Brands and Emotions

Monday 10:30 a.m: An emergency meeting in the Marketing Department. "The market performance of our brand X is continuing to deteriorate. For this reason, we need to relaunch it--the brand needs to be made sportier and more emotional. The agency is to develop a campaign that is suitable for this purpose. The best solution would be to use footballer Y, he's scoring again." A typical challenge for brand managers and agencies in this age of complex brand management, where it is becoming even more difficult to differentiate brands via objective product benefits.

This task is easy to define, difficult to execute, and even more difficult to implement successfully. It is for this reason that the band manager needs to know if the campaign has actually been successful in "emotionalizing" the brand. And, ideally, determine whether the planned campaign will achieve the desired success: Does the campaign add emotions to the brand? Will this improve its market performance?

The brand manager directs these questions to the market research institute. In order to measure the emotionalization of the brand, advertising research needs to be able to deal with two challenges: it must be able to measure the effect of the communications on the brand and it must be able to find a valid and practical method of measuring emotions.

Is it possible to measure the effects of communications on brands?

"The results of our pre-test have shown that, after exposure to advertising, 61% of respondents found brand X to be sporty, while 80% associated the brand with positive emotions. Both of these conclusions are clear from our database of more than 100 tests. Therefore, the campaign can be expected to have the desired success." The brand manager and the agency receiving this happy news are ecstatic. The campaign proposal will be able to achieve the desired results.

But does an above-average result of 80% of respondents who associate positive emotions with a brand mean that this campaign would add positive emotional value to a brand? The only thing that can really be concluded from these results is that, measured against the database, after respondent exposure to advertising, this brand is associated with more positive emotions than others are. Yet this was not the question posed by the brand manager. The key question was "How would this communication influence/change the positioning of the brand? What difference would this advertising make for this brand?"

In order to measure the difference, it is necessary to measure how the brand is positioned before or without advertising contact. It is possible that 80% or more of the respondents would associate positive emotions with the brand even without advertising. In this case, the effect of the advertising would be null or even negative. This is why pre-tests, which only measure the brand status after exposure to advertising, are in principle unsuited to determine the effect of a campaign on the brand, regardless of whether the database contains 100 or even 1,000 tests. At best, they are a tranquillizer for brand management and for the agency; at worst, the unreliable interpretation of the 80% result could lead to the brand manager making the wrong decision

The individual base level of a brand

How then is the effect of communications on a brand measured? Not only is the solution simple, but it has been in practice globally in empirical social research for decades: the experiment. To do this, it is necessary to compare a group with exposure to advertising versus a group without advertising exposure. The difference between the two groups is the difference that is made by the advertising.

The fact that the experiment is considered a silver bullet in empirical research is due to its unique selling proposition: only with the help of an experiment is it possible to measure the exact effect of a stimulus. This is the case for all stimuli, whether for the effect of a medication as well as for the effect of particular music on moods. Naturally, this is also the case for the effect of a campaign on the positioning of a brand.

Therefore, it is particularly surprising that experimental design is still uncommon in advertising pre-tests, at least when it comes to the positioning of a brand. In fact, it is actually easier to carry out experiments in the context of pre-tests than it is in post-tests or tracking. One could even say that many pre-tests do not meet the standards of empirical research in this regard. Why is this so?

There are two types of experimental designs. One is the pre-post design, where the same person is queried both before and after the advertising exposure. The other is the test cell versus control cell design, where two groups of people are questioned, one with advertising exposure and the other without advertising exposure.

With the pre-post design, behavioral changes are often measured (e.g. with a lottery where the winner gets a gift basket and where respondents need to choose between various products). Why would it not be possible to measure positioning changes as per the pre-post design? There are generally two reasons for not doing this. One is that the cut-through potential of the ad is measured in pre-tests, for example, when respondents are told to watch a television program with advertising breaks. In such cases the respondents is not alerted to the brand that is the subject of the test. If respondents were alerted to an advertising test by being prompted by several questions regarding the brand prior to seeing the program, this would essentially defeat the purpose of the disguised test design and cut-through potential could no longer be measured in a valid manner. Secondly, during the pre-tests, a number of open questions are usually posed for diagnostic purposes. If one were to pre-empt brand characteristics--in the form of an item battery, for example--the open-ended questions would be influenced as a result. (The respondents often repeat things they were told before.)

As a result, the test cell versus control cell approach is the only one that is suitable, and it is ideal for measuring brand positioning effects as well. Naturally, it is necessary to pay close attention to the basic rules of the experiment--both random samples need to have the same structure, for example. Why is this method still rather uncommon in pre-tests? One reason might be the higher costs that it involves, which results from the increase in number of interviews. Even so, the amount of the extra investment required should not be a serious problem because the duration of the interviews in the control cell is kept much shorter than in the test cell. In addition, only one control cell is required no matter how many ads are being tested for a particular brand. Brand managers who want to find out if a campaign is actually suitable for emotionalizing a brand cannot afford to avoid the extra investment considered for an experimental design, because there is no sure method without this "silver bullet" in empirical research.

How are emotions to be measured in a valid and practical manner?

Now that we have answered the question of whether it is possible to measure the effects to the brand, it is now time to confront the challenge of measuring emotions. More than anything, the brand manager needs to know how the campaign will influence the emotional experience of the brand (as with the experimental design, it is here necessary to measure the emotional experience with and without advertising contact). They are also interested in finding out which emotional reactions are triggered by the commercial. How can one measure emotions? "Emotions are understood as a current state with a specific quality and activity, directed towards an object, linked with a characteristic experience," according to Prof. Dr. Joachim Funke.

The typical response of market research institutes to take a list of statements or adjectives and present the results as an octagon or calculate a score. Is this a suitable method for measuring emotions? Can a brand manager expect that emotions will be measured in a valid manner and that their questions will be answered correctly? Let us look at how this is done by the experts in emotional psychology.

Emotional psychology recognized early on that emotions are expressed using characteristic expressions and gestures. Facial expressions in particular are largely universal for specific emotions, i.e. are the same for different people and cultures irrespective of national boundaries. The measurement of expressions is, therefore, considered to be the form of measurement which offers the greatest chance of success in empirical emotional psychology (according to Keltner and Ekman in the "Handbook of Emotions").

Biophysiological indicators offer information about an individual's excitement, but they are hardly suited to determining the specific quality of the emotion in question. Verbalization implicitly involves rationalization and is, therefore, less suitable for the measurement of emotions. It is also difficult or even impossible for the respondent to express their feelings without assistance. Thus using expressive behavior offers additional advantages.

In light of the findings of emotional psychology, it is surprising that the verbalization procedure is often the sole method used to measure emotions in communications research. This is most likely due to two factors. Historically, many of the established pre-test institutes have their roots in an age when marketing objectives were primarily concerned with rational product characteristics that were perfectly suited to verbalization. Another reason is the fact that the primary method by which data is collected during the pre-tests is through surveys. This meant that it was also quite practical to measure emotions using verbalization.

The proven measurement procedures from emotional psychology are not used in advertising research because they are rarely practical. Neither trained observers nor recording using elaborate apparatuses are currently financially feasible as standard solutions. In order to make the measurement of emotions using expressive behavior practical for market research, Ipsos ASI has taken a new approach. The respondents are asked to locate their emotions on a map of emotions by assigning them to the corresponding emoticons. These visualizations are supplemented with adjectives. Once the emoticons are structured and localized in accordance with their similarities, the result is a map of emotions, rather an "Emoti*Scape(TM)" (Figure 2).

View larger image

This new instrument makes it possible for the respondents to express their feelings by utilizing the universal expressive behaviors for specific emotions in a manner that corresponds to everyday life, without having to put their own feelings on display, by the simple process of assigning them to the corresponding emoticons (which is not overly demanding), and without falling into the rationalization trap by measuring emotions using verbalization. In addition, this method requires less expenditure than the measurement procedures commonly used in emotional psychology.

The analysis of the emotional effects is performed both with regard to the specific emotions as well as in the positive/negative and active/passive dimensions. The success of the advertising is examined in two important stages of the advertising effectiveness process: 1. Which emotions are triggered by this advertising? 2. Which emotions are experienced in the utilization of this brand?

Relating to this very subject, Ipsos ASI carried out a case study before the World Cup in Germany this year. The study involved determining the emotional effects of soccer stars in advertising (Figure 3).

The results clearly demonstrated that it is not possible to draw conclusions about the emotional effect of a commercial based on the emotional positioning of a brand after exposure to the ad. The results in the test cell taken alone offer no information on the effect. In order to measure the effect of a commercial, it is necessary to know the base level, something that is measured in the control cell. It is only the experimental design that allows the brand manager's essential question to be answered: What difference will the campaign make for the emotional experience of my brand?

Emotionalization = Market Success?

Being able to measure emotions is good, but at the end of the day, it is the market performance that is decisive for the brand manager. What would be the return on investment for this campaign? This is why the classic measures of effectiveness retain their value: the cut-through potential of the ad, its linkage to the brand, as the prerequisite for any effect, as well as the effect of the campaign on the brand's equity and on purchase behavior (persuasion).

The chances of increasing market performance though adding positive emotions to a brand are not bad. The initial analyses of the new Emoti*Scape(TM) database confirm the relevance of the emotional dimension for advertising success.

This article has translated from German and reprinted with the permission of Markenartikel. For more details, go to www.new-business.de.