When conducting most forms of research, marketers rely on self-report data to predict behavior. In other words, when conducting surveys, focus groups, and one-on-one interviews, the only source of data about the customer is the customer.
We must only look in the mirror to see why this is a potential issue. We are not robots, so when asked to evaluate something, we’re unable to do so with cold objectivity. Instead, we rely on various past experiences, our current emotional state, and context-based biases that help inform the way we think.
Some of these biases are conscious, but we don’t like to talk about them – these are things we won’t say. Some are unconscious, and we don’t know they’re altering our judgement – these are things we can’t say. We’ll consider why these biases exist, real-world examples, and evidence of how they’ve been shown to skew research.
First, we’ll examine the Won’t Say Problem.
In traditional research, there are times when participants might be reluctant to provide their opinion. This is usually based on an almost universal human characteristic called social desirability bias, which influences our likelihood to:
Social Desirability Bias can be defined as “the tendency of [research subjects] to give socially desirable responses instead of choosing responses that are reflective of their true feelings.”
Evidence shows that this bias becomes especially prevalent when individuals are asked to report on socially sensitive topics such as:
Let’s look at some examples.
I was recently conducting a focus group in which one of my participants told me she did not feel comfortable visiting one of our client’s locations on the wrong side of town.
Instead of saying, “I don’t feel comfortable” visiting that location, she said “I don’t feel comfortable bringing my child” to that location. I replied to her statement with a question for the group.
“If you had a kid, how many of you would not feel comfortable bringing them to that location?” Almost everyone’s hand went up.
At that point, it became clear that most participants weren’t comfortable there, but were unwilling to admit to any judgment of cowardice that might be associated with that feeling of discomfort. It was only in the context of children that these individuals were willing to admit discomfort because they could be perceived as protectors rather than cowards.
I was lucky that woman had a child, because otherwise I would’ve been completely unaware of the group’s impression of that location, therefore hindering my ability to provide useful feedback to our client. How often does a sentiment like this go unnoticed in focus groups?
In Dan Ariely’s book Predictably Irrational (which I would highly recommend to anyone interested in neuromarketing), he explores the concept of dishonesty.
He stipulates that, human decisions are typically made based on a cost – benefit analysis. This is similar to weighing the pro’s and con’s, but it is faster, more automatic, and less accurate (System 1 thinking – but we’ll get to that later).
The way we perceive costs and benefits (of cheating or anything else) is skewed based on personal bias and social context.
Because of our societal standards, we are inclined to help our neighbors. These standards are linked to our neural reward structures (like the nucleus acumens), which is why most of us will stop at a red light even when there aren’t any cars around, or return a lost wallet without collecting a reward.
Our cost – benefit analysis considers not only what others think of us, but also our opinions of ourselves as honest upstanding citizens. Based on the following research, Ariely believes that we will usually make slightly dishonest decisions because they don’t interfere with our positive self-image, or moral compass.
Harvard business students were given 15 minutes to take a 50-question multiple choice trivia test. When they were finished answering questions on their worksheets, they were asked to transfer their answers to a bubble sheet. For each correct answer, they were given a small amount of money.
Without the option to cheat, students got an average of 65%. With this control condition recorded, researchers created 3 additional conditions with the intent of increasing students’ motivation to cheat on their tests.
One might think that as the ease of cheating increased and the likelihood of getting caught decreased, the students would be more deceitful.
In fact, the students did cheat a little bit when given the answers in the Self-Check condition, but the amount they cheated did not increase significantly when they were also allowed to shred their worksheets or take money from the jar.
This was strong evidence for Ariely’s hypothesis that our willingness to cheat was influenced by our self-image more than our likelihood of being caught. In this way, small transgressions will fly under our ‘guilt radar.’
While most of us wouldn’t rob anyone, we might be willing to take home some supplies from work, or become complacent when we take a survey…
Over $9 billion dollars are spent on surveys each year, but new research shows that much of the data collected this way may not be trustworthy due to a lack of interest from participants.
The following experiment found self-report data collected via online surveys to be remarkably unreliable. The results of this study were stumbled upon accidentally when researchers were trying to determine the effect of public service announcements (PSAs) on tobacco quit intentions.
They attempted to collect data from two groups:
After watching the PSAs, participants were asked to rank their effectiveness and were compensated with a $20 gift card.
They used facial expression analysis to group participants into three categories:
The community survey flier was shared far beyond its intended range leading to 58% deceitful and disinterested participants, who were presumably only looking for compensation. Vetted participants had more valid data, but 13% were still found to be deceitful or disinterested.
These invalid results led to altered rankings of over half of the PSAs in the community group and 2/12 in the vetted group, which is evidence that much of the survey data collected online may be unreliable (Hammond, 2022).
However, the article also stipulates that much of the error from changes in participant attention can be mitigated by measuring other types of data, such as facial coding (cough, cough, neuromarketing).
While most of our Won’t Say behavior is ingrained in our nature as humans, it is still conscious. Therefore, we may not always pay attention to the effects of social desirability bias, but when we “cushion” the truth, we are aware that we’re bending our response based on social context.
This is not the case with the Can’t Say Problem, in which bias is created at the subconscious level.
Most people think of themselves as rational beings who make decisions based on facts and logic. In fact, we are rationalizing beings who do our best to explain our actions based on the information available to us.
To explain why this occurs, let’s discuss an example, then consider the issue from both psychological and neuroscientific perspectives.
Have you ever heard that multitasking is a myth? As humans, we’re unable to perform two tasks at once without reducing our quality of performance. This limitation is based on our ability to consciously focus on only one thing at a time. The rare few who appear to be excellent multitaskers are actually just good at switching their attention between tasks quickly.
Evidence of this phenomenon can be seen in our inner dialogue. We’re capable of speaking to ourselves internally when we’re reading or thinking. We can imagine words in our minds as if we’re hearing them, however, we cannot make this inner dialogue speak multiple sentences at once.
To examine why this is the case, we’ll first look at the psychological perspective.
The renowned behavioral psychologist Daniel Kahneman explores the topic of two systems of human thought in his book Thinking Fast and Slow (another key text for aspiring neuromarketers).
Kahneman realized that although people can think rationally, behavior cannot be explained only by rational thought because we are prone to frequent and obvious logic errors. Instead, we have two systems of thought characterized by the following:
|
System 1 |
System 2 |
|
|
Let’s examine these traits one at a time.
System 1 is always active. It’s constantly processing sensory information, maintaining homeostasis, and looking for unforeseen threats. It runs autonomously, below our conscious perception. System 2, however, is just the opposite. It turns off when we sleep and requires conscious concentration to perform tasks, often feeling fatigue after “mental work.”
While System 1 processes vast amounts of information at a rapid pace, System 2 requires time to perform even simple tasks. For example, the mental calculations required to maintain coordination while we run are extremely complex. We make small adjustments to hundreds of muscles, but it happens immediately and without thought. Alternatively, if I ask you to answer this relatively simple calculation, 35 x 18 = ?, it will take you much more time.
System 1 is influenced by the chemical balance in our brains and makes decisions based on emotion and instinct. System 2 is instead able to make choices based on mathematical structures and complex objectives.
System 1 is capable of some type of “multitasking” as it can perform many calculations at once. It’s constantly interpreting information from all senses while controlling the body’s regulatory functions and movement. System 1 is responsible for our instinctual reactions and reflexes, such as our fight/flight/freeze response when faced with danger.
System 2, on the other hand, is capable of processing only one thing at a time. Therefore, we cannot perform multiple complex tasks simultaneously. It is also ultimately in charge of the behavior we express, but because it requires slow, conscious effort to function, it cannot review all the calculations made by System 1.
The answer to this question, like many other questions in biology, is that structure dictates function. Neuroscientific evidence provides great insight into how these systems govern behavior.
Our brains have the monumental task of controlling all our bodily systems, interpreting sensory information, and still creating time for abstract reasoning. How is this possible?
Neural structures have evolved to reduce complex computations into simple on/off signals, like computers. As raw data from the body is passed through structures in the brain, it is augmented and simplified. This has positive and negative effects for us…
We can identify threats and react quickly. For example, imagine there’s someone walking towards you. In order to survive, we need to know if this person is friend or foe.
Instead of pondering over the 130 million optical sensors in our eyes, the optic nerve performs simple calculations sending only 1.2 million nerve fibers towards the brain.
These signals continue to be simplified as they pass through additional synapses (the junctions between neurons) and soon, our brains can determine the identity of the approaching person based on the reaction of a single cell. In fact, for each person we know, there is a single cell in our brains that will recognize them.
This cell is commonly referred to as the “Jennifer Aniston Cell” because one of the experiments that popularized it showed that whenever someone saw an image of Jennifer Aniston, no matter the angle the photo was taken from, that neuron would fire.
The same neuron would not respond to any other faces leading researchers to the belief that we have a cell devoted to each person we recognize.
So, in less than a single second we’ve identified the approaching person as Jennifer Aniston, who we find very threatening. Our brains can then respond appropriately with the fight/flight response.
Unfortunately, these shortcuts can sometimes lead us astray…
The Halo Effect is one of psychological biases that will be discussed in the following section but provides a clear and simple example of how oversimplification can (and does) lead to errors in judgement.
The term “Halo Effect” describes our tendency to judge the unknown characteristics of someone/something based on the known characteristics, even if the two are unrelated.
For example, when we see an attractive individual, we are likely to perceive them as intelligent and kind as well. We make this generalization even though there is no correlation between beauty and intelligence.
This is particularly important to recognize in marketing as many people are willing to answer the following questions interchangeably:
If they believe that the answer to one of the questions is yes, odds are they’ll simplify and answer the second question the same way and vice versa. Therefore, if you can get people to like your brand, they’ll believe you have a high-quality output even if they haven’t gotten to experience it yet.
In the following section, we’ll describe additional negative effects of simplification and discuss how these errors in logic have also been converted into successful marketing strategies. Before that, let’s consider why it’s so difficult to recognize when we make these errors.
The information provided to our conscious brains is effectively filtered before we have access to it. This is a result of the brains internal structure and effects everyone.
Although the brain can be divided in many ways based on function and connectivity, in the context of neuromarketing, it is generally agreed that there are three major areas:
Because most of our sensory information enters through the brainstem, it is filtered by our Hind, Mid, and Limbic brains (System 1) as it makes its way to the cortex (System 2). In effect, it passes through various filters before it reaches the conscious and logical parts of our brain.
This ties back to the psychological theory of System 1 as a first responder. When System 2 is provided access to the information, it has already been biased, so it can be difficult to recognize that bias.
The best way to free yourself from judgment fallacies is to understand when you might experience them. The following section will discuss our psychological biases and the marketing techniques that utilize them.
While this is not comprehensive list of the psychological biases created by our neural structures, it will provide insight into many of the ways our decisions may be altered by failures in reasoning. Keep in mind that many marketing strategies rely on these biases in some way.
There are various elements surrounding our prediction and decision-making capabilities that indirectly alter the choices we make. Some of these will be explored below.
Cognitive ease is a state of mind characterized by a positive mood, pleasant facial expression, and System 1 thinking. When in a state of cognitive ease, people tend to believe what they hear, trust their intuition, and feel as if they are in a familiar and comfortable situation.
Cognitive strain on the other hand, is characterized by negative mood, unpleasant facial expression, and System 2 thinking. When in a state of cognitive strain, people tend to be more vigilant and suspicious, invest greater mental effort in their tasks, make fewer errors, are less creative, and feel their situation is unfamiliar.
Cognitive ease and strain occur naturally, but they can also be induced. For example...
You may have noticed that your brow furrowed, and you frowned while looking at the second sentence. As it turns out, all these elements are connected in a feedback loop. Mental effort to process unfamiliar things make us frown, and frowning makes us exert more mental effort. The same is true for cognitive ease; familiar things give us positive facial expressions, which help us trust our instincts and reduce our cognitive effort.
One experiment that provided evidence for this feedback loop by asking students to take a math test in one of the three following conditions:
Students who held the pencil between their bottom lip and chin were forced to use their facial muscles to frown and had significantly fewer errors in their test than the other groups. Similarly, students who held the pencil between their lips were forced to use their facial muscles to smile and had significantly greater errors in their tests than the other groups.
Emotions are also part of this feedback loop. Positive mood makes us smile and reduce mental effort, but the reverse is also true. Smiling actually improves our mood and reduces cognitive strain. To take it one step farther, seeing other people smile activates the mirror neuron system (which is believed to be the cognitive basis for empathy) and makes us more likely to smile in return.
In this way, cognitive ease and mood can be altered as people watch advertisements to induce a feeling of familiarity and likability. These are the first steps to getting someone to trust your brand.
Most of us have looked at a menu in a busy restaurant and considered choosing the healthy option, but somehow, we ended up with a dark beer, fried chicken, and some ice cream for dessert. Why was it so tough to order the salad?
New research shows that your cravings can be strengthened by the volume of ambient noise, and that your self-control can be weakened by distractions that deplete your cognitive resources. While System 2 is busy allocating attention to the environment, System 1 convinces you to indulge your cravings.
PHD Danni Peng-Li designed an electroencephalography (EEG) and galvanic skin response (GSR) experiment to determine how our mental workload and emotional arousal altered our decisions respectively.
During this regulation of craving (ROC) task, participants were asked to look at pictures of food and report their level of craving while the level of ambient noise and their decision perspective were manipulated.
As expected, when long-term effects (System 2 thinking) were considered, subjects reported significantly lower levels of craving than when they considered immediate effects (System 1 thinking). This was seen regardless of sound level.
However, when ambient volume increased, so did the level of emotional arousal in participants. Results showed that the loud condition led to increased emotional arousal regardless of the decision perspective.
Increased arousal was not enough to alter the strength of cravings on its own but did have a significant effect on cognitive workload. Results showed that cognitive workload was unchanged by the decision perspective when ambient noise was soft, but when the noise was loud (and emotional arousal was high), considering future decisions was significantly more difficult.
Therefore, evidence suggested that noise in the restaurant can be distracting (via emotional arousal) and acts to deplete cognitive resources, causing people to make irrational (System 1) food choices, such as selecting cake over a salad.
This has huge implications for restaurants, as those with many unhealthy options may benefit from boosting the volume. Even restaurants that offer healthy food for high prices should consider doing the same, because an increase in cognitive workload (leading to a reduction in rational, System 2 thinking), is likely to also reduce the hesitancy caused by high costs.
If you happen to work in sales, you’ve probably heard of this classic tactic. Anchoring is based on the idea that people aren’t good at judging value and have no innate sense for comparison. Because money is an abstraction from tangible goods, our brains see everything as relative. Therefore, instead of having some internal numerical baseline, we tend to gravitate towards the first numbers we’re exposed to.
For this reason, when you sell something, it can be beneficial to anchor people’s expectations on extremely high numbers before presenting them with an actual offer. Many companies use a tiered pricing model to achieve a similar effect. Anchor expectations for high numbers with your most expensive option, then offer something more manageable for your other options. It is best to have three options and expect that the buyer will choose the middle option.
The following are two excellent research examples of this anchoring effect from Ariely’s Predictably Irrational.
Imagine you see an online ad for a magazine you’re interested in. You read the following pricing options…
At this point, you’re probably considering choosing the internet only option. Then you see the third option:
You think to yourself, wow if I get Option 3, I’ll basically be getting Option 1 for free! This, of course, is an illusion. There’s no reason you would need both the print and internet versions of the magazine, but your brain has already assigned a value to each of the first options, which makes the third option (at no extra cost) look like a steal.
This doesn’t only occur with impulse monetary transactions. In fact, research shows that those expected to be society’s most unbiased individuals, judges, are prone to a similar mistake.
For this experiment, people wrote large or small numbers on whiteboards that judges saw before making sentencing decisions. When exposed to the large numbers before passing a sentence, judges will pass significantly longer sentencing times, even when they know these numbers are irrelevant, and even if they were unaware of their exposure in the first place.
Anchoring is a powerful tool that can be good for business and detrimental for society when used the wrong way.
The Halo Effect is a term coined by Edward Thorndike which describes the human tendency to evaluate a particular trait as positive or negative, and apply that conclusion to all other traits. It is commonly known as the “what is beautiful is also good” principle.
Examples of this can be seen in studies of student appearance and academic performance. One article examined over 4,500 students and found that those who were rated as above average in attractiveness got significantly lower grades in online classes compared to traditional classrooms because their instructors could not directly evaluate their appearance (Hernández-Julián, 2017).
In this way, instructors were applying the Halo Effect by perceiving attractive students as more intelligent, even though the two characteristics are uncorrelated. This occurred because instead of answering the complex question, “Did the student perform well?” System 1 replaced it with a much easier question, “Did the student look good.”
The Halo Effect may be the psychological basis for the importance of branding. Consider the following example…
You’re in the cleaning isle of the grocery store and you can see 10+ types of detergent with only small variations in price between brands. You want the one that is most effective, but you haven’t used any of these types before. Instead, one of the brands looks familiar and you remember seeing their ad.
Despite the brand’s marketing budget being unrelated to the quality of their product, the familiarity and likability of the brand was sufficient to replace the question of quality.
While it may sound outlandish that human behavior is based mostly on our unconscious minds, errors in our ability to reason show that this is the case. Examples of this include the negative information bias and our poor ability to estimate the likelihood of events.
Negative information bias is our tendency to seek out negative information, threats and losses, and prioritize it over positive information. This is why it is so important to market the problem. Consider the following:
This is because once they had the mug the idea of losing it became much more painful. Research shows that losses are valued at about 3x higher than gains, meaning we need to gain $3 to make up for the loss of $1.
The difference between the decision made by group C and the others is referred to as framing. For group C, the question of which item to select was framed as a gain – you had nothing now you get something. For groups A and B, the question of trading in an item was framed as a loss – you had something, and you must give it up to get something else.
In both scenarios the outcome is the same; you get a coffee mug or a sandwich based on your personal choice, but people chose differently based on how the question was presented. This is why free trials are such a popular and effective marketing technique. Rather than having to lose money to get access to a product or service, you have to lose the product or service you already have unless you fork over the cash.
Another example of our unconscious logic errors is the availability heuristic: our tendency to judge the probability of events by the ease with which relevant instances come to mind. This is why people are afraid of shark attacks even though you’re more likely to get killed by a champagne cork. Shark attacks are more prevalent in the media, so they come to mind more easily, and people judge them as more likely.
There are many other logic errors to be discovered in our learning center, and this is not an exhaustive list, but now it’s time to discuss some solutions to the problem.
We’ve considered many ways in which either the structure of our brains or our social desires have altered the way we make decisions. You may at this point be wondering how these innate biases can be avoided in personal and professional life.
When it comes to your own biases, the best way to combat them is by developing habits of self-reflection. Once you’ve identified situations in which one of these biases is likely to occur, you can intentionally activate System 2 to analyze your decisions using logic rather than instinct. Of course, it will never be possible to eliminate personal bias completely, but it is a skill you can build over time.
When it comes to detecting biases in others, there are two possible methods. The first is through direct measurement of behavior. While people can come up with various rationalizations for their decisions, the true test of marketing effectiveness is not what someone says about your content, it’s what they do after they’re exposed to it. You can eliminate the bias seen in traditional market research by measuring behavior directly. This can sometimes be costly, time consuming, and unpractical in marketing research contexts unless research participants are asked to track their own behavior. Of course, when that happens, you open the experiment to additional types of bias.
The second method to avoid bias in others is to directly measure their subconscious reactions. Because nearly all the biases discussed in this article are created by the part of the brain which controls our body’s automatic functions, measuring those automatic functions can provide detailed information on how someone is responding to marketing messages. With the right technology, you can monitor people’s attention, emotional state, mental workload, or even whether people like or remember things.
In a world where marketing research is full of conscious and unconscious bias, the ability to directly measure how someone reacts to marketing content is a powerful tool. It allows us to make predictions about the effectiveness of ads, websites, or other media before spending the money to produce and distribute that content. It allows us to determine how specific audiences will react differently to the same content based on their personal biases. It allows us to make informed decisions based on science rather than wasting money on the inferences of marketing experts. If you’re interested in an unbiased understanding how your content will be perceived, it might be time to Get Started.