We Don’t Know What We Want & We’re Unhappy When We Get It: “The Art of Choosing,” by Professor Sheena Iyengar

I went back to my alma mater, Columbia University’s Graduate School of Business, on March 25, for a presentation by Sheena Iyengar, S. T. Lee Professor of Business, on her new book, “The Art of Choosing.” Her framework about choice and how we choose was thoughtful, insightful and, most importantly, clearly articulated several concepts that I had been grappling with recently.


My notes are below, with the caveats that they are my best rendition of the discussion, with no claim to accuracy; much of the discussion has been paraphrased, rather than being a direct quote of the participants; and I bring my own viewpoint and perspective to this discussion, which influences my perception of the discussion. I am sure that many of these studies and anecdotes are much more compellingly set forth in her book, which I have not yet read, and that I have hardly done them justice in my rendition of her presentation. I look forward to reading her book and attempting to more fully understand the fascinating work she has done.


Professor Iyengar was introduced by Eric J. Johnson, a founding co-director of the Center for Decision Sciences at Columbia Business School.  Professor Johnson hailed Professor Iyengar for having introduced one of the most famous theses of decision research - too much choice.


Professor Iyengar began by discussing the broad general theme of making choices.  She began her remarks by pointing out that if anyone in the audience had never made a choice, then her presentation would not be a relevant session.  As expected, not a soul left the room.


We all have dreams about how our lives will unfold – she used to dream about becoming a pilot.  After all, our schoolteachers said that we could be anything we set our minds to.  Presumably at a fairly young age, she realized that she could not in actuality become pilot without her eyes – Professor Iyengar being blind.

In truth, we spend our lives struggling against our limitations.  Sometimes we succeed, sometimes we don’t.  As she traveled, she found that her lack of eyesight prompted unexpected reactions from local residents: in Spain, people gave her money for lottery tickets, while in Japan, people expected her to give them massages. [Caveat: these observations of hers are totally foreign (no pun intended) to me, so I can only hope that I heard them correctly, since they seem quite bizarre to me, culturally conditioned as I am.]

In general, people kept telling her what she couldn’t do due to her blindness, so she was constantly trying to understand her true limitations vs. those that were merely perceived.  In her quest, she attended the Wharton School at the University of Pennsylvania where she confronted the question of what she could be when she grows up. [Note: the minefield of expectations (see later reference to the “Hurt Locker”) from which we choose is littered with false negatives and false positives – we are neither capable of “being anything we want to be” nor necessarily limited by outmoded stereotypes.]

She took a course on social psychology that changed her life.  No one had ever said that a blind person couldn’t be an experimental psychologist [although I’m sure that it was not on the top of anyone’s suggested career list for blind students, even at Wharton] so she asked to be a lab assistant for the course.  An experiment was created utilizing her somewhat unique situation that compared how subjects regarded feedback received from a blind person vs. that received from a sighted person.  The experiment failed but had the pleasant outcome of launching her on her career path.

For one thing, it was the first time that her blindness was perceived as an asset.  She chose [if you’ll pardon me for using that word] to become a researcher of the role that choice plays in our life.

In her view, choice is the tool we wield when we come up against our limitations.  We want to choose our way to happiness, but in reality, choice isn’t everything.

She feels that she took advantage of the career choice that became available to her because she knew what her true limitations were.  She didn’t have the same career choices as a sighted person, and, in that sense, it made her decision easier.

At the same time, she felt that her physical limitations placed a higher premium on her choice – she couldn’t afford to choose on a whim.  The language of limitations, and presumably the how she coped with it, became second nature for her.

We Don’t Know What We Want:

Her thesis is that we can take the most advantage of our choices by acknowledging the limits of those choices.  We believe that we’re experts in knowing what we want, so more choice is therefore better.

But do we know what we actually like?  Do we even know whom we find attractive, a choice that would seem obvious, ingrained and immutable?  She cited an experiment in which students had to select between two photos of reasonably comparably attractive people, such as:

Subsequently, the subjects were asked to explain their choices.  The experimenters, however, had switched the pictures.  In other words, a subject may have initially selected A over B, but later was told that they had selected B, whereupon they proceeded to explain why they had do so in preference to A, when in fact they had not.  [Note: I think that the mark of an excellent presentation is one that answers lots of questions yet raises many more.  (At least it helps keep me awake and engaged.)  I would love to know what percentage of subjects couldn’t remember or were easily influenced into false memories, something that I am certain would never happen to any regular reader of this blog!]

Net net, we all [well, maybe you.] have trouble remembering and reconstructing our preferences after the fact, even on a very basic level, as in how we would like to continue to propagate our species.

Of course, as Professor Iyengar put it, this phenomenon also happens with “important” choices.  In another experiment, she tracked college students during their graduation year.  They were asked about their job preferences and attributes at several points during that year.  She found that they kept changing their minds to the extent that the correlation of their choice at the beginning of the experiment in September to their choice at the end in May was 0.15.  In other words, next to none.  [Note: it would be interesting to have asked the students whether they recalled their earlier choices and whether they thought that they were being, inaccurately, consistent, or whether they recognized their changes in preference.]

If she is correct that we don’t actually know what we want [or it keeps changing even when we think our preferences are immutable] are we better off getting what we want or wanting what we get?  [Note: perhaps her research points to the answer to that age-old musical lament: I Can’t Get No Satisfaction.  Maybe Mr. Jagger should just be satisfied with what he has, which I’m sure is plenty, in any case.]

Professor’s Iyengar’s initial remarks implied that our choices are somewhat random or not well-grounded, even though we consider them to be the product of careful thought and deliberation.  She next set forth a framework that, in my mind, showcases the human condition in an even less-flattering light. 

We Let Others Decide for Us:

She shifted into a discussion about how choices reflect back on the person who made those choices.  Our society, and by extension, each of us, associates choice with freedom – in other words, choosing is perceived as an act of freedom.

We therefore use choice as a means of self-definition: by asking ourselves what we want, we are asking: Who are we?  What should I choose? 

We try to express ourselves through our choices in order to assert our individuality and distinctiveness.  But this self-expression can actually be an obligation rather than a freely selected “choice.”  As an example, Professor Iyengar cited an episode involving her husband.  She had gone to the Apple Store to purchase an iPhone for him.  He was quite emphatic that he wanted a black one, allegedly due to certain characteristics of a black phone, such as being less likely to show dirt and scratches.  While she was in line, he rushed into the store to find her and tell her that he had changed his mind to white.  His rationale apparently was that everyone else is buying black; therefore he needed to purchase an iPhone of a different color. 

She sees a similar phenomenon among patrons who order food and drink in restaurants – they do not want to be seen as copycats.  Therefore, if someone else selects the item they had originally intended to order, they will often either change their original preference or feel the need to justify it on the basis of having made their decision prior to the other patron’s disclosure of their order. [Note: is our self-esteem so delicate that we would be considered feeble-minded to say about a fellow diner’s order, “Gee, that sounds good.  I’ll have the same!”? Apparently so.]

Professor Iyengar characterized this behavior as the Drinker’s Dilemma – in situations where patrons at selected tables secretly wrote down beer orders, overlap often occurred among the orders, as would be expected.  These patrons, however, were generally more satisfied with their beer-guzzling experience than patrons who had given their orders out loud. The exceptions among the publicly-announcing patrons were those who had ordered first.

In other words, if someone who orders before me selects the brand of beer I was planning to quaff, I will deliberately “choose” a sub-optimal outcome of a malt beverage of less preference simply to avoid the appearance of aping my fellow patron. Scary.

Please note the definition of “aping” in Merriam-Webster: “to copy closely but often clumsily and ineptly.”  Well, of course, who would want to be accused of such behavior?  It calls to mind the episode of “Seinfeld” in which our eponymous hero proclaims loudly, “I am not an animal!”  Apparently, according to Professor Iyengar’s research, ordering the same beer as my drunken buddy constitutes disqualification from the Homo sapiens club!

Again, the question of choice seems to be less about what do I want and more about who am I and what do I know.  We think of choice as a way to individuate ourselves.  In this construct, we never choose alone.  Choice is a means of communication, consciously or unconsciously.  They constitute gestures that are interpreted by people around us.

In this process of self-definition and signaling to those around us, we want people to see us as unique individuals, but not as outcasts.  [Note: this was one of the key insights of the evening.  It’s obvious in retrospect, but that simply speaks to its insightfulness.]

Professor Iyengar, in conjunction with Professor Dan Ariely of Duke, had done a study of choices of baby names and neckties.  In general, they fell into three categories: plain, a little unique and bizarre.

Preferences generally seemed to be for unique, with a general avoidance of bizarre.  [Note: is this a tautology?  Generally by definition, very few of us want to be viewed as “bizarre.”  Similarly, behavior with which we would not want to be associated we would characterize as “bizarre.”  Which comes first, the behaviorial chicken or the characterizational egg?]

When asked to characterize themselves and their choices, subjects saw themselves as having more of a preference for unique choices than other people.  [Note: does this mean that we are all “special.”]  But at the same time, they generally liked the same things as everyone else.

In summary, we all aspire to be unique – but we also want our choices to be understood by those around use.  Achieving that societal positioning requires balancing the fine line between unique (“good”) and bizarre (“weird”).

In order to choose, we each have to answer for ourselves questions including: Who am I?  What does that person want?  What should that person choose?  What would other people choose so as to be different, but not too different?

Professor Iyengar seemed to imply that these sorts of calculations often resulted in answers that differed from our “real” and unencumbered choices – presumably the choices we would have made if no one were watching?  [Note: remember her earlier research that indicated we don’t actually know what we want, although we think we do.]  This sort of dissonance can result in our choice of things we didn’t really want and ending up being dissatisfied [Note: but wasn’t that by our choice?  About this time, I feel the need to choose an aspirin!]

[Note: I am reminded here of a characterization of winning in the stock market – the best choice of a stock is not the company that YOU think is the best;

it’s the company that you think EVERYONE ELSE thinks is the best!]

Summarizing the presentation so far: We don’t really know what we want, although we think we do.  We are also willing to make choices that we think will impress or influence those who are aware (or even care!) about our choices.  Could it get any worse?

Let's Add More Choices, Just Because We're So Good at Choosing

Of course, because Professor Iyengar went on to demonstrate that the process of choicing is further complicated by the explosion of choice.  [Remember, since we think we are free-thinking actors and choice is an expression of our individuated special selves, more choice is better as it gives us more opportunity to express ourselves, especially in the purchase of items we didn’t need in the first place!]

She cited the exponential growth of the number of items in a typical grocery store from 3,700 to 45,000 [Note: sorry, I missed the time frame, but even such growth since the emergence of the modern supermarket would be pretty impressive.]

She apparently became interested in the implications of this sort of phenomenon during graduate school at Stanford University in northern California.  She found herself going into a local grocery store with lots of choices in a variety of product categories, yet she noticed that she often came out empty-handed.  She asked the manager if people were taking advantage of the multiplicity of choices and buying things, such as a selection from among the store’s 348 kinds of jam. 

Professor Iyengar decided to use jam, an innocuous product, to test the impact of choice on consumer purchases.  She and her team set up a tasting station near the entrance to stop customers on their way in to the store and offer them a selection of either 6 or 24 flavors of jam.  They then tracked the proportion of incoming customers who stopped at their station to sample some jam and those who subsequently purchased jam, identified through special coupon designed to track sales.

The results were that more people stopped at the table if there was a higher selection - 60% of incoming customers decided to check out a sampling of 24 flavor, while only 40% of customers paused to inspect a selection of 6 flavors. [Note: obviously, customers were presented with one selection or the other as the contrasting sample sets were never presented simultaneously.]

The plethora of choices, while encouraging customers to stop, had the reverse effect on their propensity to purchase.  When there were 24 choices – only 3% of customers subsequently bought jam while 30% of thosre presented with the relatively meager selection of 6 flavors ended up making a purchase.

In summary, while having more options is viewed by consumers as being initially more attractive, they are conversely more likely to choose from among a few selections than from a lot of options.

Professor Iyengar has hypothesized that our cognitive capabilities are overwhelmed when we have to do the math of comparing and contrasting numerous options.  How many choices can we handle?  According to a gentleman named George Miller (whose bona fides I did not catch), we are limited to what he calls The Magical Number: Seven plus or minus two.  We are restricted by our limited capacity to store information in the DRAM active memory of our brain.  As a result, when confronted with too many choices to easily process and from which to make a selection, we default to the “no choice” option and decline to choose at all.

She cited the example of declining participation in corporate 401(k) plans that directly correlate to the increased number of  fund options – bonds, stocks, emerging markets, oil & gas industries, pharmaceuticals, junk bonds, etc., back in the day when people actually harbored a hope that they could provide for their retirement through prudent investing. 

Regardless of domain, as the number of options increases, people delay or choose things that are worse for them.  [Note: I suppose that, as the number of choices increases, the opportunity for sub-optimal selection increases even faster.  If I have two choices, assume one is better than the other.  If I now have 10 choices, I’ve increased my pool of options by 400%, but the number of sub-optimal choices has increased by 800% from one bad choice to nine potentially bad choices.]

Even when we make the “correct” choice, we are often less satisfied, presumably because we are fretting about whether a different choice would have been preferable.

So how do we deal with a situation in which we have to know ourselves and what we want, compounded with multiplicity of choices?   Professor Iyengar posits that people get better at choosing with practice.  For example, we can acclimate ourselves to choosing from a wide selection of options.

She cited the example of German car buyers who, on the plus side, are able to custom-design their own automobiles.  On the negative side, that means making 67 separate decisions across the range of trim, engine, body paint, etc., with each decision having between 4 and 56 options each – such as a selection of audio options including satellite radio, in-dash CD player, 8-track tape player [OK, maybe not that one.], etc.

Her team did not change the substantive options themselves, but tested the impact of how the choices were structured.  Their first approach was to “Start Deep” – begin with decisions requiring choosing from among many options (let’s say 56) and proceeding to decisions with fewer options, such as 4.

They also tested the reverse “Start Shallow” approach – start with a smaller number of choices and progress to decisions involving larger numbers of  choices.

In each case, the number of total decisions and choices remained the same – they simply changed the order of presentation.  Yet they found that the differing frameworks of choices had profoundly different impacts on ability of consumers to wend their way through the labyrinth of self-definition through personalized expression on the Autobahn.

Starting shallow seems to enable consumers to become acclimated to the process of sorting through choices and making decisions.  By easing them into the process, the auto retailer was in effect teaching the consumers to get comfortable with forming preferences.  As Professor Iyengar phrased it, “learning to figure out what I want.”  [Note: how many of us will voluntarily admit in mixed company that, not only do we not know what we want, but that we don’t know how to figure out what we want?]

In other words, they experienced less decision fatigue than if they started deep with multiple choice decisions, in which case, decision fatigue kicked in sooner.

Not only that, but those who started shallow scored higher on every measure of satisfaction with the outcome.  Net net, a more pleasant process and a perceived higher quality outcome.

Conclusions of this sort are not of interest only to pointed-headed academics (I’m just checking to see who’s reading this!).  Armed with this knowledge, we can re-orient our behavior to conform with our heretofore unknown cognitive limitations and preferences.  For example, Professor Iyengar felt that, when shopping, we tend to start with the stores that have the most options.  We are probably better off starting with stores that have fewer options and working our way up to the more complicated stores – a tall Cinnamon Dolce Frappuccino Light Blended Coffee, anyone?

Another terrific coping mechanism she discussed was that of “categories” that enable our minds to differentiate sets of choices, presumably by grouping choices so that we are not necessarily making a multitude of individual choices but are filtering groups of choices instead.

She cited a wine store called Best Cellars.  While it addresses the excess choice issue by only having 100 selections of wine, its true innovation seems to be in having created 8 categories of wine from which patrons can make their selections - fizzy, fresh, soft, etc.  Once a category has been selected, one presumably then chooses a wine from among those that fit that category.

In broad terms, categorization enables you to learn more about yourself and what you want – in other words, it helps you to choose. [Note: an interesting circle – you choose in order to define yourself; knowing yourself helps you to choose.]

Meaningful Decisions (or, Maybe Hobson Was Right):

Professor Iyengar then switched gears entirely to a much more profound discussion about the concept of choice.  She argued that we can’t or shouldn’t always want to choose, that there are some choices that we’re not prepared to make or would want to prepare ourselves to make.  She felt that we may be confronted by what she called “choosing domains” – that an obligation to choose requires that we choose as often as possible.  This issue was encapsulated by her question: if you had to make a life-altering choice where no options were desirable, would you want to have to make a choice?

She gave the example of a baby born with a cerebral hemorrhage.  The choice to be made was to remove life-support from the infant, leading to death, or to wait.  The outcome of the latter would range somewhere between death and life in a vegetative state.

She illustrated the perils of the implications of choosing by comparing parents in Paris and Chicago.  In all cases, life support was removed and babies died.  The difference in the framework of choice was that, in Paris, the doctors decided, while in the U.S., the parents made the choice.  [Note: again, I found her presentation compelling because it raised additional questions: haven’t the doctors essentially done the choosing in each case by the way they framed the question and making the choices so stark?  If I give you the choice of bad option A or bad option B, which includes A with some worse stuff thrown in, I’ve basically guided you towards option A.]

The parents were the subject of a follow-up study six months later, and significantly different outcomes were found among the parents centered on the issue of who had made the life-or-death decision.  The findings were that the American parents were more depressed than the French parents, who were more likely to have moved on with their lives, perhaps even to the extent of planning for other children.

In looking back at this experience, the study found that the French parents focused primarily on the experience of the baby and the place it held in their memories.  The American parents’ memories were focused on the choice that they had been required to make.  Under those circumstances, the French parents could understandably have taken some solace from the introduction of a new life into the world, albeit briefly, while there seems to have been little upside to American parents endlessly revisiting their choice of horrendous option A and even worse option B.

Interestingly, the parents were also asked their opinion of the framework of choice that they had experienced.  When asked if they would have preferred about reversal of choosers, each set of parents felt that the appropriate parties had been given the responsibility of making the choice.  The French parents felt that the doctors were in a better position to choose.  They did not, however, seem eager to make the choice in any case.  The American parents, on the other hand, felt that it was their duty to choose, but felt trapped by this choice.

If choosing is supposed to set us free, why do we allow ourselves to become slaves of choice?

As a society, we reject the language of limitations – we want to be able to choose, or at least, so we think.  In viewing our lives as objectively as possible, we sometimes might be better off to say that we CAN’T choose.  After all, not every opportunity to choose is an opportunity to improve our lives. [Note: that’s true, but everything is relative.  Choosing between two negative outcomes is still an opportunity to improve the net quality of our lives.  Living with the burden of having made that choice, however, can understandably outweigh the perceived benefit of choice and control.]

Choice is a powerful tool – it enables us to go from who we are to who we want to be.  All the same, however, it doesn’t solve all of our problems or fulfill all of our needs.  Sometimes it’s too much.

Another wonderful insight from Professor Iyengar: To get the most of out of choice, we have to get choosy about when we choose.  [Note: the concept of meta-choice – choosing when we want to choose – therefore becomes a fundamental guiding principle of our lives.]

She played a brief clip from the Oscar-winning move “Hurt Locker” showing the protagonist dealing with a situation where he has to make lots of decisions and there are a multitude of choices among which to decide.  The contrast she was illustrating was that the hero makes decision after decision with calm in Iraq when he is confronted with life-and-death situations.  But upon his return to the U.S., he is paralyzed by the choice of cereal in a grocery store.  As a result, he apparently returns to Iraq to resume his earlier mission of defusing and destroying hidden bombs, in other words, to restore meaning to his choices.

We begin to believe that all of our choices are important to us.  We must focus on the choices that really matter.  In that way, we can balance our hopes and desires with clear eye on our limitations.

Questions from the audience: 


Q: how do we let others choose?  For example, Obama’s health-care insurance bill received Congressional approval this week.  An issue seemed to be that of choosing between the government or insurance companies making decisions regarding our health.

A: In the West, we value our freedom.  Under the status quo, we had the perceived freedom to choose our insurance company, who would then make choices about our health.

Q from me: her earlier example of students and career preferences indicated that students convinced themselves that they were happy with their choices, while the beer drinkers were less than happy because another tippler had ordered the brew which they actually preferred.  When do we rationalize our choices and when do we grouse about them?

A: Her study indicated that students were happiest when they believed that what they got was what they wanted.  She differentiated between maximizers, who seek more choices, and satisficers, who are more willing to be less thorough.  Apparently, the students who were maximizers got more offers, as would be expected.  The outcome, however, was that they were less happy with their ultimate choice, presumably due to concerns that a different choice would have been preferable.

Furthermore, even if one had such tendencies, it’s not practical to attempt to maximize on every decision with which we are confronted – there are simply too many decisions and choices.  In many situations, we’re just going to have to be satisfied to be a satisficer.

Q: people own what they choose.  People don’t resist change, they resist being changed.

A Going from zero choice to some choice is enormously powerful.  Therefore, we don’t want to get rid of choice, we just need to manage it.

Author’s Summary: The act of choosing has a myriad of implications.  It connotes freedom.  It gives us an opportunity to define ourselves.  At the same time, it is fraught with negative implications.  I can’t choose because it’s too complicated.  I can’t choose because the consequences are too severe?  What if we make the wrong choice and I’m unhappy?  Or people think we’re bizarre? 

The good news is that we can train ourselves to both recognize the choice frameworks we are given and decide how, or even whether, we choose to choose.  A trivial example: if I can’t decide between the ribs and the fish & chips, I might as well choose at random.  It’s not likely that I’m going to get any additional information that will help me decide.  And, what’s the worse than can happen?  To the contrary, what’s the best that can happen – making the right choice is likely to only have minimal impact on my long-term well-being, so why expend what few neurons I have left agonizing.  Furthermore, I am trying to train myself to live with that decision – even if the choice turns out poorly, let’s not have a nervous breakdown over it. It doesn’t really matter, and besides, the other choice could have been worse.

When I chatted with Professor Iyengar about this issue during the reception, she said that she allows her husband to make menu selections for her because, for whatever reason, it’s more important to him than it is to her. 

Now, that’s a woman who knows how to choose!