drowning by numbers

TV research is under the spotlight as BARB’s cost and effectiveness comes in for tough scrutiny from both the ITV sales groups and the BBC. Will a new, improved system be the result?

SBHD: TV research is under the spotlight as BARB’s cost and effectiveness comes in for tough scrutiny from both the ITV sales groups and the BBC. Will a new, improved system be the result?

Television had better watch itself. As the shape of broadcasting changes, increasing pressure is being placed on its systems of measurement, the currency that is supposed to show whether or not it’s working. The TV industry has now begun to ask how much it really needs to know.

The present BARB panel was invented 15 years ago to provide a comprehensive gauge of who watches what on TV. Such a measure was necessary to provide a sound currency according to which ITV airtime could be bought and sold.

At the time, ITV was a patchwork federation of regional monopolies, which relied on a sizeable sample size (the so-called panel) of respondents within each of its 15 regions to guarantee accuracy. In theory, these conditions still apply. But in practice, last year’s consolidation of the ITV sales created three megaliths, and simultaneously removed the need for such a detailed regional analysis of the viewing public.

Only halfway through the present BARB contract, which was awarded to AGB in 1991, TV companies and agencies have begun to discuss how it may be simplified, although there are concerns on all sides that by the end of the contract in 1998, much of the data it generates will cease to be relevant.

“This particular sample was designed to give improved regional measurements. If they now tell me there are only three regions, I will get a design of research to suit that,” says BARB director Bill Meredith.

The UK’s panel was enlarged considerably in 1991 to encompass a larger number of sub-groups within the population as a whole. The size of the panel (more than 4,000 respondents) is very large by international standards. The driving force behind the formation of the current BARB was clients’ outspoken desire to tighten the targeting of their advertising as the recession bit into profits. It is also very costly. At about £8m a year, BARB is coming in for some tough financial scrutiny both by the ITV sales groups that use it, and the BBC that doesn’t (very much).

The problem for television research is that it is expected to do two jobs at once. And the commercial sector’s need for a trading currency is at odds with a real measure of TV advertising’s impact on viewers.

“There is a clash inherent in all discussions about TV research between the need for a currency and the need for effectiveness,” says The Billett Consultancy chief executive John Billett. TV airtime could be traded on a much simpler basis, but the resulting information would give no picture of effectiveness. But if effectiveness was the sole criterion, the system would cost too much, says Billett.

The cost of BARB is a problem ITV could well do without, as its constituent companies seek to weigh their unyielding franchise repayments against an increasingly rigorous ITC remit. Thanks to the Station Average Price system, broadcast directors at agencies are constantly tweaking their buying pattern based on statistics that are just not accurate enough to warrant it, according to one senior ITV executive.

“I’d prefer a system that was about advertising effectiveness rather than a trainspotting mentality of whether one watches 30 seconds or 31 seconds of a programme,” says the source.

Where sales are concerned, BARB has very little to say. Last year, AGB launched Mediaspan, a research product that links BARB data to consumer research from its own panel, Superpanel. In theory, this should give some idea of the effect of advertising on grocery purchase. It measures consumption and it measures advertising, and matches the two. Unfortunately, the tie-up actually works only on the basis of similar demographic households behaving alike. The direct impact of airtime exposure remains a mystery.

For the broadcasters themselves, the nearest thing to a measure of the quality of television comes in the form of Appreciation Indices (AIs), designed to measure an audience’s enjoyment and interest of a programme. To their continuing frustration, neither advertisers nor agencies are privy to this analysis – BARB continues to refuse them access to it.

Agencies will never really have access to effectiveness data until it is possible to correlate particular homes receiving a particular ad against buying habits – an option which is only now forming the basis of experiments in US cable regions. But by the stage such measurement is technologically feasible, will it be relevant to commercial television?

United Artists Programming vice-president of research Dave Brennan has suggested that the only solution for the fragmented cable companies – which are not measured by BARB – is to establish an audience research system for themselves. He has set up a working group to examine the idea.

“What the cable industry needs from research is not relevant to the rest of television. By definition, it is clustered and the way the companies and viewers communicate with each other could be very different,” says Brennan.

As the nature of TV changes, research must ask itself what it really needs to know.