Marketers are divided over new ways of compiling data on the relationship between the ads that a sample of consumers see on TV and the brands they buy each week.
For marketers, single source data looked like a dream come true. For decades, they have been agonising about which half of their advertising budget works by how much and where, but information has never really been to hand to let them get a grip on the answer. Recently, however, new technologies in the form of so-called 'single source' data have promised to realise that dream. This should be of huge value to a wide range of brands including big names such as Ariel and Persil soap powders, Nescafe coffee or pet foods such as Whiskas or Chum.
Single source researchers use fancy widgets to gather data on exactly which ads a sample group of consumers actually see (or don't see) on television and exactly what brands the same consumers buy each week, thus allowing researchers to analyse the relationship between the two. It has only really taken off in the last two years but it's already setting that cat among the pigeons.
Markets that once seemed highly stable, with the brand shares of main competitors hardly shifting at all from year to year, are revealing astonishing levels of sales volatility from week to week. And analysis of consumers' actual behaviour suggests that, despite everything they say, they are influenced by the ads they see - enough of them go out to buy the brands they see advertised to produce immediate, measurable sales results. A recent study of over 50 brands by Andrew Roberts, technical director at UK research firm Taylor Nelson AGB, shows that, on average, sales increased by 6.1% after an ad was aired, for example, and most of this sales effect occurred within three days of consumers seeing the ad in question.
Even more startling is research from the US, which suggests that many advertisers are wasting enormous amounts of money by making two simple mistakes. First, says John Philip Jones, a researcher from Syracuse University, doing ad campaigns the traditional way - in bursts - is almost a waste of time. Each ad burst generates a positive sales blip but, once it's over, competitors are left free to gain back the share that they lost.
Second, he says that advertisers are wasting a second tranche of money by repeating their ads too often within those bursts. Single source data suggests that most of the big sales effects follow from just one 'opportunity to see'. Put these two together, he suggests, and the elusive goal of long-term brand share growth suddenly becomes simple. 'If you have many short-term effects, you get a longer term effect.' The strongest year-end sales improvements occur with brands that have a strong short-term advertising strength allied to above-average media spend: in other words, a strong campaign reinforced by media continuity.'
Balderdash, say a host of angry critics. Indeed, a team of Leo Burnett researchers led by that doyen of the art of measuring advertising effectiveness, Simon Broadbent, claims that its analysis of the purchasing and television viewing habits of consumers in single source panels reveals previous in-built biases. 'We are not observing so much the effect of more exposures on the same people as a different sort of people,' the team says.
And the net effect? The discovery of the holy grail of single source data has simply sharpened the divide between different camps. The notion that advertising actually persuades consumers to go out and buy brands in mature, established markets is absurd, says Andrew Ehrenberg of London's South Bank University. Most consumers already know the brand that's being advertised, he points out. Advertising simply nudges their memories of it, without which sales slowly begin to sag.