I had a chance to interview Barry Schwartz who is a guru on strategies of discovery. This is the first of a series of Podcasts on Search and Discovery. I enjoyed the conversation. Thanks to Aggregate Knowledge for introducing me to Barry.
Barry Schwartz, award-winning author of The Paradox of Choice: How More is Less, explains how the increasing demand for options actually decreases consumer confidence and satisfaction: people can be overwhelmed with too many choices. In this podcast, Schwartz discusses his research and how it can apply in the online world. He gives practical advice for retailers on how they can delight their customers and be successful in the online world.
Barry Schwartz is the Dorwin Cartwright Professor of Social Theory and Social Action at Swarthmore College. He is the author of several books, including The Battle for Human Nature: The Science, Morality and Modern Life and The Costs of Living: How Market Freedom Erodes the Best Things in Life. His articles have appeared in many of the leading journals in his field including the American Psychologist.
Full Transcript Here:
We’re here with Barry Schwartz, the author of The Paradox of Choice. Barry, welcome to the podcast. Tell us a little bit about yourself and your background, your academic background, and your book, Paradox of Choice.
I got my Ph.D. in psychology from The University of Pennsylvania in 1971 and I have been teaching at Swarthmore College ever since. For 20 years, I have been interested in how well the assumptions that economists make about human nature actually fit with what we know about human nature, The answer to that, by the way, is not very well.
In the last few years I got interested in a specific [assumption], which is the idea that the more choice people have the better off they are. [This idea] is deeply embedded in economic theory, and there’s evidence that’s accumulated in the last few years that that assumption is also false… that you can give people too many choices and it screws them up in various ways. My book, The Paradox of Choice, is a summary of that evidence and an argument about why people can be overwhelmed with too many options.
Talk about The Paradox of Choice in an offline world, and that’s mainly where all your research has been driven on in terms of your examples. Give us an example of some of the things that happen quite frequently.
Well, in an offline world, you’ve got 175 salad dressings in the supermarket, 250 kinds of cereal. You go to Circuit City and you can construct six million different stereo systems out of the components on hand. You look in the newspaper and there are 10,000 mutual funds – stocks – to choose from. There is no less than 30 kinds of dental floss. There’s simply no area of life where people don’t have an extraordinarily large number of options in the offline world.
And, again, the assumption was that adding options will make people better off. If you’re alternating between Corn Flakes and Rice Krispies and I introduce Sugar Frosted Flakes, if you don’t care, you can ignore it, and if you do care I’ve just made your life better. That’s been the operative assumption. So in the offline world, you can’t walk into a store without experiencing this.
What does your research show in terms of the consumer?
Well, there are three different effects that having too much choice produces. I should emphasize, most of this work has been done in labs, not in real retail environments, although some work has been done in real life environments. The first effect is with all these options to choose from, people end up choosing none. They simply pass.
So if you’re a retailer what that means is that you actually sell less. People just don’t know how to choose, they walk out and they throw their hands up in dismay, they walk out empty handed. That’s one problem.
Second effect is that if people overcome this indecision and paralysis and choose, they may choose badly. And here’s what I mean. If you’re trying to buy a stereo system, let’s say, there are certain things about it that are easy to evaluate and other things that are hard to evaluate. Sound quality, fidelity, loudness, those are more difficult than, say, physical appearance. If you have a lot of options to choose from, it’s just not possible to do the difficult evaluation of each of the features that matter. So what you’re likely to do is simplify the task and choose on the basis of what’s easy to evaluate, even if that’s not what’s so important.
So you might choose by brand and price and ignore all those wonderful subtle features that producers of these goods take so much trouble to create. Now if it happens that all you care about is the simple things like brand and price, you’re not harmed by this simplifying but if you actually care about other things, you’ll end up making a less than optimal choice.
So first is paralysis, the second is worse decisions.
The third thing, which is in some ways the most surprising, is if you overcome paralysis and manage to choose, and you manage to choose well, you’ll be less satisfied with what you’ve chosen if you’ve chosen from a large set than if you’ve chosen from a small one.
So even though you do well, you end up disappointed and of course you’re likely to blame that disappointment on the thing you’ve chosen and not on your own psychological processes, and the reason for this is that it’s very unlikely that there’s one stereo or one cereal or one place to go on vacation or one restaurant that is in every respect the best. So you have to make trade-offs. And if you’ve considered a lot of options and you say “No” to most of them, you’re saying “No” to a lot of really attractive features of the other options.
And even if you make the right choice, you end up thinking about all those other wonderful things you’ve passed up, and that makes your choice less satisfying.So in the marketing world, they talk about cognitive dissonance. In a way, more choice creates more of that.
It does, except that if cognitive dissonance worked, upon making your choice to go to the Bahamas for example, you would immediately forget about all the alternatives or convince yourself that the alternatives really sucked and there really was only one option. That’s what we’re supposedly doing all the time to reduce the dissonance that we experience when we have a tough, conflicted decision. But I believe that doesn’t work when there are lots of options to choose from.
Or if it does work, it works very temporarily. I mean, in the long run you are sitting on the beach in the Bahamas thinking about how great it would be in the Rockies.
How have you looked at some of the things that you’ve discovered on the offline side. Can you share some of your thoughts on the online choice issue?
Absolutely. Now I haven’t done any research in online settings. I would love to be able to; still hoping to be able to find the right partner. In theory, I think online is both better and worse than offline.
It’s worse because the number of options available for you to consider is essentially infinite. Everything is there and the next source of alternatives is just one mouse click away.
So you could spend your whole life looking for a stereo if you were of a mind to do that. That’s what’s bad about it. You know, if you’re going retail, eventually you just say I’m tired of getting into my car and driving across town; I’m just going to pick one from this store, I’ve been to five already but it doesn’t take any work to go to five, 10, 20 or 50 websites. So that’s why it’s worse.
The respect in which it’s better is that the online retailer has the tools to very much customize what the individual visitor gets to experience. So even if you’ve got thousands of stereo systems available, you don’t necessarily have to show them all to each customer. If you ask the right questions in advance, you can turn your super department store into a boutique. I get to see five, you get to see five, and they’re not the same five. And so we are choosing from a small set, not a large set, thanks to the clever way in which you’ve organized your site.
So I think that we need retailers to think of themselves much more as agents or curators rather than simply providers if they want their customers not to be paralyzed.So group one sees, say, 20 suggestions on a website. And group two sees five suggestions….
Right – well, our prediction is straightforward. If you show me 20 things, I’m less likely to choose any than if you show me five… For pretty much any category. No one has done that experiment to my knowledge, but that’s what I would expect to happen.
Some of the things that Aggregate Knowledge is doing, (such as) user behavior collected in an aggregate way, — the human collective becomes a filter. Is that something that can solve the paradox?
I think it can; it really depends. From my understanding, for example, of how Netflix works – you get DVD recommendations based on what you’ve chosen in the past; how much you’ve liked what you’ve chosen; and how much other people who have chosen the same things liked these recommendations. So [Netflix] can give you a list that is as long as the equator or they can give you the ten things that, based on your past behavior, you’re likely to find satisfying.
I think that’s a terrific way to solve the choice problem, and if you’re the kind of person who doesn’t want to be told what you’re supposed to like, well then, all 80 million titles are available for you to inspect. Not everybody will think this is the Holy Grail right now, coming up with the best system, best algorithm, for making recommendations to people that are individualized. The knowledge problem has been solved, mostly (thanks to Google) and now people realize that once you make every piece of the world’s knowledge available you’ve created a new problem. Which is– making it useable. And that means managing it, organizing it, editing it, filtering it, and so on.
And there is not yet a Google equivalent, at least in the retail space that I’m aware of.
So are you planning any tests with Aggregate Knowledge to test the paradox at all?
I’m hoping to. I’m hoping – I mean, the ideal form of a test is to go to some highly trafficked website and create alternative visions that, for example, either vary how many options people see or vary the way in which those options are organized. You can organize them hierarchically so that it doesn’t look like such a long list. And for a day, a few hours, depending on how much traffic the site gets, show random people one or another version, and then measure the things that are so easy to measure online: how much time people spend and what the conversion rate is. And you can ask them how satisfied they are with the site, whether they would recommend this site– you could even ask them further downstream how satisfied they are with whatever they bought.
So that would be my ideal. Now most of the things I’d want to measure, I suspect, are already being measured by these various sites.Yeah, the collective, anonymous group model, which basically drives discovery around consumer behavior–identifying patterns around anonymous data–is less profiling and more of the “wisdom of crowds.”
Well, it can be anonymous. I have other interests because I think the choice problem is worse for some people than it is for others. It would be nice, for example, to give people the scale that I’ve developed that distinguishes people who are out for the best, I call them “maximizers,” from people who are just out for good enough. My prediction is that maximizers will be especially plagued by large choice sets. So then people wouldn’t be quite anonymous because you’d give them a little survey before they even started, and then sort them on that basis.
In your book, Paradox of Choice, you talk about “satisficers and maximizers.” Talk about the difference between the two types of people.
A “Maximizer” wants “the best” – the best vacation place, the best restaurant, the best stereo, best job. And in order to find the best, you need to evaluate every possibility, otherwise how do you know it’s the best? And of course the whole point is that with a trillion choices out there, examining every possibility is just not possible. It produces a sort of paralysis of indecision, exhaustion, huge amount of time spent choosing and then finally you pull the trigger and you pick something and there are still dozens of things you haven’t considered, so you imagine that you could have done better if you’d looked a little longer.
A “Satisficer,” by contrast, is only looking for something that is good enough. And your standards could be low or they could be high. You can be very discerning or not. The point is you know what you’re looking for and as soon as you find something that meets your standards, you just choose it and you don’t worry about what else is out there. So you don’t need to do an exhaustive search. That means a large number of alternatives may not be nearly as much of a problem for you.
And what we find is that people differ in this regard. People who score high on this maximizing scale take more time to choose, look around at what other people are choosing, and are, in general, less satisfied with what they choose, and less satisfied with their lives than people who score low on this maximizing scale.
In the online world I think a lot of retailers get caught-up on metrics like time on site…A Maximizer would spend more time on a site – the other category they want to get out of there as fast as possible.
Well, they might want to get out of there as fast as possible; you know, if you’re a Maximizer and you’re looking to buy a $25 toaster and you figure this will take five minutes online and you come down four hours later…
On the one hand, you know, it’s not worth four hours to choose a $25 toaster – on the other hand, damnit, you want the best toaster.It’s a conflict. Do you think you’ll be able, in an online setting, to tell the difference between a Satisficer and a Maximizer?
Well it’s easy enough to tell the difference just by giving them the questionnaire–it’s just a 10-12 item survey and there’s no reason to think they’d be any more serious about it online than they are when they’re sitting in a room.
The question is how does their score on this measure correlate with other things that are of interest like: how much time they spend on the site, how much and how often they buy, how satisfied they are with what they buy, and so on.
But many people in the tech business talk about algorithms and data mining. How important are the human factors in making good suggestions like product placement or number of items?
Well, I think that both things matter. If you don’t have the wherewithal to do an assessment case by case and individualize, tailor your site to each customer, then knowing in the aggregate what produces satisfaction and what produces conversions, you just design your site around the aggregate (behavior data). It ought to be possible to take advantage of what you call the human factors and give me the right site for me and you the right site for you. There’s no doubt that people differ from one another substantially and you can’t rearrange the shelves of your brick and mortar store for each customer that walks in, but you ought to be able to rearrange your shelves of your virtual store.
What I’m saying is that if retailers took the time and trouble to do that and they did it well, they would not only sell more but they’d have much more satisfied users.
You mentioned that you’re looking to work with folks online like Aggregate Knowledge to test some of your theories. What are some of the tools that you’re looking for?
Well, there are various things that have been studied as psychological characteristics that all ought to have an impact on how people use sites and how satisfied they are with their use. One of them is this scale that I developed with my colleagues, “Maximizing.” Another one is a scale that measures what’s called “tolerance for ambiguity.” Some people just can’t handle conflict, and of course the more information you give them, the more conflicted they are, so you might expect if you gave people that scale it would predict how they use a site and how satisfied they are with it.
There’s another scale that distinguishes people who are kind of approach-oriented – they are looking for something good from people who are avoidance-oriented. They’re trying to prevent something bad. So you’re a pessimist and your main mission is to prevent a catastrophe versus you’re an optimist and your main mission is actually to get something that will be satisfying. That’s a dimension on which people differ and that might also have an impact on how people use these web spaces.
So I’d love to – again, in an ideal world – get profiles of some of these users by giving them some of these scales and then track their behavior and see which of these psychological dimensions can predict the things that the retailers care about. Of course the retailers would be interested in this because it might help them design their sites more effectively. I’m interested in it because it’s a way of connecting these psychological variables to real decisions that people are making.
Right now sites rely heavily on the search bar as a crutch. Aggregate Knowledge talks about this concept of discovery–finding that unexpected result versus typing in a keyword. How was discovery in The Paradox of Choice related?
I think that Aggregate Knowledge is right about the kick people get out of discovery. I’m blocking on the name, it’s a very large big box store that’s based in Seattle – it’s mostly groceries. People leaving that store are more satisfied than people leaving pretty much any other ordinary traditional retail store and there seem to be three reasons: One is that the prices are good, not quite as good as Wal-Mart, but very good. The second that selection is limited unlike other places– there aren’t 100 kinds of toilet paper, just a few. And the third is that they always have surprises available. Even though it’s kind of a bargain store, you may walk in there one day and see that they’re offering expensive smoked salmon. At a very reasonable price.
They get a deal, they buy it, they offer it to customers, a week later it may not be there. And so people do their regular shopping getting the stuff they know they need and also expecting that they will encounter a few things that are going to be pleasant surprises, little treats, little luxuries. And people walk out of there with a smile on their face.
And it’s very rare for people to walk out of shopping experiences with smiles on their faces these days.
Well that’s what Aggregate Knowledge is talking about– the idea that here’s the surprise for you, that we’ve done work for you.
That’s exactly right and I think that if you could pull that off, you’d win real loyal customers and the more complicated the choice environment is, I suspect, the more satisfying it is to people to have these pleasant surprises.
So final question on that note, five years from now on the online side, knowing what you know with The Paradox of Choice, what is the environment going to look like?
My best guess to what it’s going to look like is a lot of boutiques, both in the retail side and even in the information side. People are not going to want sites where you get everything; they’re going to want sites where you can get just the things they’re looking for.
And the trick will be to figure out how to organize what you have, what you know, and what you make available so that for each and every person who comes, it’s a little mom and pop store. And just the right mom and pop store. I think that if something like that doesn’t happen, then it’s just going to implode. People are going to throw up their hands and conclude that it’s just not worth the work. “I’ll just call my friend and ask her.” So I think that we need sites to develop that act as if they are our friends, know us well enough to make the choice of appropriate relevant and manageable.
Barry Schwartz, the author of Paradox of Choice: Why More Is Less. Thank you so much and good luck with your new book.
Silicon Valley Technology Blog - John Furrier 