As I explained in the preceding chapters, the fact that people can post-rationalize their behavior, or will answer a question on the basis of what they believe they think, doesn’t mean that they do so accurately in terms of the way they will subsequently behave. Beyond that, taking a person out of the environment in which they make judgments about consumption creates an even greater risk that, however well intentioned they are, their responses won’t reflect how they will think and act when those influences are present. As I discussed in the last chapter, these problems are compounded when the process of questioning alters what people think and say.
Most research is preoccupied with getting people to answer: to say something, anything, that can be analyzed. The barometer of validity that tends to be applied is whether the same response is heard consistently. This may very well indicate nothing more than that people respond in the same way because the process of research was more or less the same, rather than because the repeated responses reflect some underlying truth.
The best time for asking questions is when the behavior of the person being questioned has been observed. That way, any claims made can at least be compared and contrasted with what was seen and, to an extent, validated accordingly. The best place to ask a question is when the respondent is as close as possible to the environmental and contextual elements that influenced their behavior; unlike with research conducted in any other location, the only additional sources of inadvertent influence are then the questions and the person asking them.
It is also advantageous to be asking the questions relatively soon after the consumer choice or experience has taken place. Given our capacity for consciously rationalizing a nonconflicting, positively embellished perspective of the things we’ve found ourselves doing, the longer we have to construct an apparently sensible rationale for our actions, the greater the likelihood that we will do so.
When considering emotional responses, such as how someone feels about a brand or an advertisement, there is a strong argument for only paying significant attention to their instantaneous reaction. The longer people have to involve their conscious mind, the more likely they will be to adapt that reaction to one that is influenced by social factors that would ordinarily not be involved, such as who else is present and how they would like to be perceived by other people. Think of it as the difference between that moment when someone makes a tremendous belch and is really quite pleased with the sound they’ve produced, and the moment they remember they should be embarrassed at something generally considered socially unacceptable.
A number of studies have shown that our unconscious mind’s response occurs some time before we reach a conscious conclusion about something. In addition to the decks of cards experiment mentioned in Chapter 1, Benjamin Libet and his colleagues scrutinized the brain and muscle activity of people asked to tap their finger at random and discovered that the conscious experience to move the finger happened a third of a second after the activity in the brain that initiated it.1 More recently, researchers in Berlin found that brain activity preceded the conscious awareness of selecting one of two buttons by as much as seven seconds.2
More evidence of the important link between speed of response and the unconscious mind can be found in the Implicit Association Test, developed by Greenwald, Banaji, and Nosek. It was devised to reveal the underlying unconscious associations that influence our beliefs and behavior and does so by asking participants to categorize words as quickly as possible and comparing reaction times; where the response is quicker the unconscious association is stronger.3
Of course, with traditional approaches to soliciting consumer views, the efficacy is associated with the depth of probing and the cost of the research is closely linked to the length of interview or discussion. But with the perils of introspection and post-rationalization, ironically the mechanism through which value and quality is implicitly judged is indicative of a barrier to accuracy.
In a nutshell, the longer the pause between the question and the response, the greater the likelihood that the conscious mind has intervened and wielded its duplicitous influence on what’s going to follow.
A lot of research sampling and recruitment screening concerns itself with asking the right person, but how do you ensure that you are then asking the right mindset of that person, and how do you do this if you can’t catch them just after the moment of interest?
Fortunately, while it’s impossible to force someone to shift into a different mindset against their will, it’s surprisingly easy to encourage them to do so. If you think about an emotional experience you’ve had in the past – a powerfully unhappy one will perhaps help make the point best – you may look back on it with a degree of sadness. As you reflect, you know that it was a sad event but you are anesthetized from it by the passage of time. Your unconscious mind has done its job in assimilating the feeling and returning you to the same level of happiness you were always at.4 However, if you start to recall what you did at the time – in other words, the behavioral experiences you had – the feelings you experienced start to return; perhaps not as powerfully as they did originally, but your basic frame of mind will shift to the one you occupied then and the emotions you experienced will return.
Even if you can’t take yourself into this mental state, you can easily observe the process in interviews with people who have had traumatic experiences. Recently I watched two documentaries about people who had been directly or indirectly caught up in the terrorist attacks on London in July 2005. As they relived their experiences of that day for the journalist, their emotional state shifted to the one they had occupied at the time. The wife who couldn’t get hold of her husband started to cry again; Susanna Pell, an extraordinary woman who had found the composure and bravery to walk into the bombed carriage and save lives, exhibited the resolve and calmness she had discovered in herself that day. In both of these cases the people concerned had no desire, derived no benefit, and made no conscious attempt to shift back into that mindset, it happened unconsciously.
These extreme cases reveal that people can be reversed into an unconscious mindset by a questioning process that causes them to reflect on their behavior, rather than post-rationalize it. In my experience this technique doesn’t only work with experiences involving significant emotions; people will also shift their frame of mind to relive more mundane experiences, particularly when they are relatively recent.
In his book The Feeling of What Happens, one of the world’s leading experts on neurophysiology, Antonio Damasio, reveals why emotions are so critical to how we process information:
In a typical emotion, then, certain regions of the brain, which are part of a largely preset neural system related to emotions, send commands to other regions of the brain and to most everywhere in the body proper. One route is the bloodstream, where the commands are sent in the form of chemical molecules that act on the receptors in the cells which constitute body tissues. The other route consists of neuron pathways and the commands along this route take the form of electrochemical signals which act on other neurons or on muscular fibers or on organs (such as the adrenal gland) which in turn can release chemicals of their own into the bloodstream.5
If the process reveals the mindset that a consumer was in at the time of interest, it is then possible, with an understanding of how Eric Berne’s transactional states interact with one another, to talk to respondents in a way that encourages them to stay in the frame of mind concerned and elicit further information about what they were thinking and feeling. When combined with observation of actual behavior, this approach provides a combination of factual evidence and psychoanalysis that can be particularly informative.
I have explored countless reasons why it is reckless and misleading for consumer research to work on the principle that consumers can tell us what they think and feel. I find it is invariably better to design investigations into consumers’ thoughts, feelings, and behaviors on the basis that they can’t tell us what they think. Whenever it becomes helpful or necessary to question consumers, it is essential that what they say is treated with enormous skepticism.
One way of assessing the accuracy of a response is to pay close attention to the language the person is using. Psychoanalysts are wary of people who say they will “try” to do something; as commitments go it’s not exactly a powerful statement of intent. Similarly, when respondents are talking about their attitude to a brand, product, service, or piece of marketing communication, their words may reveal that they are being socially considerate rather than entirely honest. When people drop pronouns such as “I” and “we” in association with an opinion, it suggests a personal affinity with what follows. In contrast, when they unconsciously distance themselves from a statement by saying something like “It’s nice” rather than “I like it,” there are grounds to be suspicious.6
It is also helpful to look for correlations between people’s behavior and their claimed attitudes or values. It’s easy for us all to make claims (based on the not-always-unhealthy mechanism of self-delusion) that we are striving for something positive when there may be significant clues that suggest we aren’t quite attaining the values to which we aspire. If someone sets out to tell a lie, then it takes a reasonable amount of practice and skill to establish a base line and detect the micro facial reactions, verbal pauses, and stress cues that can indicate an attempt at deception. However, when people are lying to themselves it is generally easier to identify. Indeed, the main challenge is lifting ourselves above the acquired level of social interaction that leads us, by and large, to accept what other people tell us is true, particularly if it’s relayed in a compelling way. It does feel as though you are doubting the person concerned (naturally enough, because you are) and is not necessarily going to result in an exchange that endears you to the other person; although in my experience people aren’t usually hostile when, through the process of asking the right questions, they realize that they have been misleading themselves too.
For example, in a project testing advertising, someone dismissed an ad and claimed that he wanted a more factual and informative advert from the company concerned. By distracting him with an irrelevant topic for a moment (so that he didn’t see the association) and then quickly asking him to state what his favorite advert was, I established that that ad was in no way factual or informative. Evidently he didn’t like the ad I had shown him, but had I advised my client that it should produce an infomercial based on comments like this one, I would have been doing the client a huge disservice.
Sometimes it’s not even necessary to ask a question. When someone says that they have been buying a product for years because they’re on a diet, the statement seems reasonable enough. But when that person is significantly overweight, it is apparent that, unless they started out at a size that would have made the national news, they aren’t shedding pounds at any discernable rate; it is likely that they are actually buying the product for another reason altogether.
Another way in which people’s conscious misinterpretations can be manifested is when they are inconsistent with other behaviors. For example, in one project looking at supermarket shopping, I spoke to several women who, despite taking the time to write a shopping list, routinely forgot to take it with them. They rationalized this as trivial forgetfulness rather than anything more duplicitous, but when I asked them how many times they forgot to take their money or to put clothes on, the answer was never. I deduced that there must be an unconscious basis for forgetting the lists and, by exploring their behavior on shopping trips, could start to explore why this really happened.
Another technique for getting behind the conscious veil people like to keep in front of themselves, particularly where you suspect that they are putting on a show for your benefit, is to switch the focus of the conversation from them to “other people.” On the basis that what people see in others reflects their own perspective, asking consumers what they think other people’s motives are can be enlightening. Customers who are unwilling to reveal their own confusion with a product display will often be happy to point out that “other people” would find it confusing. One word of warning though: it’s important to distinguish those responses that are the result of you having asked the respondent to represent the views of others from when they voluntarily do so. The latter can be a form of social politeness, for example when they think something is hopeless but they try to soften the blow by suggesting that someone else (who isn’t present) would think it was terrific.
Asking indirectly related questions can also help distinguish genuine motivations from positions adopted to be socially acceptable. Few people would admit that they choose a brand for reasons of elitism or snobbery, but, where you suspect this is the case, asking them if they would describe users of the brand as more successful or more intelligent than nonusers can be revealing. It’s important to play close attention to the total package of the response to such questions. In one case where I used this approach the woman replied “I wouldn’t say that,” while shifting her position into a more upright and assured pose, and placing a little more emphasis on the word “that.” She wouldn’t necessarily say that she was better than people who didn’t use the brand, but her reaction suggested that she was happy to think it.
In a project for a vegetarian food product, one pack design was being rejected out of hand by a subgroup of the target audience and favored by the others. There was something distinctive about the type of people involved, though: their behavior and style of dress suggested that they saw themselves as set apart from the others. Through asking them questions that were about other aspects of their lifestyle, it emerged that their vegetarianism was a reflection of a greater stance against mainstream culture and, as a result, they didn’t like the pack design because it made the product more commercially attractive. Being commercially attractive was exactly what my client wanted, and with the evidence I obtained the company could at least make an informed decision about what proportion of its existing customers it risked losing by changing to a pack that had the potential to appeal far more widely.
Taking a skeptical position on consumer post-rationalizations is eminently sensible. Recently, a client of mine described research he’d conducted suggesting that people leaving his company’s stores without purchasing were doing so because the store didn’t have the specific product they wanted. This simple and eminently reasonable explanation, captured from interviews with customers as they left the store, could easily have been accepted at face value. However, my client has a healthy skepticism when it comes to consumer research and commissioned a further study to ask people entering the store what they had come to buy. It transpired that most customers didn’t have a specific product in mind. It’s much easier to tell yourself that you haven’t bought something because the shop didn’t have what you wanted than it is to appreciate that the store environment didn’t make you want to spend longer, the salesperson failed to influence you in the right way, the product display confused you too much to be able to make a choice, or you were scared you’d buy the wrong thing (all much more likely reasons for not parting with your money).
For understandable reasons, most people who come into contact with market research regard leading questions as a potential source of bias rather than a tool for getting an accurate insight into the consumer mind. However, in the right context, leading questions have the capacity to provide more powerful and more accurate insights than balanced ones.
When behavioral evidence has been collected, even if it’s at a market level rather than for the individual concerned, building this knowledge into questions can have the effect of giving the respondent tacit permission to say something that they otherwise wouldn’t. For example, if the behavioral evidence shows that most people walk around part of a store looking confused, it can be much more effective to say to people “Most customers find this part of the store really confusing, why do you think that is?”, rather than to allow conscious vanity to present the respondent as a rare example of someone who could cope with the experience. In any event, if they haven’t experienced the problem their answer will tend to reflect this.
Similarly, rather than using a totally dispassionate and balanced style of questioning, it can be beneficial to adopt a position that reflects behavior. When interviewing people about financial services products, I have often found it more helpful to be empathetically confused about product details, rather than allow the respondent to feel stupid on their own because they don’t understand the technicalities of a product they will only encounter once or twice in their lives.
One of the most helpful uses of questions is to confirm and clarify what has been observed at a behavioral level. Very often this involves confirming what you believe consumers haven’t taken any notice of. For example, it is much more useful to put a piece of point-of-sale communication advertising – say, the availability of next-day delivery – into the environment in which it will have to operate and ask: “How soon could you get that delivered if you wanted it?” If people say “Tomorrow”, ask if that information is provided anywhere in the store. You will recall from the phenomenon of inattentional blindness that people are perfectly capable of not seeing something that they have looked at. From my own work I know that people can select from a list of a dozen television programs, having looked at the list for over ten seconds, and be incapable of recalling any program other than the one they selected once the list has been removed. One of the ways I confirm this is by challenging people to name what’s on a given channel to win £100. Not only can they not do it, but on occasions they will claim that a program guide from which they have chosen thousands of times doesn’t have information about the channel each program is on, despite its always having been there.
Conversely, just because something hasn’t been consciously noticed doesn’t mean that it hasn’t influenced a consumer. Issues surrounding quality of finish of the materials used in the environment, packaging design, and the nature of surrounding products can all influence consumers’ unconscious perceptions. This is why having the true behavioral measure of observed consumer activity is so important, and is where having a carefully constructed test-and-control methodology can be revealing.
When the subject of such tests needs to be explicitly processed, such as with specific promotional messages, clarifying awareness in the way I have described is worthwhile. However, questioning can also be useful to support tested elements that one would expect to be processed predominantly at an unconscious level. Here, the challenge is to identify the correlated factors and compare them in test-and-control conditions. For example, people’s assessments about the size of a range may be inversely proportional to the actual range if an aspect of the fixture enables them to appreciate more of the products available when there is more space between them. Asking them how many products they thought were available shortly after they’ve left that part of the store can confirm that one display is working better than another in this regard.
It is possible to identify confirmatory questions for many of the factors that influence the unconscious mind:
Establishing that someone has bought the same product that they did previously is a clue that risk aversion may have driven choice. For purchases that aren’t routine, if people only have a point of justification for the product they have bought, as opposed to a comparative measure with an understanding of the data for the competitor they rejected, the chances are that they have played it safe.
Asking where else they have seen a product or who else they know who owns it can help identify where social proof has played a part.
Understanding the amount of attention paid to competing products can indicate how unconsciously “fluent” the purchase is. When I watch people walk straight up to a product and take it without following the action of their hand and arm, I can be fairly sure that the customer is sufficiently comfortable with their choice to leave the physical act of taking it to their unconscious mind. When I ask people who have shopped in this way what competing products were available, they often can’t name any at all.
Identifying how someone references a purchase in relation to other products can indicate where frames and extremeness aversion have been influential.
Asking someone what other (unrelated) products the shop sells can highlight (through the order in which they’re accessed and the nature of those products) how the customer perceives the purchase more broadly, and may indicate whether other products have contributed to the unconscious perception of the one selected. A retail client asked me to explore the appeal of a competitor which was always mentioned in glowing terms by its customers. Among many factors I identified, it was apparent that people referenced the premium brands of the competitor and, despite the fact that those same brands were available at both stores, the unknown brands that my client pushed to the fore because they were relatively inexpensive. As a result of this frame of reference, all the products in my client’s store seemed less attractive.
Some of the most powerful insights occur where questioning reveals a contradiction between behavior, attitudes, and experience. For example, one friend of mine is almost evangelical about the Apple brand and, following his purchase of an iPod a couple of years ago, has rapidly bought desktop and laptop computers, an iPod shuffle, and an iPhone. He is swift to tell me that, whatever gadget I may be considering, I should get an Apple version if it exists. Recently, I watched as a colleague of his asked him for advice on which laptop to buy. Sure enough, he recommended an Apple product, and regaled the person with an explanation of how they were immune to viruses and how stable the operating system was. Shortly afterwards, I was with the same friend when, to his evident irritation, his iPhone locked up. It transpired that this had happened several times over recent weeks. This unwillingness to attribute the problems he was having with one product to the brand as a whole are a powerful indication of his relationship with the brand and reflect the bias that we all tend to exhibit (except when we’re forced to be artificially rational by consumer research).
When observing consumers in retail environments, I often find that their behavior isn’t congruent with what they tell me. A customer may justify their choice as being “the best product” for them, but if they haven’t done research prior to visiting the store and haven’t spent an adequate amount of time looking at the alternatives available, it is an indication that they have a deeper relationship with the brand they’ve chosen than its competitors. The right probing usually reveals what form this relationship takes and how it has influenced their behavior.
The accuracy of a customer’s critical assessment of their experience, in comparison with what has been observed, is also usually enlightening. Since for the most part our minds work to select the evidence that reinforces our initial perceptions, what customers allow themselves to notice reveals a great deal about what they feel about the product, brand, or retailer concerned.
There is still a place for asking consumers questions, but that place isn’t anywhere near the start of the process of understanding consumers and it should never involve taking what people say at face value. The fundamental nature of the interview must be totally shifted. When the unconscious mind is involved in consumer behavior – and it always is – it is futile to believe that respondents can accurately supply the information required to guide commercial decisions. Instead, the right questions can help substantiate or expand on what has been observed. It is essential to start with the principle that consumers can’t tell you what they think, not that they will.
In my experience, people answering market researchers’ questions aren’t usually actively attempting to deceive anyone (including themselves). When a respondent says “I didn’t realize I did that, but I do,” it’s a good indication that you’ve reached a point of insight into the workings of their unconscious mind.
Respondents are, unfortunately, far too willing to answer questions. One can only wonder about the spectacular event imagined by the respondents who said they were likely to go to visit the Millennium Dome. What is clear is that they were happy to imagine and speculate when the only logical response would have been to say: “How the heck should I know if I’ll go to this event that hasn’t been conceived yet?”
In cases where the only route to reassurance about a future consumer-directed initiative is through an artificial research process (as opposed to a live trial), by taking care to replicate the likely frame of mind and contextual influences, and by being mindful of the potential biases and influences inherent in the research process, it is possible to reduce the risk of getting a misleading reading. In this sense, a convenient approach can be made more reliable than most current approaches manage to be.
Careful consideration needs to be given to the sequencing of questions in research. It is foolhardy to discount the unconscious sensitization and associations formed in a respondent’s mind by one question or comment prior to another – be it what is asked of them or what they hear themselves say in reply. There is much to be said for the one-question survey.
If the subject of research is to do with the present – current attitudes or feelings, for example – the process must start with questions that focus on behavior rather than post-rationalized thoughts and feelings. It is imperative that a respondent isn’t aware of the subject of the research, which makes it much more difficult for them to filter and frame their responses. People frequently think about why they do things, justifying that to themselves and others, but far less about what they have done. This not only provides a basis against which to judge the congruence of subsequently expressed attitudes and feelings, it also helps move the frame of mind from one conditioned by the research process to one in line with the consumer experience of interest.
Following this, I advocate asking questions to explore what elements of unconscious influence might have motivated the behavior. Again, this gives priority to the role of the unconscious over the conscious constructions that may have been made after the event to support consumers’ view of themselves as autonomous, consciously driven, independent beings. Subsequent questioning can invite the respondent’s post-rationalizations and conscious analysis, because by this stage there is a base against which to judge its congruence.
Had the Dome research discovered that respondents’ previous visits to theme-park-style attractions had been infrequent and spontaneous decisions, the researchers might have been less tempted to suppose that their own attraction would draw so many people. If they had doubted the veracity of the research process, they might have asked the same people how many were likely to visit an existing attraction and, when the real numbers of visitors were available, concluded that theme-park visitors weren’t reliable predictors of their own behavior. Nevertheless, with the added complexities of its not being an established attraction and the likely consequence of priming respondents to contemplate an event laden with contextual excitement (essentially the researchers were asking: “Are you going to celebrate the year 2000 by going to the Millennium Dome or can you think of something better?”), the results would still have been misleading.
In many ways, when it comes to asking people questions to inform decisions, it is a case of recognizing the involvement of the unconscious mind and digging around for clues about how it has been influenced. Equally, an acute understanding of the way in which this part of our mind, to which we have no direct access, drives our behavior generally can help identify the likely challenges that any new initiative will have to overcome. Against that background, there is another aspect of human behavior that is far more influential than most of us are willing to concede. To be successful any organization must connect with it but, paradoxically, never ask for its perspective. We must understand the crowd.