Companies and advertising agencies would like to know whether their subjects resonate with consumers and achieve the desired effect: After all, helmets should be worn, condoms used, mineral water drunk and cars bought. To achieve this, advertising has to work. However, the performance of advertising media cannot yet be measured. Providers of internet quick tests, pre- and post-tests claim that they can find out how effective advertising is. But what do the percentages behind awareness, uniqueness, brand fit and likeability tell us? Exactly: nothing at all. The values determined are usually in the low one to two-digit range, a campaign never reaches 90 or 100 percent, even if it is good. Nobody can explain why. Nor can the providers of the posttests tell companies and agencies what they need to do to increase the values for the measured parameters.
Pretests also fail. This is shown by cases in which an advertising motif that was actually shot down accidentally comes into circulation - and unexpectedly increases sales. This was the case, for example, with an advertising campaign by Reynolds for its Winston cigarette brand: in the pre-test, a subject performed well for French-speaking Switzerland and poorly for German-speaking Switzerland. The company decided to use it only in French-speaking Switzerland. However, due to an oversight, the posters were also put up in German-speaking Switzerland - and achieved a sensational recall. In short: the existing tests cannot explain the effect of advertising and they also do not allow any predictions to be made about how subjects will be received by consumers. Advertisers and companies have to hope that their campaign works. The proof of the pudding is in the eating. But that was yesterday.
Practical data"For the first time, I can determine valid data on the effectiveness of advertising measures in good time," promises Dr. Patrik Zwahlen (44), founder and Managing Director of the Zurich-based research institute for visual communication Bildkom. Zwahlen is a trained photographer and studied the psychology of perception, marketing and artificial intelligence. He financed his studies with his own DTP agency. During his studies, Zwahlen worked on measuring the impact of advertising. In his dissertation, he developed the theoretical basis for his new survey software, which is tailored to the needs of communication specialists in everyday advertising.
Today, eight years and 450 test subjects later, Zwahlen offers his measurement method for sale. The scientist assumes that advertising material works if the subject conveys the message intended by the company. He can check whether this is the case with his test. Here's how: around 50 test subjects are shown the subject under investigation. Bildkom accompanies the perception process, monitors the flow of information from image, film, text and sound and uses scientific methods to record how the test subjects actually perceive the subject. All components of the advertising material - image, design, structure, copyline and the interaction of all elements - are systematically examined for all aspects that influence the impact of a subject. In the case of a poster, for example, there are certain focal points that the recipient must look at in order to understand and remember the image. This has been scientifically proven. If Zwahlen's test subjects look elsewhere, it is clear that the subject cannot work - to put the complicated scientific truth simply. Zwahlen can also find out whether the amount of information conveyed fits the advertising medium. For example, an advertisement in a trade journal can take more text than a poster. Zwahlen knows exactly how much. "My survey software examines the perception process," he summarizes. "When evaluating the data, we combine proven knowledge from the psychology of perception with findings from business quality assurance."
Subjects under the magnifying glassThe existing pre- and post-tests all assume that the recipient has seen an advertisement and that they have understood what it is supposed to convey. But according to Zwahlen, this is often not the case: "Only 50 percent of the subjects and 20 percent of the commercials are perceived by the target group in the intended sense. All previous tests ignore this fact. That's why they don't work." Zwahlen replaces pre- and post-tests with analysis and quality control. His first question is whether the subject is seen at all, and if not, why not. "If consumers don't perceive the message that a company wants to send out, I can explain to the art director why this is the case," says the scientist. "This makes it possible to minimize the usual gap between target formulation and target achievement." It sounds as if every creative just needs to apply a few simple recipes to guarantee their campaign resounding success. Unfortunately, it's not that simple: "It may be that image elements are normally perceived particularly well when they are positioned at the top right," Zwahlen gives an example. "But if they show a naked man at the bottom left, everyone looks there. So the designers' work is absolutely essential. But advertisers and companies should carry out a thorough analysis to find out whether their subject really works the way they want it to."
Everything is nothing without perceptionZwahlen sees both advertising agencies and advertising companies as potential customers for his services. Agencies should actually offer their clients not only their creative work, but also the monitoring of it. This would allow them to prove that they work better than their competitors. However, creatives cannot carry out this quality assurance themselves, because anyone involved in a project is blind to the business. An independent auditor would therefore be called for. In addition, the audit would give agencies security: if an advertising agency can prove that it has conveyed the message specified by a company as desired, it is in a safe position, even if the campaign falls flat. Take Rivella Yellow, for example: pretty much everything that can go wrong has gone wrong with this product. Despite many years of research, Rivella has developed a drink that is clearly not to the taste of the masses. What's more, most consumers didn't want to hear the "soy" message, even though the positioning as a stylish drink with a touch of the Far East fits in with the zeitgeist. The Reinhold agency, which implemented the campaign, was therefore dealt a very bad hand. The agency conveyed Rivella's objective - the fact that it was the wrong one is not its responsibility. Now Rivella has "refreshingly changed" the wrong product with the wrong positioning, adapted the taste and bottle layout and thus actually launched a new drink. Rivella could possibly have prevented this disaster with an early, accompanying test.
Like the creatives, corporate market researchers are also unable to provide useful data on the impact of their advertising media. "Market researchers aim for representative measurement results," explains Zwahlen. "They would have to interview 750 people just to test one subject, but usually several people are tested. That's unrealistic and takes far too long." But instead of involving an independent auditor, advertisers often do without a check. The result: "The probability of a film or poster not working is quickly 80 percent. A company that invests a million francs in a campaign and has a target achievement quotient of 20 percent has wasted 800,000 francs," Zwahlen calculates impressively. "It should be in the company's interest to maximize the target achievement rate. I can guarantee this with my test."
Zwahlen knows what he is talking about: shortly before the launch of the Federal Office of Public Health's (FOPH) Smoking Harm Campaign 2005, Bildkom checked the film spot that had been developed. It was intended to draw attention to the passive
smokers who cannot escape cigarette smoke, for example the waiters in a restaurant. The result of the check: "The message of the film was not optimally tailored to the target group," says Zwahlen. The agency responsible, Byheart, revised the film within a day. The next day, a commercial was available that clearly conveyed the FOPH's message (before and after: webcode a 3009bagbefore, 3009bagafter). "In most cases, something can be done at short notice, for example by adding or sharpening information," says Zwahlen. "Of course, it's better to carry out accompanying interim tests. Then you're never so far away from the desired end result that you have to start all over again."
Prevent stove firingBut how does the scientist manage to close the gap between the desired goal and the achieved goal with just one test? One example: If Henniez wants to convey the naturalness of its mineral water with images of the forest, but the consumer has mountains in mind when they hear the keyword "naturalness", then the positioning of the mineral water will not work. It follows from this: If you want to sell your product, you have to position it correctly. And if you want to position it correctly, you need to know how it is perceived or could be perceived. Finding all this out is the task of the company's market researchers. With this knowledge, they can develop an optimal positioning for their product, define the target group and determine the message of the campaign. Finally, the advertisers only have to implement this message one-to-one - which sounds simple, but rarely happens, according to Zwahlen. "Market researchers want to present positive and constructive results to their management, provide new impetus and point out all possible lines of thought. They are researchers, not controllers. The advertising agency can then pick one out of a sea of possibilities and visualize it, more or less at random." If the agency is lucky, it picks the right message and the subject works. Most of the time, however, it is unlucky - see the target achievement rate of 20 percent. "In order for what was planned to come out, certain work steps would have to be followed," says Zwahlen. "It has to be defined: This is our competition, this is our target audience, this is our product, this is our desired positioning: mountains or forest? If all of this is clear and the process steps run smoothly, all you have to do at the end is check whether the campaign clearly conveys the set objectives. Then you know whether the subject works or is doomed to failure." Zwahlen can use his test to check whether the technical process, from the positioning of a product to the definition of the target group and message to the visualization, is running or has run smoothly. "You can also use our methods to check a communication strategy or a positioning concept," emphasizes the scientist. "However, Bildkom is usually only called in as soon as the advertising agency starts working. We then accompany the process until the campaign is launched. It's not a bad thing if mistakes happen during the work. But you have to correct them."
The makings of a revolutionZwahlen's test is available in three versions: first vision, deep vision and multi-vision. The First Vision test is a quick test for almost all questions. Due to its simplicity, it is well suited to accompany the design process. The deep vision test is an extension of the first vision test. It provides reasons as to why something works or not. The multi-vision test is a special test that Zwahlen recommends using for the assessment of many subjects. All tests offer genuine cross-media testing. They provide a comprehensive analysis of perception, understanding and reaction to one or more subjects as a basis for a long-term positioning and image strategy as well as key figures for controlling. As an additional service, you can also book the target vision, which offers evaluations, key figures and additional analyses for communication controlling or communication evaluation. Bildkom prepares the test results in detailed, easy-to-understand documentation in French, German, English, Spanish or Italian as required. Bildkom requires between one and three weeks for all three test procedures. The costs are significantly lower than for the usual pre- and post-tests: the first vision test of a print subject, for example, costs 3,000 francs, the check of a film 3,500 francs.
Old theme, new ideaThe advertising world seems to have been waiting for Zwahlen's test: The desire to be able to make clear statements about the impact of advertising measures is almost as old as advertising itself. Nevertheless, Zwahlen's approach is completely new. This surprises the scientist himself: "I asked myself what has worked in the last 50 years and what has not. To my amazement, advertising impact cannot be measured yet. Yet we have seven million advertising experts in Switzerland, all of whom offer advertising effectiveness research."
Zwahlen's test software has already proven itself in many small, medium-sized and large companies. Nevertheless, the scientist is only now launching his service to the general public. "I just wanted to be sure that everything was mature and elaborate," he explains. "Presenting initial ideas for a new test with a PowerPoint presentation in hand is not my style." Zwahlen has a tongue-in-cheek explanation as to why no one else has yet come up with the idea of finally developing a valid instrument for measuring advertising impact: "Because no one else is crazy enough to invest ten years of their life developing something in their closet without knowing exactly what the result will be - driven only by the question: why is this subject good, why is that one not good, why does this one achieve its goal and that one doesn't?" If the research method delivers what Zwahlen promises, it will be a real hit - with the potential to revolutionize the advertising world.
Anne-Friederike Wilhelm
In a nutshellBildkom was founded in 2001 by Patrick Zwahlen. In addition to the managing director, the Bildkom team consists of three consultants and 40 employees who use the new test procedure to survey and evaluate the impact of advertising subjects. Bildkom's existing customer base includes national and international companies, and for online surveys Bildkom cooperates with large institutes such as Research Now. Bildkom makes its method book and the documentation of its test method publicly available:
> Method book> Test system