The Line Between Clarity and Chaos

Posted by

“The problem used to be, ‘how do we get information out to people?’ That problem has now been solved in spades. Now the problem is, ‘how do we filter the information so that people can actually use it?'”

In order to read this article, you got a pointer. Whether that pointer was from a newsletter, your RSS aggregator, a blog, a search, or even just a recommendation from another human, somehow you got here. You got here despite the piles of other stories and articles competing for your attention. This one made the cut.

How did you make that decision? Can we make successful decisions, or are we doomed to indecision? What does the now-terrifying number of options available to us do to our ability to make successful decisions?

How do we, well, choose?

Boxes and Arrows talked with Barry Schwartz, author of The Paradox of Choice, to get insight into new filters, successful strategies, and how the silly number of choices we now have is affecting our everyday lives.

Liz Danzico: Can you choose some key concepts from your book, The Paradox of Choice, that really drove the idea?

Barry Schwartz: We’ve always taken for granted that being able to make choices is good for our well-being. And that, as it turns out, is true. We couldn’t be human if we weren’t able to make choices; trivial ones and significant ones—where to live, what kind of work to do, who to marry, what kind of cereal to buy, the whole gamut. So choice is good. And that’s a truth.

In 50 years of research and psychology, there is study after study showing that people who are able to choose X were more satisfied than people who simply got X. But in all of those studies, the contrast was always with two options. And if two options are better than no choice, then three must be better than two, and four must be better than three, and so on. But no one ever studied that. The empirical basis for the idea is that the more choice people have, the better they are. And it seems perfectly reasonable.

What economists have said, more as a matter of theory than as a matter of empirical evidence, is that if you add options, you can’t make anyone any worse off. If you’re happy alternating between Cheerios and Rice Krispies, you can just keep doing that. And, if I add 50 other cereals, you’ll ignore them. And if I don’t like Cheerios and Rice Krispies, chances are that one of those 50 cereals that have been added will be just the ticket.

Adding options is bound to make somebody better off, and further, it won’t make anybody worse off. The more choice people have, the better they are. So how could it not be true?

It’s not true.

But it’s only in the last five years that people have started doing research where instead of having two options, people have 20. Or 200. And when you cross a line (and you are probably going to ask me “where’s the line?” and I’m going to say, ”I don’t know; nobody knows”), choice goes from being beneficial to being paralyzing. So one effect of too many choices is that people can’t choose at all.

Is that a bad thing?

“So, you overcome paralysis, you make a good decision, but for a variety of reasons you’re less satisfied with the results than if you had fewer options.”

It’s kind of trivial if it means that you don’t get cereal, but it’s not as trivial if you don’t participate in your company’s 401k plan and thereby pass up thousands of matching dollars from your employer. We have evidence to show that’s what happens. Throw 200 mutual funds at employees, and they end up saying, “I’ll decide tomorrow.” Yet tomorrow isn’t any different from today, so they just don’t do it, which is a disaster. So paralysis is one problem.

The second problem is that if you overcome paralysis, and you do choose among a lot of options, chances are pretty good that you will adopt a strategy to simplify the task of choosing. You’ll choose on the basis of criteria that may be easy to evaluate—even though they aren’t necessarily the most important criteria. The result is what turns out to be a worse decision.

There’s a speed dating study that shows when you’re looking at a lot of potential partners, you choose entirely on the basis of looks. It’s hard to evaluate how kind, how funny, how smart, or how nice 12 people are. But it’s easy to evaluate how good-looking they are. You’re meeting more people than you can take the time and effort to evaluate, preventing you from making more difficult assessments of character that are important to you. So you end up simplifying. And that’s not necessarily going to mean a worse decision, but it will likely mean a worse decision than if you could have given it all the complexity it deserves.

So, you overcome paralysis, you make a good decision, but for a variety of reasons you’re less satisfied with the results than if you had fewer options. It’s easy to imagine that one of the options you said “no” to would have been better than the one you chose. So those three things—paralysis, bad decisions, and dissatisfaction with good decisions—all kick in when some line is crossed, and people have more options than is good for them.

I was struck by the passage in your book where you talk about the Internet being “democratic to a fault.” I wondered at which point that fault started being introduced. And how has that changed, even since this book was published?

What’s great about the Internet is that all the barriers to entry that exist in the material world are taken down. These can be barriers to entry with regard to products or with regard to ideas. You have gatekeepers who keep interesting, new ideas out, and thanks to the Internet, you can get around those gatekeepers. The problem is when people start to get around them; there are hundreds of thousands of blogs. Now, the question is which ones do you actually look at? What kind of audiences do these blogs have? How trustworthy are the arguments and the empirical claims? You’re on your own.

So when I say “democratic to a fault,” what I mean is that you come to appreciate that the gatekeepers serve an incredibly important function: most of the stuff that they keep out is crap. They’re filters.

They’re going to filter out things that we’re going to be sorry got filtered out because no one judgment is perfect. But on the other hand, most of what they filter out is stuff that no one should be wasting his time examining. I read 10 professional journals. Each one comes out once or twice a month with 10, 15 articles. And the articles are hard to read. If I read all carefully, reading would be all I did. I don’t read all carefully. I read a few of them carefully—the ones that are right in my area of interest—and the others I kind of read. I look for the executive summary, and I trust that the claims made are validated by the data that were collected. Why do I trust that? Because there’s an editorial process that scrutinizes each one of these things with extraordinary care. Chances are pretty good that if somebody made a mistake, it would have gotten caught, and the paper would have never been published.

So what this means is that I can actually stay on top of my field and have a life because I let the editors do the work for me. In the world of Internet and blog and no barriers to entry, either I’m going to do all the work myself, or I’m going to walk around with a lot of bad information, or I’m going to be so paralyzed by all that’s out there that I’m going to become a shoemaker.

That to me embodies what the problem is. And there’s some evidence out there that news-type blogs have increased in their number. People gravitate more and more to websites that have a print-media presence.

Now that’s interesting. I assumed you were going to go the other way with that. I figured people were going away from websites with a print component.

Newseek. Nothing could be better for them than having a million blogs out there because people don’t know how to choose. So what do they do? They simplify the decision, and choose on the basis of brand. And that’s exactly what will happen if there are 30 different kinds of cola in your supermarket. You’ll buy Coke.

Does that hold true for millennials, the generation that has grown up with this kind of choice, the technology to support the finding of this information, and the number of choices? Do you think that generation also looks at these brands as trustworthy?

They may not trust Newsweek, but they are going to be brand dependent; maybe a different brand. It may be what gives a site credibility will be determined by key cultural influencers of their generation that they take seriously. But do I think that what’s going to happen is that they’re going to shop for their information by brand. Absolutely.

Now, Chris Anderson and I had a debate about this on the radio because his whole premise in The Long Tail is that this is all good. He acknowledges my arguments, and he’s optimistic because he thinks that the problem technology created, technology can solve. And that may be true. It may well be that Google and Google-like things that are better than Google (although that’s hard to imagine) will provide the sort of filtering devices that make us less brand dependent.

The problem used to be, “how do we get information out to people?” That problem has now been solved in spades. Now the problem is, “how do we filter the information so that people can actually use it?” And that problem hasn’t yet been solved.

But since search companies did such a great job in providing the information, why not trust that they’ll also do a good job in filtering it? It’s possible. But if that’s the case, there’s going to be a technology fix to this problem. The great Democratizing of the Internet will be much less of a democratizing because some entity like Google will be doing the filtering, so there now are barriers to entry again. And we’ll be at the mercy of filtering algorithms that we don’t know how they work. We’re going to have to trust that Google has the interest of truth at heart. It’s going to be presenting us with a structured list whenever we type something into search. And we’re going to have to hope that’s appropriate for what our interests are.

And that doesn’t strike me as democratic.

More like traditional editorial.

“The consistent problem in all of this is that people don’t know what’s good for them.”

That’s right. But it’s under the guise of being democratic. It’s much less obviously managed than a newspaper, but it’s not less managed. It’s possible that some really incredible, smart person will come up with a way of doing this so even the filtering is transparent, so people know what the editorial criteria are, and can choose their search engine based on the criteria that are most important. But that hasn’t happened yet.

So it’s possible that people in their 20s and 30s (who are much more comfortable with this sort of technology than I am) have already figured out ways to grapple with the choice-overload problem in a way that people my age haven’t and won’t.

In your book, you talk about an evolution: an evolution from when people foraged for food, the intro of manufacturing. And now, or at least when the book was published, the trend is moving back to this time-consuming behavior of foraging for information. Do you feel like people are now conditioned to not be able to forage? You talk about people falling down when it comes to these choices. Is this just a result of us being conditioned to not be able to do so?

I don’t think so. Back in the foraging days, the criterion was, when you went searching for food, “Is this good enough? Can I eat this? Will this not poison me?” You weren’t wandering around in the forest looking for the best berries; you just wanted berries. And as long as that’s true, people can forage just fine in the information world. But increasingly, people believe that good-enough berries are the point of a search. You need the best berries, and people have the illusion that it’s actually possible to find the best, and that it’s worth the effort to look for the best. And that’s what does us in.

You’re talking about maximizing, the attribute of a person who seeks and will accept only the best. Are you suggesting that the number of choices we have is conditioning us to be maximizers?

I don’t have any evidence that that’s true. People differ from one to another to the extent they feel they have to maximize. And my hypothesis is that when there’s an extraordinary range of options, people maximize. I mean, what sense does it make to maximize when there are only three kinds of jeans? What does it even mean? When there are 3,000 jeans, well, now you think, “One of them is probably perfect. I gotta find it.” Whatever “perfect” means.

But I don’t have evidence that that is true. The predication would be that in less-affluent, less-industrially developed societies, fewer people are going through life maximizing. So I don’t know that we have generated this problem for ourselves, but I suspect that we have.

During your talk at Google earlier this year, you talked about the idea that people no longer have to deal with the hand they’re dealt, and you cited plastic surgery as an example. That’s disturbing. I wonder what the implications of that behavior might be for us as designers. How can we design for people who know that they don’t have to deal with the hand they’re dealt?

The consistent problem in all of this is that people don’t know what’s good for them. If you offer people a limited range of options and a large set, most people will choose the large set. They’ll go and try to pick something, and they’ll walk out empty handed shaking their heads. So everyone’s kind of swallowed the ideology that more is better than less.

And so they’re going to demand from you, as designers, something that makes everything obviously available. And if you’re going to be successful, you’re going to have to provide people with what they want even though it’s not what they should want. Now, this may be changing. The response I’m getting to the book suggests that I haven’t told people anything they don’t already know. What I’ve done is identify and name something that almost everyone experiences in some or many aspects of their lives.

People tell me I like to shop at boutiques. I used to think I like to shop at boutiques because I got pampered. Now, I think I like to shop at boutiques because I’m, in effect, hiring someone to edit the choices for me. Anything that’s in this store is worth looking at. And that’s good.

So I think there’s this self-awareness that’s now developing in people that more isn’t necessarily better. In how many people? I don’t know. And in what areas of life? I don’t know. But if that were to happen, then designers could actually think about designing for what really is in the best interests of people rather than what they think is in their best interests. Because what is in their best interest is what they’ll be looking for.

I think there are ways to do this, although it’s a technological challenge, which is, you can make everything available but make it hard for people to get access to everything. And you make it easy for people to get access to things that almost everyone cares about. Another way of saying this is an enormous amount of attention should be paid to what the defaults should be on any website. What happens if you do nothing? You want to design a website so that what happens when people do nothing is what they want. Then leave it to people to act if the defaults aren’t what they want.

I read an interview with you and Mark Hurst before the GEL conference and your homepage idea: listing the top five bestsellers on the Amazon site, and having the rest available, but just making them a click away.

We tried to get [Jeff] Bezos to let us do an experiment that was exactly that experiment. We said just give us two hours instead of presenting the 20 bestsellers, present 10. And we predict that you’ll sell more books. He wouldn’t do it.


No, he wouldn’t do it. You’re not making the other stuff unavailable; you’re just making it a click away.

Well, the studies would indicate that you’re exactly right in terms of sales. The classic jam study.

Now books may be different from jam because there is not a sense that when you buy Stephen King, you also can’t buy John le Carre. Books are not competitors with one another in the same way that flavors of jam are. So it’s conceivable that the jam study is appropriate for those kinds of goods, but not for things like books. But it certainly would have been worth two hours of Amazon’s time to find out.

Seems like you would have competing stances. People like Paco Underhill might recommend putting the bestsellers at the back of the store with the milk because people know that they’re there. And then Chris Anderson might suggest bringing forward the titles that no one knows about. I would love to see the three of you design a homepage for Amazon.

The power of defaults is this:

When you enter your drivers’ license, you get asked if you’d like to be an organ donor. And if you do, you have to check a box and sign a form. And in the United States, 90% of people approve of organ donation but 20% are organ donors. There are several European countries that also use the drivers’ license as the opportunity to sign up organ donors. In those countries, organ donation is 90%. And the only difference is, in those countries, you have to sign a form and check a box if you don’t want to be an organ donor. Otherwise you are.

Designing websites is the same way. What you get if you do nothing is what you want, and that will have an enormous impact on how user-friendly people find the websites.

Your book has been incredibly well received across different industries and disciplines. Did you have any idea it would be as widely received as it has been?

I wrote the book mostly with an eye toward the people who are tortured by all these choices—people like me who face what modern America has become. It was an attempt to explain why it is a problem and then to make some suggestions about how people might cope.

It never crossed my mind that there might be an audience for this in the world of business that is creating the problem. Then, it ends up getting called one of the “ten best business books of the year” by Business Week and Forbes. And I keep getting invited to talks at supermarket professional trade associations and luxury travel trade associations and Microsoft and Google and user experience and wholesale flower industries.

So it turns out (much to my delight) that the merchants in every conceivable industry who found out about the book realize that what I was telling their customers was also a lesson to them about what customers might actually want and appreciate from their providers. Whether anyone actually has the courage (or can get them through a complex organization) to make the kinds of suggestions from my book that will make customers happier remains to be seen. At the very least, lots of people are taking the ideas quite seriously, which is, I must say, fantastic.


  1. Great interview. Thanks, Liz.

    In terms on navigation and page design, I think you also have to take the mode of seeking into account when deciding on the number of options. For known-item seeking, having more trigger words on the page should increase information scent. That’s why Jared Spool and even Forrester recommend making your home page the site map.

    But if a visitor is just casuallly browsing best sellers, then yeah – having fewer might be better. It all depends, of course, but challenging the notion of “more is always better” is an important consideration. This is why a thorough understanding of what people are doing on your site is so important. Otherwise, you’re just guessing.

  2. A great interview, Liz! I have referred to Dr. Schwartz’s work on a couple of Podcasts I’ve completed to date and his book on the paradox of choice.

    Another interesting book, related to these ideas is “Enough – Staying Human in an Engineered Age” by Bill McKibben. As he states – “We need to do an unlikely thing; we need to survey the world we now inhabit and proclaim it good. Good enough. Not in every detail; there are a thousand improvements, technological and cultural, that we can and should still make. But good enough in its outlines, in its essentials. We need to decide that we live, most of us in the West, long enough. We need to declare that in the West, where few of us work ourselves to the bone, we have ease enough. In societies where most of us need storage lockers more than we need nanotech miracle boxes, we need to declare that we have enough stuff. Enough intelligence. Enough capability. Enough.”

    Once we have limited choice to improve overall satisfaction with our decisions, we also need to proclaim, as Mr. McKibben points out, we have enough of what we need. I think from that point, we can start to look at how much more we can contribute, rather than trying to keep up with the pace of life, technology, and the like.

  3. Thanks for this interview, and props to Liz for being able to collar the man for B&A.

    I won’t pretend to know more than Shwartz on his subject, but I think his comments on the question of the Internet being “democratic to a fault” and the importance of gatekeepers and editorial are somewhat beyond his ambit, so I feel justified in being able to disagree with them. There is not a binary choice between trusting a single gatekeeper, editor or “brand,” and drinking from a fire hose. Collaborative filtering can be just as effective (and transparent in operation) as centralised filtering, and methods of calibrating trust and editorial strength dependent on perceived importance (to the individual consumer of that information) can and are becoming quite sophisticated. I find that people like Shwartz often don’t seem to understand networks and technology properly because they choose to see them in terms of their observation of centralised or linear systems (like publishing) or established models like classical economics. The situation here is different, and his mistrust of technology does him no favours in that respect.

    On a minor point, I’m also adding this to my collection of quotes from people who think that Google is the last search engine we will ever have.

  4. Great interview! Schwartz’s insight definitely help inform some of the design decisions we make but I have to agree with Jonathan on the comments Schwartz was making about gatekeepers. Depending on the gatekeeper’s agenda, information can be as erroneous or misleading as having an abundance of choice.

    Schwartz mentioned his preference for a range of editors who help him filter what he reads. As Jonathan already alluded, this happens constantly and in more interesting ways on the web. Implicit actions like collaborating filtering or explicit actions such as voting can act as editors in their own rite. Regardless, the book definitely helps our case against “features for the sake of having features”.

  5. Very cool article, Liz. What captivated me was the concept that overburdening people with choices can actually limit decision-making.

    At my agency, we operate in part on an experience theory called “directed choice”. Oversimplified, directed choice is a method of presenting brand benefits that takes some liberties with assumptive user tasks–much like a salesperson would engage a prospect.

    The design paradox is that we intentionally limit the choices users can take, discard the traffic that isn’t really interested at that moment (or better yet keep them around to draw into a marketing database, viral tools) and then move warm prospects into buying or communicating with a real person. One of the more effective examples of this is

    This type of high-bandwidth interaction certainly does not work for all users, information-gathering, task scenarios, or decision-making styles, but limiting choices and encouraging exploration can often be highly effective for branding and selling interactions.

    Keying off of Jonathan’s comments: one thing I would LOVE to try is some kind of combination of rich media directed choice and collaborative filtering like Amazon or Netflix. How cool would that be?

  6. I don’t see “information overload” as any problem whatsoever. There are filters that we use mentally, individually, called web savvy. Thinking and judging is good for the mind and society. Why should information always be easy, simple, and limited?

    Information wants to be, and always already is, infinite. There is some increase in information, but also increase in hunger for information, and increase in new ways of distributing and accessing information.

    To bewail the abundance of content, to regret the influx of garbage, is not the best use of our time and effort. All communication channels, including fully present speech, is loaded with noise. Not just blogs and YouTube.

  7. I also take exception with the very dangerous wording of “democratic to a fault” which strongly implies a distrust of “too much” democracy, like “too much” civil rights for blacks, or “too much human rights”, “too much education”, “too much happiness”, or “too much political freedom”.

    How is something taken “to a fault”? Is the fault pre-arranged, pre-existing? Or is the fault an outstreaming of a contamination inherent in a corrupt system? Is democracy such a system?

    What does it mean to be distant from a fault, and is there safety in distance? If something can be taken “to a fault”, is there not the assumption on the part of the author that that democracy is already “faulty”, since it was vulnerable, perhaps secretly wishing, for this fault to come along and seduce it?

Comments are closed.