The Big Lie – the complete book online - 1 Effectiveness

Article Index


Chapter 1

The Holy Grail


The time has come when advertising has, in some hands, reached the status of a science . . . Certainly no other enterprise with comparable possibilities need involve so little risk.

Claude C. Hopkins, American advertising pioneer, Scientific Advertising, 1923

Forty executives sat in the darkened conference room in Camden, New Jersey. There were no women – it was 1960. You couldn’t see their eyes; each man wore a pair of bright orange cardboard 3D spectacles with lenses of smoky film. They couldn’t speak either, because their teeth were clenched on white lollipop sticks. Inside their mouths, instead of a sweet, a narrow ribbon of gauze was wound tightly round the end of each stick. The audience stared at a white screen on which a colour slide showed a stack of pancakes dripping with maple syrup. The projector clicked and the pancakes were replaced by a mouth-watering chocolate cake.

The man by the screen, wielding a pointer and shambling across the beam of light, was Horace Schwerin. A pioneer of advertising research, he ran a leading firm which specialised in the testing of television commercials, using a singular method of his own invention. Today he was demonstrating something new to the marketing department of the Campbell Soup Company, manufacturers of the famous canned soups and a wide range of other popular food products. It was an experiment devised by his 75-year-old father, “Pop” Schwerin, a retired Bell Telephone developmental engineer. Based on techniques used by Ivan Pavlov, the Russian physiologist, the idea was to measure the rate of salivation triggered by showing people appetising photographic images. When the lights were switched on in the conference room, the men from Campbell Soup took off their cardboard spectacles, removed the lollipop sticks from their mouths, unrolled the gauze ribbons, and solemnly counted the number of saliva stains that had penetrated through the layers.

Schwerin was using his “salivation tester” in experiments to study the reaction of adults, infants, and neurotic rats to various stimuli. If a valid measurement could be derived, it would be advertising’s Holy Grail, a system of measuring the effect of food advertising not by asking people questions, but through physical measurement. If advertising could make you drool, the salivary glands couldn’t lie.

Four years previously, a different physiological experiment in a New Jersey drive-in cinema had ignited a firestorm of public controversy. For six weeks, at an outdoor cinema in Fort Lee, New Jersey, a motivational research company run by James Vicary had flashed the injunctions “Drink Coca-Cola” and “Hungry? Eat Popcorn” on the screen every five seconds throughout the feature film. These messages appeared for 1/3000th of a second, calculated to be below the threshold of visual perception. The hope was that the message would nevertheless penetrate the subconscious brain. Vicary, who had already attracted attention for his studies of the eye-blink rate of female shoppers browsing through store displays, reported that the sales of cola drinks rose 18 per cent over the period of the test, and popcorn consumption went up 58 per cent. The nation which had invented advertising became alarmed about the threat of “subliminal” manipulation. It was another new word, like “brainwashed”, which had been recently invented to explain why some US soldiers had rejected American values in captivity during the Korean War.

Horace Schwerin was obsessed with what he called “action measurements”. To him it was patent that what people said they did, or would be likely to do, under given circumstances was burdened with so much psychological baggage as to be almost totally unreliable. During the Second World War, as an enlisted man in the US Army attached to a psychological unit, he devised a way of testing this belief. Soldiers in training camps were routinely issued with two pairs of boots, and in the interests of foot hygiene the army medical service recommended that the pairs should be worn on alternate days. For the GI, however, there was a massive disincentive: the pair he wasn’t wearing had to be displayed, properly shined, beneath his cot. It was considerably more convenient to wear the same pair of boots every day, and keep the second pair on display at all times, pristine and ready for inspection.

To resolve this conflict of its own devising, the US Army resorted to advertising. Messages were broadcast in mess halls urging the GIs to change their footwear daily. To determine the influence of this propaganda, a company of soldiers which had been exposed to the messages was matched against a control group which had not. The next day, when the First Sergeants asked which men had changed their boots, every hand was raised – in both companies. One hundred per cent compliance is an unusual result in any behavioural survey, so Sgt Schwerin devised an “action measurement” to check this finding. While the soldiers were out on training, his researchers entered the barracks and chalked the soles of each pair of boots under the bunks. Following further test broadcasts, they made return visits and found that almost all of the boots under the bunks still bore chalk marks – in both companies. The soldiers had lied with their boots on.

After the war, the technique which Schwerin developed to test the effectiveness of radio commercials, and later television, while less absolute, recognised that self-interest is also the consumer’s principal motivation. He called it “competitive preference”. For decades, Schwerin conducted tests almost every weekday night in New York and other North American cities. A random sample of 300 to 400 consumers was attracted to a “preview show” in a cinema. Before the show, several tombola lotteries were held. The prizes were large supplies of various consumer products, and for each prize draw audience members were given a list of major brands in a product field and asked to tick which brand they would like to receive if they won the draw. They then saw a television programme and a few television commercials, one for each of the relevant product categories. Afterwards another series of prize draws was held, and the respondents were again asked to indicate their brand preferences in the same way as before.

Schwerin’s theory was that the prize draw placed the consumer in a situation where he or she had to make a decision of some personal importance. The individual’s response would be well considered, because it carried with it the prospect of penalty as well as reward; a frivolous choice could result in a large quantity of an unfavoured brand of shampoo arriving on the doorstep.

In these tests consumers did regularly change their brand choices after a single such exposure to advertising. Schwerin’s measurement of effectiveness was the difference between the percentage of respondents who chose the advertised brand before and after seeing the commercial. This figure could be compared with the results of other commercials tested for the same brand or competitive brands in the same product field, tested under the same conditions amongst different but statistically comparable audiences.

Intellectually, the technique had great appeal. While there existed plenty of measurements of various effects of advertising exposure – for example, whether people remembered seeing an advertisement, whether they recalled its message, or simply whether they liked it – Schwerin’s “competitive preference” zeroed in on what most people think most advertising aims at: creating a change in consumer behaviour. There was encouraging resonance with real-life behaviour, too. The pre-choice levels, before advertising exposure, roughly resembled brand sales shares. Commercials in product fields which were known to have only loose brand allegiances, such as plastic clingfilm wraps, usually gained high preference changes, while advertisements for cigarettes, for example, where brand loyalties were known to be very strong, produced minimal shifts. And commercials for new brands with attractive new features would usually outperform the norm in any product field.

All advertising research techniques attract methodological criticism, and the Schwerin method, probably because of the scale of its ambitions, drew more than most. Complaints focused on the self-selecting nature of the audience and practical difficulties in controlling tests in the “test theatre” environment. Nevertheless the technique had certain inherent advantages. One of these was that it could demonstrate reliable results. The same commercials could be fed into Schwerin’s research factory again and again under similar conditions to see whether the result could be replicated. This is a basic criterion of all scientific measurement, yet repetitive testing is flatly impossible or prohibitively expensive in most other kinds of advertising research. The Schwerin method offered practical help, too. The test commercial could be varied to evaluate various creative factors – the selling message, the style of its presentation, or its length. Or the advertisement could be held constant while its programme environment was changed. In America, commercials are shown within the context of programmes, and advertisers spend huge sums sponsoring these. As well as considering the types of viewer attracted by these programmes, the advertisers were concerned about the compatibility of the viewing environment with their brands. To help make programme choices they used Schwerin tests to explore whether, say, a commercial for Betty Crocker Cake Mix would be more or less effective in a relaxed situation comedy or a high-tension thriller.

Most advertising researchers are content with the limitations of their measurements. Mindful of how many other factors can affect a brand’s success in the marketplace, such as pricing, distribution imbalances and competitive sales promotional activities, they are generally loth to venture any direct connection with sales results. Researchers usually defend their results on theoretical grounds: by measuring one or more factors which may play a part in how advertising works, they can make some sort of contribution to a judgement of an advertisement’s effectiveness. Schwerin’s goal was loftier and more empirical. His firm continually encouraged advertisers to test whether commercials which had earned higher “competitive preference” scores would actually produce more sales in the marketplace, through controlled market tests using different commercials in different sales regions. Successful results were presented as validation of the technique.

In the mid-1960s the quest for validation led the Schwerin Research Corporation to conduct the most ambitious attempt ever undertaken, before or since, to relate advertising test results to sales results in the marketplace. Unilever funded a nationwide study in West Germany, selected because severe federal restrictions on the value of consumer give-aways meant that sales promotion activity would be a less disruptive factor. Four major consumer product fields were studied: laundry detergents, washing-up liquids, margarine, and toothpaste. For eighteen months Schwerin tested every television commercial broadcast and every national press advertisement published on behalf of every major brand in these product fields. Some of the usual difficulties encountered in controlling “test theatre” audiences were eased in Germany. While New York audiences in the 1960s could be restive, even mutinous, in Hamburg they were docile and interested. (Once, expressing the ultimate “action measurement”, a man died in a rear row; ambulance attendants entered quietly and unobtrusively removed the corpse while the show went on.) In repetitive tests, results in Germany were consistently more reliable than anywhere else.

Schwerin carried out 179 press and television advertising evaluations for 24 brands; the share of advertising expenditure by each brand in each medium was also tabulated. These key variables – test scores and spend on press and television – were compared to quarterly retail sales share movements in a multiple-regression analysis. Taken together over eighteen months, they predicted 24 per cent of all brand sales share movements in the laundry detergent field, 41 per cent for washing up liquids, 59 per cent for toothpaste and 80 per cent for margarine. Across all four fields, these measurements explained slightly under one-third of all sales behaviour, which was seen as a promising result. Significantly, if the Schwerin measurements of the effectiveness of the advertising content were disregarded, advertising expenditure alone had a much poorer, and in some cases an inverse, correlation to sales movements.

Today, an adaptation of Schwerin’s “competitive preference” technique called “persuasion/attitude shift” is still used to test advertisements by the American firm, McCollum Spielman Worldwide, which has ­inherited leadership in this field, but it is just one of a basket of measurements used by this firm. In the UK, serious efforts to validate techniques with reference to sales results are out of fashion in the industry.


Forty years on, people still worry about subliminal advertising, but usually the term is a misnomer. They generally mean irrational appeals to the emotions rather than subconscious stimuli which actually fail to cross the threshold of physiological perception. James Vicary refused to release details or statistical data from his 1956 study of subliminal effects. Despite several attempts to duplicate his cinema advertising experiment, no one was ever able to produce similar results. Apprehensions about true subliminal advertising faded, though it was not until 1974 that the Federal Communications Commission in America finally banned the technique, because “Whether effective or not, such broadcasts clearly are intended to be deceptive”. Subliminal communication was dubious in any case, for the same reason that “Pop” Schwerin’s taste-testing technique never bore fruit: individual thresholds of perception are different. In a 1956 BBC television experiment, some people glimpsed the hidden message while others didn’t; and people salivate at different rates, too.

Research methods have not changed much over the years: they still offer to measure many different effects of advertising. Applied to the behavioural problem the US Army encountered in Fort Benning, Georgia in 1944, all these techniques would falsely indicate, in their various ways, that soldiers change their boots every day. They do not predict behaviour. Horace Schwerin’s dream of a single, validated action measurement of advertising effectiveness remains a chimera. Indeed, few advertising people seem to care much about any kind of empirical measurement of the influence of their expensive black art.

Why is that?