I recently received an email titled “Hello from the North Pole (well almost).” It was from Dan Landin, a planner with Stockholm agency Åkestam Holst. In his email, Dan asked if I thought people could really report their reactions to a piece of advertising. Because I had just spent a week abroad, discussing that very point in both Denmark and the U.K., I didn’t have any trouble formulating my answer, which I would now like share publicy: Yes people can report their emotions (though they may need a little help).
Here is Dan’s question in full:
Is it really possible for people in a pre-test situation to understand and accurately report on their emotions in relation to a piece of advertising? Considering the fact that most advertising is aimed first and foremost at influencing people emotionally rather than conveying information or a “message,” this strikes me as a crucial question. Do people actually know how they react emotionally to a TV spot? Can they tell us? And do their accounts of their reactions have any real bearing on their future purchase decisions?
Why am I so confident that the answer is yes?
Because if the emotion you feel watching an ad is strong enough to leave a lasting impression in memory then - de facto - that emotion is accessible for introspection. As I have mentioned in previous posts, you may need a little help - like words on a card - to identify the exact emotion you felt, but you will be able to do it.
Emotion is fundamentally important to advertising for two reasons. The first reason is the one that everyone focuses on, while the second is the one that many conveniently ignore.
The first reason has to do with one of our survival mechanisms, our “attentioning” system. This system, which functions at a non-conscious level, causes our attention to be directed by the emotional potential of what is happening around us. Advertisers, focusing on the importance of emotion to this system, understandably try to trigger a reaction with evocative content.
However, if an emotional response is to enter long-term memory and have a lasting effect, we must have some conscious appreciation of it when we experience it. Even if the response triggered by the attentioning system is an instinctive one, like jumping out of the way of a falling object, we are still able to reflect on how we felt at the time (scared, then relieved). The event and the feelings associated with it are now encoded in memory to help us recognize and react to similar events in the future. This is the second reason why emotion is important to advertising, the one that is too often overlooked.
So yes, the use of appropriate cues in advertising - visual or auditory - attracts attention, but the story does not end there. If the advertising is going to evoke a response that will last longer than a few seconds, it must cause a feeling related to the brand to be planted in our memories. If that occurs, then when we think about the brand at a later date, that memory will help to shape our predisposition toward the brand. The fact that the feeling has been encoded in memory means that the feeling is available for introspection, and thus respondents can accurately report it.
Many pundits confuse the issue by equating conscious thought with rational thought. However, the fact is that we are conscious of both thoughts and feelings. If we were not conscious of our feelings, they could not affect our future behavior. And advertising works by affecting future behavior, i.e., making people want to buy a brand.
One last point - while emotional responses are available to recall and report, we do sometimes need to help people voice their emotions. On an unaided basis, it can be tough for people to find the right words to describe how they feel. But with the appropriate prompts, they can readily identify their emotional response to an ad. This is just one of the reasons why I believe pre-testing is valuable. It gives us insight into how the target audience really feel about an ad or campaign. (Click here to see my presentation from the Admap conference on that topic.)
So now, over to you. How would you answer Dan? Is your understanding of how advertising works different from mine? If so, how?
(5 votes, average: 4.2 out of 5)
Loading ...
Email This Post
October 10th, 2008 at 9:30 pm
don’t disagree on the fundamental aspects Nigel
but how do we know in the aided conditions that we aren’t planting suggestions for an emotion that otherwise wouldn’t have formed or if formed - would have lasted.
Not unlike the tree falling in the forest with no one around to hear it question, if we need to coax the naming of the emotion - does it count?
cheers
Miro
PS. ready for tougher math questions on the spam protection quiz
October 11th, 2008 at 4:01 am
The biological function (and therefor the way it works for advertising) is to bring something in the environment to our attention.
If we do not know we had an emotion, and what it was, then we would not survive.
We can all describe the emotion we had, we might just use different words. Prompting alows us to normalise over the sample.
October 12th, 2008 at 5:43 pm
Similar to Miro’s point above, I take on board your points but I have a few lingering doubts
1) Between conscious and unconscious there is subconscious - would this not be able to affect our judgements?
2) Even if we could pick out subconscious effects along with the conscious, can we rationalise them in the form of a survey structure? Conscious does not equal rational as you point out, but we need to formalise the thought in order for it to be recorded
But your post has certainly given me food for thought
Simon
PS Congrats on the book launch/world tour!
October 13th, 2008 at 6:22 am
[...] Hollis of Millward Brown defends the rational AIDA approach of the Link Test on his blog. He raises some valid points, but I remain slightly sceptical. Saying that, I cannot [...]
October 13th, 2008 at 8:03 am
what about brands that we have engaged positively with since a young age? that kind of nostalgic feeling that brands will go to great extremes to acquire, were these formalized at the time?
October 13th, 2008 at 12:37 pm
Let’s talk about the second one, the feeling related to the brand as a result of exposure to a stimulus. Our research (at your sister company Added Value) has shown us that the yang is often consciously noted but seldom the yin. Here’s an example of what often happens.
A car manufacturer created a safety ad, and, in testing, everybody agreed that the ad provided reassurance and confidence, a very rational interpretation of the cues. The company ran the ad with disappointing results. We were asked to look behind the obvious using our indirect questioning technique. We saw that the ad had associated a heightened level of fear with the product at the unconscious level. The problem is that, in order for the concept of safety to be meaningful, our brains need to consider the prospect of danger. Of course they do this subconsciously, nobody visualized the product as being a source of danger, and yet the indirect questioning showed a significant rise in fear. The anticipation of fear, a really powerful emotion, seemed to outweigh the upside of the safety message. The client was intrigued by this unexpected finding and so decided to investigate further using fMRI. The fMRI revealed that final beauty shot of the product coincided with a sharp rise in the threat response. The correspondence between the emotions measured by fMRI and the indirect questioning was close across the board and substantially different from the reported emotions.
Here’s another example of observing the yang and missing the yin. A women’s clothing store was showing off it’s line-up in a TV ad. Post exposure, in response to direct questioning, viewers agreed that the clothes were pretty and that they would enjoy wearing them. But the indirect questioning showed that a big chunk of the target was unconsciously associating a sense of discomfort and anticipated low levels of acceptance wearing the clothes. The ad was effectively scaring off customers. In fact the store ended-up having to undergo major repositioning. The interpretation of the finding was that the contemplation of wearing such pretty clothes succeeded in raising all sorts of subconscious fears about the ability to look so good. The idea of looking great can only be relevant in the context of the alternative.
Just as in the car safety ad, the reported emotions showed us the rational yang without quantifying the depth of the unavoidable yin. We have since measured loads of cases where the reported emotions all seem totally sensible in relation to the ad, and yet the subconscious emotions reveal a different and often a more powerful picture.
Best wishes.
October 14th, 2008 at 9:35 am
Nigel,
Thanks so much for picking up my question in a blog post! I really appreciate it.
Personally, I don’t yet have a complete understanding of how advertising works (will I ever?). All I’m doing is seeking the truth. I’m adding pieces to the puzzle day by day and the one thing I know for sure is this: it’s complex.
In part, you answered my question right at the beginning by stating that people often need help to verbalize the true nature of their emotions. Through words on a card or using other techniques. More often that not, people doing pre-testing (or tracking, for that matter) don’t even get that part.
But the two things that, to me, really stand out in your reply are these:
1. “However, if an emotional response is to enter long-term memory and have a lasting effect, we must have some conscious appreciation of it when we experience it.”
2. “If we were not conscious of our feelings, they could not affect our future behavior.”
Is this really true? Intuitively, I’m not so sure. But it doesn’t really matter how I feel about it - it will need some further looking into on my part. The interplay between emotions, memory, consciousness and behavior truly is a fascinating one. And when you add to the mix the fact that advertising per se is of very limited interest to people (as opposed to, for instance, being attacked by a snake), you get an even more interesting cocktail.
So I’ll dig a little deeper and try to get my head around this and then get back to you with my toughts as soon as time allows. I hope that’s ok.
Thanks again!
Dan
October 14th, 2008 at 1:58 pm
Some great issues being raised here, I want to pick up several: 1. Emotion isn’t unconscious, and consciousness isn’t rational. Some of the physiological aspects of emotional responses response aren’t conscious, but our experience of emotions, our feelings, ARE. Consider the last time you had an argument with your partner: Were you being rational? Probably not. Were you aware you were angry? Almost certainly… and you could tell other people that you were angry too.
2. Implicit methods do add value – but not always. ‘Implicit’ associations are often defined by omission – what people didn’t say. But if we ask the right questions, we might bring out more. My experience with the many different forms of implicit approaches that we have experimented with at Millward Brown is that the times that they do ad value are: Dealing with sensitive material – this is when people are likely to edit their responses and give the answer they thing we want to hear, or which they want to hear themselves say Picking up transient effects that may happen too quickly or be too small a part of the whole for people to report to us. Respondents tend to answer explicit questions about ad communication based on their response to an ad as whole, not the specific elements that may have contributed to that overall impression.
3. We need to combine explicit and implicit associations to arrive at a full picture of response to an ad or brand. Explicit, unaided questions like reveal the ideas the ad or brand may evoke, which will readily come to mind when people think about it later. Prompted questions help people articulate difficult ideas, but also show what the ad may come to mean for the brand if people are given pause for thought. Implicit methods can reveal the ideas that have been planted that weren’t explicitly said or demonstrated. These may be harder to access but are still part of the raw materials the brain uses later when making judgements. But neither are they the ‘real’ response, because BOTH cognitive and emotional processes are used when people make decisions. We need to use all three methods, especially when an ad is taking a less than explicit approach to communication.
For instance, to take an ad that Nigel has blogged on before: the Skoda Fabia ‘Car Bakers’ spot. Unaided questions revealed limited explicit communication – mainly that the ad was clever and the different, and it evoked VERY powerful positive emotions. Prompted questions revealed good communication of extensive features. The implicit measures we used (which use reaction time rather than direct question to pick up implicit associations) revealed a strong association with the idea of femininity, and to a lesser extent, family. Which is the ‘real’ response? All of them – but some will need some coaxing to affect decisions. People walking into a dealership will feel good about the car, but might see it more as a family or women’s car. Certainly a dealer who pitched it as a family workhorse might be well received. Likewise the dealer who talking about he spec-list, will find it’s seen as credible. I think what the combination of measures shows is both what’s top of mind, but also what’s also there latently, to be picked up by other media or elements of the mix.
October 14th, 2008 at 3:32 pm
As Dan points out this is a complex but fascinating topic. Maybe that is why this post is attracting some of the longest comments on the blog!
Thanks to Graham for chiming in on this one. He has been actively led Millward Brown’s exploration of biometrics and assessment of implicit communication and how best to measure it. And that is a point worth clarifying, Graham the idea of femininity was implict in the ad, but the emotional response to it was not. The ad scored very strongly on our emotional response metrics.
At the risk of over-simplifying I think the key issue here is whether or not we correctly anticipate how people will react to the ad. If we think to ask the right questions ahead of time we can elicit the explicit and implicit communicattion. If we don’t do this we risk missing out on what is not readily accessible in people’s mind. The Link pre-test includes a number of standard questions that use sets of words to assess engagement and emotional response. It is useful to have a standard set of words when comparing across ads but if the ad is intended to elicit a specific response not on the list we should add additional sets of words.
Now to pick up on specific points made by other commentators.
Simon, glad to have given pause for thought. What about the subconscious? Most things are subconscious, i.e. we can only consciously process a few ideas at a time, but they are still accessible with the right questionning.
Seth, I think nostalgia is the an emotion produced by the act of thinking back on something that you felt good about previously. In other words, we don’t feel nostalgic about using the iPhone today but we might do in 20 years time when it is long gone and our phones are implanted in our skulls!
Mark, thank you so much for commenting on this issue. We don’t often get people from other WPP companies joining the debate. However, at the risk of losing a commentator, is what you do really that different? Don’t you still use a set of words to assess people’s reaction to the brand on its own and then the brand as portrayed in the ad and then look at the delta? People still need to be able to tell you how they felt. Or am I out of date with the Added Value technique? Maybe you are using something similar to the reaction time test Graham mentions?
October 15th, 2008 at 10:24 am
Mark,
what were the “disappointing” results of the car ad?
The problem is that, in order for the concept of safety to be meaningful, our brains need to consider the prospect of danger. Of course they do this subconsciously, nobody visualized the product as being a source of danger, and yet the indirect questioning showed a significant rise in fear.”
by the scenario you outlined above, Volvo would never have gotten off the ground and yet…
are there other reasons/factors at play or are we simply offloading on the ad to find some post rationalization?
Does the same mechanism work in reverse with ‘dangerous’ brands?
October 15th, 2008 at 11:57 am
Nigel, in response to your question, first, to be absolutely clear, I was talking about the ability of an ad to exert pressure on how people anticipate feeling using the brand; this is discrete from measuring emotions evoked while watching an ad. We concentrate on this because we believe that all human choice is ultimately determined by how we expect to feel. Ads like the iPod launch showed a bunch of cool and exciting product features in a fairly non-emotional way. Many of us responded by anticipating the fun and joy that we would get from using such a device and that is why we lined up around the block to get one (well, not me personally). A very emotional response to a largely rational ad.
Our methodology goes like this. We ask people to imagine a situation in their lives where they would be using the product. We then have them elaborate on this situation, with questions about where they are, who they are with, when is this happening, and so on. Once they have constructed a detailed story we ask them, using a fixed battery of highly tested and very simple scales, how they are feeling in this imagined situation. We then compare the exposed group to a non-exposed group (or to other ads or other brands) to see if we get a significantly different set anticipated emotions.
Without making too long a point of it, it is perhaps unfortunately true that human discourse requires us to make sense, both to ourselves and to the receiver. So we will always have a tendency to recognize and acknowledge feelings that make sense above feelings that might suggest we are heading towards insanity. If I were to ask you directly how you think you might feel driving the new Porsche 911, you might have a tendency to say things like thrilled, excited, a bit scared, and so on. But when we ask people to imagine a situation in their lives where they might be driving the car, we get all sorts of things. Being made fun of by co-workers or impressing valet parkers at restaurants, are just a couple that frequently appeared. For this particular product, given its obvious cues, people are far less conscious of the emotion of acceptance than they are of joy or anticipation.
As I said earlier, it is the expectation of emotion that determines our choices. In choosing a restaurant to dine at this evening, you must run a little micro-second movie in your head as to what it would be like to go to one restaurant over another. Maybe you see one as being crowded and noisy and so you anticipate feeling irritated and uncomfortable. So you rapidly substitute another which seems quiet and peaceful and you can visualize yourself relaxing with a glass of wine and quite conversation. When we tested a restaurant ad recently, we found that it clearly communicated a sense of family, good food, and friendly warmth. But to one group of people, the young, rebel, go-my-own way crowd, this engendered a whole battery of negative emotions in their imagined situation.
The point of all this is that if I were to ask you a direct question like “how do you imagine yourself feeling using brand X” I will often get a different answer from “how are you feeling in this imagined situation”. The reason for this is that the second version isn’t necessarily going to be a rational extrapolation from obvious brand cues. If the subject’s story is about being laughed at by their colleagues then it is totally sensible for them to say they are feeling awkward and rejected. So, if we detect significantly more stories that evoke awkwardness and rejection among our exposed group, we would conclude that there is something in the ad causing this.
It’s interesting that we often find it really difficult to express how we are feeling. “Are you happy?” can be a horrendously complicated and difficult question to answer. Psychotherapists, EEG and fMRI owners have made good capital out of this problem. And yet, we don’t seem to have nearly as much difficulty when we remember events or anticipate events. A question like “were you happy when we went to the beach last summer?” seems much easier to answer, as does “do you think you would feel happy if we were to go to the beach next summer?”. It seems like our brains have a way of simplifying the somatic marker when attaching it to our event memories. But then it kind of makes sense. Human survival depends on remembering or learning how an event is going to make us feel. But it doesn’t depend on us actually being aware of how we are feeling in-the-now. We are conditioned to believing that memories can only be created by conscious experience, but it is not impossible that our brains are programmed to remember our body state, our emotions, even when we are not consciously aware of them. Have you ever had that experience where you are looking forward to eating this wonderful meal only to find that you have ingested the whole thing unnoticed while having a heated discussion about the existence of god? And yet a few days later you will swear blind how great the food was.
Best wishes.
October 16th, 2008 at 5:36 am
It seems to me that Added value and Millward Brown are methodoligcally but not ideologically different.
I agree with Graham’s 2 conditions for implicit measures to add value but would probably add a third - “Implicit measures (techniques) can be useful when consumers are not sufficiently motivated to provide introspective responses”. The added-value method may provide more accurate responses in some cases as it engages the consumer for greater introspection. Under the Millward Brown survey method, consumers may become misers dispensing emotions based on norms/expectations etc. The differences between the two systems may be observed in something like the car ad Mark mentions but may be a non-issue in other cases. As said though, I see no fundamental problem with the Millward Brown technique just a challenge of providing the right stimuli to elicit in-depth responses.
Would be fascinating to see some side-by-side tests of the two techniques…
October 16th, 2008 at 10:04 am
Mark, thanks for the clarification.
To some degree I think we are comparing apples and oranges. Our focus tends to be on understanding the role of emotions at the time of viewing in order to help produce a more effective execution. Yours focuses on how the ad will make people feel about the brand. To Chris’s point, both assume that people can report how they feel, or think they will feel, accurately.
I agree that what motivates people is the expectation of how something will make them feel when experiencing the brand. Because of the way you ask the question, i.e. setting people up to consider how they will feel, I can imagine that you will get useful additional insight in some cases. But shouldn’t that have been learnt at an earlier stage of development?
And if I am honest I think that the methodology used may be overkill - making things more complicated and costly. It seems to reflect the same mindset as other pre-tests which literally set out to test people’s reactions to an ad. As Millward Brown demonstrated with persuasion it is perfectly possible to ask people a direct, introspective question and get the same result that one would have done with a test/control methodology. What’s more we’ve repeatedly proved it is predictive of in-market results. Maybe something to consider?
October 17th, 2008 at 1:34 am
Miro, I’m guessing that Volvo made the benefits of safety stick better than the threat of danger, unlike this ad. We had strong mirroring of the people in the vehicle, the people in danger. One theory was to not have actual people being threatened. Volvo used dummies. I saw a Honda ad recently that showed the sophistication of their safety testing facility, again another way to do it without actual people.
Nigel, it has been my experience that clients and agencies have a belief that ad testing has a tendency to kill good ideas. They maintain that our industry has always been guilty of favoring ads with clear, concrete, tangible arguments of superiority compared to ads “with an emotional tone”. A great example of why they might be right is the Corona beer brand. Corona has never bragged about the quality or freshness of its ingredients, the precision of the brewing process, the care taken during transportation, a great taste, or any other concrete reason why we should buy their product. As far as I know, Corona is the number one imported beer in a host of countries. Somehow Corona has managed to attach the “coolness” of its ads to its product. And yet it makes no sense for any of us to believe that the consumption of this beer will transport us to a sun drenched beach where we are making foghorn noises into a bottle. Further most subjects would like to establish that they aren’t so easily manipulated. So a test needs to bypasses logical processing of ad content, however introspective, and quantify the degree to which the emotional payoff of consuming the product has changed.
If your prediction of in-market results doesn’t have an R-squared of one, perhaps there is room for extending the armory?
Chris, we have done some side-by-side testing using our classic diagnostic system which I believe is a little closer to Millward Brown’s system. You are absolutely right in that sometimes both systems easily lead to a similar set of conclusions. But I disagree that its a question of finding anything “in-depth”. People are absolutely capable of telling you how they are feeling in an imagined situation; that’s what they do all day, every time they make a decision. I have no idea why I sometimes choose a cup of tea over a cup of coffee and I’m not sure what kind of in-depth probing would get me to work that out (after all I’m not a psychologist and absolutely no good at self-analysis). But if people consistently imagine tea situations in which they see themselves as feeling more optimistic than coffee situations (massive oversimplification going on here), then I can draw a conclusion (although clearly no explanation as to why tea should be appropriate for optimism).
Best wishes.
October 17th, 2008 at 3:23 pm
Seth, I think nostalgia is the an emotion produced by the act of thinking back on something that you felt good about previously. In other words, we don’t feel nostalgic about using the iPhone today but we might do in 20 years time when it is long gone and our phones are implanted in our skulls!
Nostalgia definitely contains the positive emotional reactions from the past, although often combined with a sense of sadness or regret about the loss of the past. The original emotional reaction may also have involved negative emotions, but they tend to be filtered in the case of nostalgia and dominated by memory of the positive emotions. Unless the object had clear emotional associations originally (whether directly related to the object or related to its context), it is not likely to evoke nostalgia at a later point in time.
October 21st, 2008 at 5:23 am
Dan questions whether we need to be concious of emotions for them to have a long term effect.
I don’t know.
but Damasio has looked into this extensively and writes:
“the full and lasting impact of feelings requires consciousness,” “consciousness must be present if feelings are to influence the subject having them beyond the immediate here and now.” This is why he argues that, “when consciousness is available, feelings have their maximum impact,” and “the conscious component extends the reach and efficacy of the non conscious system.”
(All these quotes are from his book “The Feeling of What Happens”).