Dan Ariely - Market Research - Predicting The Irrational

NGMR Marketing Guru Interview – Dan Ariely and Tom H. C. Anderson discuss Predictably Irrational and market research

Today I have the pleasure of speaking with Marketing (Behavioral Economist) Guru and author of Predictably Irrational, Dan Ariely. Importantly, his book is also currently on the Next Gen Market Research Top-10 list (according to our recent survey among 850 Next Gen Market Researchers). Dan is also professor of Behavioral Economics at Duke University Fuqua School of Business and visiting professor MIT The Media Laboratory.

Tom: Welcome Dan. I really enjoyed the recent interview you did with yourself. It’s nice to see a business guru with a good sense of humor.
What’s the difference between “behavioral economics” and “consumer insights” or “market research” etc.?
Dan: In reality, there are a lot of similarities between different versions of social science. They all try to understand and predict human behavior. The particular nuance that is specific to behavioral economics is that it examines behavior from a perspective of rational economics and looks at that as a starting theory (often wrong). There are also multiple versions of behavioral economics…

NGMR Marketing Guru Interview – Dan Ariely and Tom H. C. Anderson discuss Predictably Irrational and market research

Tom: Welcome Dan. I really enjoyed the recent interview you did with yourself. It’s nice to see a business guru with a good sense of humor.
What’s the difference between “behavioral economics” and “consumer insights” or “market research” etc.?
Dan: In reality, there are a lot of similarities between different versions of social science. They all try to understand and predict human behavior. The particular nuance that is specific to behavioral economics is that it examines behavior from a perspective of rational economics and looks at that as a starting theory (often wrong). There are also multiple versions of behavioral economics but my particular version is concerned not only with changing our understanding and reliance on economics but also thinking about what kind of changes we would make in the world if we understood that people behave in a consistent, predictable way that is not captured by standard economic theory.
For example, we can think about how we want to change our financial system to take into account that people cannot deal well with conflict of interest. Or we can think about how to think differently about health care if we realize that people who are sick cannot weigh the cost & benefits of different treatment options, etc.
Tom: Tell me a little about Social markets vs. financial markets. You’ve conducted several experiments, one in which you paid people either $0.0 (nothing), $0.50, or $5 to do a very simple task and you found that those who got paid $0.50 did the least work. What does this mean for us as marketers?
Dan: The basic idea is that we have two types of rules for our behavior. One type concerns market norms which involves how much you pay for things, how much people charge you and so on. The other type is social exchanges that have to do with fairness and warm fuzzy feelings. Both of these relationships are perfectly reasonable and have advantages and disadvantages. Marketers need to understand the particular advantages and disadvantages that come with both of these relationships and perhaps more importantly the failure that can occur when the relationship is in the middle.
Tom: Thinking about this specifically in terms of market research, we often incentivize people for participating in survey research or focus groups. We might pay from nothing to $15 to take a survey or as much as $200 to participate in a focus group. We know incentives can influence results, but perhaps this is having more of an impact than we think?
Dan: When people are paid money, they’re going to feel some obligation to the researcher who gave them this gift. We should particularly worry about his in the domain of focus groups – when the amounts are larger, and there is also a direct relationship to the person paying them, I suspect that this can further get people to do what they think is pleasing for the research entity.
Tom: Perhaps a good reason to do more web ethnography/screen scraping and test mining in the future.
But, how about in a larger perspective. How do you think marketers can leverage social media with the idea of social markets? Is this something that NGO’s/Non-Profits could leverage? Are there any companies out there that you have seen leveraging this idea better than others?
Dan: Social media is a general term for lots of different things. There’s no question that companies can use social media to better understand what their customers are complaining about and what is on their mind.
Take the example of Comcast tracking customer complaints on Twitter. That’s just one type of social media use where companies are listening and trying to be proactive. There is also a way for companies to create services and products that can be promoted on social networks.
Take for example, a software program like TurboTax that helps people complete their taxes in a comfortable, quick way versus going to an accountant. Companies like that can create more opportunities by telling their existing customers “If you want to tell your friends about how great we are, here’s a way to do that.”
Tom: Switching topics a bit, I love your example video of visual illusions and the cube which explains how even our experiences can’t teach us the correct way. Can you tell us a bit about how this is relevant to consumer behavior?
Dan: The idea of illusions is that our senses feed us information and that is not a direct reflection of reality. Information is provided to us courtesy of our brain, which has a set of rules for situations so that what we see is very different from reality in specific, repeatable, predictable ways.
The same thing applies to consumer behavior. For example, when something costs more and people expect it to be better, they actually end up seeing and experiencing it as being better. It’s like an illusion in the way that we process the information is not a function of what is out there, it’s a function of what is happening in our brain. Most of our understanding of the world comes from our brain not from our senses. We think we see with our eyes but much of what we see is happening with our brain.
For example, when you lie on your back and look at the sky, you think you’re seeing blue but in reality only a small part of your eye can detect blue. If you extend your arm out in front of your face and hold your hand in a fist, that’s the only portion of your eye that can detect blue. The rest of your eye isn’t supposed to see blue. We don’t see blue because of our eyes, we see it in spite of our eyes. It’s our brain that is doing all the work to help us detect blue. The same thing happens when we process other information: Price, quality, etc. In all those cases, it is our brain that drives our expectations and determines much of our final experience.
Tom: If people make irrational decisions continuously, how can we expect to use traditional survey research to ask them their opinions? Are there effective ways to use traditional market research techniques?
Dan: The first point is to think about which areas of life people have good intuitions and which areas they don’t. You can expect, for example, that in areas where people have lots of experience, deliberate a lot, or are experts that they might be able to articulate some of their opinions in a more coherent way.
For example, someone who has used many iPhone apps might be able to understand what attracts them to some applications and not others or why they are willing to pay for some applications and not others. In cases where people don’t have experience or good intuitions, it’s not useful to ask people their opinion.
The second point is that we need to think of what we are asking people. If you ask people what they did yesterday, what they’ve eaten today, how much they’ve walked today, how much they’ve paid for a product . . . things that are easily measurable, we will get bias results for sure but to a lower degree. But in this case we are asking what is better than why.
Finally, we need to remember that surveys are a cheap substitute for a real experiment or field test. They are a compromise We are sacrificing accuracy and ability to predict because we are trying to do something quick that is not time or money consuming. If we have the money we should do the whole study.
Tom: So, what are some ineffective ways you have seen?
Dan: One practice I’m most disheartened about is the heavy reliance of focus groups-the idea that twelve people who know nothing about a business, answer random questions and then substantial, important business decisions are made based on this It sounds crazy but organizations do it all the time.
I think there is something intoxicating and fulfilling about focus groups. When you do a study or survey you are able to extract hard numbers .6 or 35%, for example. But focus groups give sentences or testimonials “I like using this product because…” Since we are all storytellers in basic ways, marketers can include such sentences on their PowerPoint slides and present it upper management. The fact that they have this sentences also gives marketers in the organization confidence that they have a sense of what is going on, which is not necessarily true.
Tom: What do you think about structured choice research like Conjoint or Discrete Choice techniques where different products, levels, attributes and prices are tested against each other. Often later a what-if market simulator is created. This type of research is very popular, especially in B-B research.
When if ever does this type of research make sense if we are irrational?
Dan: One of the big lessons from behavioral economics is that the decision environment makes a big difference in people’s choices If the natural conditions in which people make decisions about buying a car is by going to the Web, comparing three types of models first, making a decision, and then doing a price comparison later then a good market research technique should follow the same choice procedure. If, for example, when people are choosing which soft drink they want, they also think about what pretzels or chips to get with it, it would be good to represent this other part of the decision as part of the decision environment.
If the decision environment plays a big role in what people end up choosing, it’s important to model or represent the decision environment. To the extent that conjoint analysis correctly represents decision environments that people use when making particular decisions (when people consider different computers they look at the whole profile: resolution, memory, hard drive space) then it’s a good method. But to the extent that it uses different decision-making processes then it’s not accurate and can actually be misleading.
Tom:Have you seen any company advertising lately that clearly has taken our weaknesses in behaving predictably irrational into consideration? What are some good ones why?
Dan: In Predictably Irrational, I presented this ad from the Economist. The magazine added a decoy and the experiments showed that adding an alternative that people don’t want actually changes what people end up deciding. I recently saw an ad for The New Yorker that basically used the same trick: for the price of X you could get The New Yorker in print or also for the price of X you could get The New Yorker in print AND online.
TEST ADVERTISEMENTS:
ACTUAL RESULTS:
Tom: What are some bad ones and why?
Dan: Another example of people who take irrationality into consideration are people who are basically selling placebos. Placebo is one of the ideal examples of how the power of belief and persuasion can actually create a real effect. There are many companies with questionable medications with questionable efficacy but some seem to understand the psychology of individuals and how the power of belief can get people to use them their belief that the medication will work actually helps medicine become effective.
In terms of ineffectiveness, I think the industries that have the least understanding of behavioral economics and what people can and can’t do themselves as well as the way that people reason they reason are, perhaps not surprisingly, the financial and insurance industries. The financial industry has basically created a whole ecosystem around the idea that you can ask people how much they will need for a comfortable retirement and to identify their risk tolerance and that people can answer these questions in accurate, meaningful ways. Similarly speaking, the insurance industry assumes that describing insurance programs in terms of deductibles, limit and so on makes the most sense. In reality it’s incredibly hard to think about what these things mean as a consumer. The insurance industry assumes that people perfectly understand their options and can reason about them. And often, though not always, they advertise by explaining those attributes and assuming that people understand what they mean and can make a decision based on that information.
Tom: How does consumer segmentation fit into being predictably irrational, if at all? Is it possible/wise to segment customers into different groups and treat them differently?
Dan: Segmentation is alluring, When we look at the world and see differences among people we tend to ask questions to help us examine and focus on the differences between people, gender, socioeconomic status and so on. The reality is that we are much more similar to each other than different from each other. Though we are naturally not attuned to seeing these similarities. We aren’t as interested in a system that detects the similarities because it’s uninteresting and uninformative. So our system focuses on differences. But even though we have more similarities than differences, we aren’t as designed to see them.
That’s one reason it’s so tempting to look at the differences. But I suspect we would make much more progress if we spent more time examining how we operate in similar ways than in different ways. That’s not to say that we don’t have our differences, it’s just that we have so many similarities that we should not be tempted to just focus too much on differences and, in the process, neglect similarities.
Tom: If we do so many things irrationally, what if anything are we rational about? Do we vote rationally for instance?
Dan: Voting has been one of the big puzzles for economists because, strictly speaking, nobody should go to vote. A single vote will not change an election one way or another. Plus it’s a costly process. Voting is a demonstration of an irrationality that economists have been struggling with for a while. I think that voting illustrates another important point about irrationality-people think irrationality is always a bad thing but in reality there are lots of beneficial things about. It’s good for people to stay irrational in some respects. Voting is one of those things. It’s nice that people want to vote and that they care about it. Irrationality is not always bad.
Tom: Thanks Dan. Great speaking with you!

@TomHCAnderson

[I’ll leave you with a final video that illustrates some illusions that keep us from thinking rationally in the market place. Dan has several others on his blog and on Youtube]