The accuracy of these audience figures is paramount, but they have been called into question after some fluctuations that have seen stations swap places in the audience pecking order, seemingly without obvious explanation. Volatility has become the new word in the industry cannon.

The battle for London dominance could only be measured with a slide rule. The most recent Rajars for the second quarter saw Capital 95.8 snatch top spot, overtaking Heart 106.2, which slumped from first to third, with its audience share sliding from 7% to 5%.

Meanwhile, talk station LBC's share was 85% from a 2% to a 3.7% share of it's Total Share of Audience, kids channel Capital Disney's audience exploded with a 500% rise, while digital station Choice FM doubled its listeners.

Casting doubt

Such dramatic shifts within a three-month period are unlikely to result from genuine switches in listening habits. They cast doubt not only on Rajar's sample sizes and weighting, but the entire diary-based system of recording radio listening.

Last year, Kelvin MacKenzie, the former Wireless Group chairman, led a campaign against the diary system which ended up in the courts.

He remains vocal in his criticism of the latest results, insisting that the system is badly misfiring: "A diary can't keep pace when there's too many stations and too many listening opportunities. It's a disgrace and I think if we were in America, advertisers would be taking this measurement body to court. They say they can measure the audience, but they just can't prove it anymore."

Aside from MacKenzie, media owners have tended to stand behind Rajar, until now. Following the latest figures, Phil Riley, chief executive of Chrysalis Radio, owner of Heart 106.2, conceded that the Rajar system was now showing increased volatility and that there needed to be a system to cope with the demand of the increased number of radio stations currently measured by the survey, especially across the London market.

Rajar managing director Sally de la Bedoyere admits that there may be problems with sample sizes in certain sectors. The Rajar board will discuss the problem at its meeting on September 16 and will explore the possibility that there is an unstable sample size in London.

"I have no doubt that some of the volatility is real, because [London] is a very competitive market," says de la Bedoyere, "but also it can be a feature quarter-to quarter of the sample."

"If you have a different sample, every time every week, like Rajar, rather than a fixed panel like Barb, then the proportions of heavy listeners can fluctuate, which can have quite a dramatic effect on the figures."

An insider from one of the radio groups battling for dominance in London is more vocal regarding the volatility argument: "Of course, there is some volatility due to shifts in audience, but it is tiny compared to the volatility being caused by the statistical vagaries of the survey," he says.

Emap Radio programming boss Mark Story believes its is market forces that are ultimately causing the shift in dominance in the London market. "I think we're going to be in this state of flux, where each of those three players are going to be number one at any one time over the next two years," he says.

Three-pronged approach

De la Bedoyere says Rajar will look to address its critics with a three-pronged approach. The first possible change might be a move to a longer reporting period from three to six months, to try and iron out volatility.

"One of the options is going to six months, another option is to increase your sample, while another is to weight your hours and your reach in a different way. They're all being looked at and I'm sure the board will take a view on it."

However, any actions won't be rushed to augment the next set of Rajar figures, adds de la Bedoyere.

"Everything we're looking at would come into play as soon as we could bring it in and has been approved by the Rajar board."

The ultimate solution is the replacement of diaries with electronic measurement through audiometers, but this isn't scheduled until 2007 at the earliest.

This would mean the current system has to hold up for nearly two years, or must be patched up in some way in the interim.

However, further extreme results will continue to add pressure for changes.

Proposed short-term solutions

? Extend the Rajar reporting period from three to six months to flatten audience fluctuations

? Increase the audience sample size among specific demographic groups, or regions, particularly London

? Change the weighting of hours and reach in the Rajar survey