Choiceology with Katy Milkman
Hosted by Katy Milkman
Why do tiny gaps near round numbers alter how we perceive value? Learn how small thresholds can drive big differences in decision-making.
5,546 words · 8 speakers
Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.
David Gold grew up in Cleveland, Ohio, helping his parents run their general store. They moved to LA when David was a teenager, and by his 30s, he was running the family's liquor store. Gold noticed that whenever he posted a price ending in 99 cents on a slow-moving wine, it sold out. He tested round-number pricing, too, but that didn't work as well. Prices ending in 99 cents brought the magic. He thought, wouldn't it be fun to have a store where everything was 99 cents? And so, the 99-cents-only store was born.
Egg hunting is a lot like deal hunting. So this Easter, 99 your basket.
You do the 99. So really, it's just another day at the 99. Or better yet, I can 99 my birthday bash, right?
The store opened in 1982 with a line outside the door. Within two years, two more locations opened. And by 1991, there were two dozen. The company never hired a publicity agent. Instead, they relied on word of mouth and humor. They took puns on the number 99 to the max with signs that read, "Open 9 to 9, 9 days a week", or "99.99% satisfaction or your money back." So much love. Go 99ers. It wasn't a business model that could last forever. The chain closed due to the pressures of inflation on their unchangeable prices. But at one point, there were over 300 stores located in California, Texas, Arizona, and Nevada.
David Gold had tapped into something we all recognize. In this episode, we'll look at why a price, or an age, or a test score that falls just under our round-number threshold can have an outsized impact.
We'll describe a bias that can influence our decisions about the music we buy,
infringements aren't slowing down Napster, the song's swapping platinum...
how we categorize diamonds,
top-quality diamonds to optimize brilliance, not care at weight, resulting in remarkable shiny.
and even the health care we receive.
I'm Dr. Katie Milkman, and this is Choiceology, an original podcast from Charles Schwab. It's a show about the psychology and economics behind our decisions. We bring you true and surprising stories about high-stakes choices, and then we examine how these stories connect to the latest research in behavioral science. We do it all to help you make better judgments and avoid costly mistakes.
They were scrambling, trying to come up with ways to deter people, which included suing all sorts of people, including 12-year-old kids and grandparents.
This is Melina.
Hi, my name is Melina Palmer. I'm founder and CEO of The Brainy Business and author of The Truth About Pricing.
Melina is remembering the late 1990s and the chaos in the music industry as it made an awkward transition into the early days of digital music.
Sites like Napster had popped up where consumers were able to get free music, really wherever they wanted. Problem, of course, being it was stealing, right? So, not good on that side. But, you know, from a consumer perspective, you were able to go get whatever songs you wanted for free at any time. The issue, of course, then for the music industry is they were losing a lot of money and really in a free fall.
Consumers were ready for digital music, but their options at the time weren't great. They could buy physical CDs and turn those CDs into MP3 files. But that was time-consuming. And if you only liked one song off of an album, it was expensive. And peer-to-peer file-sharing sites like Napster were not only illegal, but also buggy and full of problems.
They would be bad quality, or they were labeled wrong. If you would go to search for a song or an artist, you think it's there, but there are 50 or 60 versions of the same song. You then go and pick one. It takes, you know, four minutes, and then you find out that the last four seconds are missing, and you don't have the full song. So it was a really terrible user experience.
Consumers wanted something better, an easier way to manage their old and new music. And record labels wanted some level of control back. And of course, they wanted payment for their artists and song catalogs. In 2003, Steve Jobs was about to solve everybody's problems by solving his own. Apple at the time was relatively small, with around a 3% share in the personal computer market. But iPods were becoming popular. Jobs's idea was to create a complete system that would include the iPod, music management software, and a store to buy music. But first, he needed access to music. And for that, Jobs needed the five major labels to sign on.
He proposed that songs could be purchased individually.
An executive from Warner had suggested the 99 cent pricing. Jobs ended up working with the Warner executives to come up with the initial pitch. And then when they agreed and signed on and had this 99 cent price, that was then leveraged to go through the rest of the big five music companies and get them to agree one at a time.
99 cents per song became key. Jobs explained to the labels how just below $1 was an emotional threshold for people for an impulse purchase. People were more likely to click and buy a song for 99 cents than to buy the same song for $1. And for anyone else who thought 99 cents was too much, he put this to them.
How many of you had a Starbucks latte this morning? And that's $3, which we can now all remember when that was $3. To which he said, you know, that's three songs. How many lattes got sold across the US this morning? A lot. 99 cents is pretty affordable. And so this was really paramount in the way that they pitched and talk about how easy it was to just grab a song if you want it when it's just 99 cents.
The plan worked. During opening week, the iTunes Store sold 1 million songs. And this was at a time when only 2% of computer users and 5% of laptop users owned Apple computers. Once iTunes was made available to people using PCs,
PC users added another million songs within just three days. So people were really hungry and excited to get that music and they started picking it up at a very, very quick escalating pace. And it changed our whole experience and the way that we access really all media these days.
Melina Palmer is the author of The Truth About Pricing. She's also the host of the Brainy Business podcast.
The 99 cents pricing strategy is one of the oldest tricks in the book. It's widely presumed to communicate a bargain, a good deal, because wow, it's just pocket change. It's not even a whole dollar. For the iTunes Store, this approach to pricing helped revolutionize the entire music industry. But falling just below a round number influences more than discount goods. Take diamonds. The price of a diamond is determined by several qualities: its cut, clarity, and color. But there's another factor that significantly influences the value: its carat number, or weight. And slight differences in the carat number can have an outsized influence on the price.
There was discreet price bands that go up, sometimes in tenths of a carat, sometimes in half a carat, sometimes in a whole carat. And that means that within each band, there is a particular baseline price.
This is Joshua.
I'm Joshua Friedman. I'm senior analyst at The Rapper Book Group. We are a network of companies around the world providing services to the diamond industry.
The Rappaport Group publishes a standard price list used by dealers to set diamond prices. For example, there's a price for a diamond in the 1 to 1.49 carat range, and another price for a diamond in the 1.5 to 1.9 range, and so on.
The way I have described it in the past is as a side view of a badly-built flight of stairs, rather than a ramp. So as you go up with size, the price increases, but it doesn't increase consistently as you go up, because consumer desire to get the round 1 carat or the round 3 carat. And just below those round numbers, that's the point where the consumer can get a good deal.
Here's the thing: a 2.99 carat diamond is almost identical in weight to a 3 carat diamond. And to the naked eye, it sure looks the same. But the price goes up substantially when you hit that 3 carat mark. And there are cultural and psychological reasons for this.
You don't want to have to explain why it's actually a 2.99 carat. You just want to be able to say you've got a 3 carat diamond. This is mostly an issue in engagement rings, where there are these social expectations going back to the 1930s.
We can thank De Beers and a successful ad campaign for inventing the idea of the diamond engagement ring. That and the suggestion that the engagement ring should cost at least a month's salary. Today, the suggestion is even higher. But if your concern is value, it's the oversized diamonds. That is diamonds at the end of their price brackets, a 0.9 or 1.9 or 2.9 carat diamond, that are the better deal. And not just for your bank account.
Often, these diamonds will be better quality when you're manufacturing a diamond. When I say manufacturing, I mean cutting or polishing a diamond from the rough. You often have to compromise on something, either on size or on quality. So if you go for that extra size and try and hit the round number, you might have to compromise a bit on the quality. So it will often turn out that those diamonds that are just shy of a round number will actually be the best quality with the best proportions and the best cut quality.
So the next time you're in the market for a diamond, consider buying one just under the carat value you're looking for. And then maybe just round up when you're showing off. Alternatively, if you're like me, you'll know you've found exactly the right person to spend the rest of your life with when they propose and tell you the diamond ring they picked was just barely below a round number threshold to outsmart the system. Yes, that's actually how my husband picked my ring, and we've been happily married for nearly 20 years.
So far, we've been focused on how our heightened attention to leftmost digits in a number and our underattention to the ones farther right affects our perception of value and pricing. But our tendency to focus on numbers on the left also shows up when life is on the line.
My name is Babu Jena. I'm an economist, physician, and professor at Harvard Medical School. Medicine is interesting in that a lot of the decisions that are made ultimately rely either explicitly or implicitly on numbers. So one measurement that we look at for kidney function is something called the creatinine. If you have a creatinine of 1.9, that's considered to be elevated, not normal. And if you have a creatinine of 2.0, that's also considered to be elevated. But you might imagine that when a doctor sees a creatinine of 1.9, they may be less concerned about it than if it's 2.0, even though those numbers are pretty much the same
and they reflect similar amounts of kidney dysfunction. But when they see the 2.0, they might react a little bit more aggressively and say, all right, well, I need to perform a kidney ultrasound. I need to work up this problem a little bit more.
Doctors have to make a lot of quick decisions with the data in front of them, including data about a patient's age.
So imagine you're a doctor. You're working in the emergency department. And the nurse comes to you and says, there's a new patient who just arrived who has chest pain. There's a lot of things that would be going through your head as to what could be causing chest pain. But a key thing to think about is, is this person having a heart attack? So you go into that room. You ask them some questions about their chest pain. When did it start? How does it feel? What makes it better or worse? And then you start to formulate an idea of whether or not this person could be having a heart attack. And then you look for things like, does this person have high blood pressure, diabetes, high cholesterol? Those are risk factors for a heart attack. And the other thing you might consider is, what's this person's age? Is this person 40 years old, 50 years old, 70 years old?
The older people are, the more likely they are to have heart disease. And heart disease, meaning blockages or narrowing of the heart vessels, is ultimately what will cause someone to have a heart attack. So those are the things you're thinking about. If you see someone who's 79 years old, and you see someone who's 80 years old, you might think of these two people as being more different than they actually are. They're really only separated in age by at most a year. But the person who's 79 might feel like they're in their quote-unquote 70s to your mind, and the person who's 80 years old might feel like they're in their quote-unquote 80s. And the older these people are, the more likely you are to think, alright, this person might have a heart attack.
And moreover, once you diagnose a heart attack, what do you do about it? Do you do a surgery? Do you just watch them? The older someone is, the less likely you are to want to crack open their chest and do a cardiac bypass surgery. So if someone is in their quote-unquote 80s, you might not want to do that, because you might feel more comfortable doing that if they're in their quote-unquote 70s. And if you look at studies, that's what we see.
Babu and his colleagues looked at Medicare beneficiaries, people who are typically above the age of 65 in the United States, and who are coming into the hospital for a heart attack. They found that patients aged 79 in 50 weeks were 20% more likely to receive cardiac bypass surgery than a heart attack patient aged 80 and two weeks. That's a huge difference for two patients who are just four weeks apart in age. And cardiac bypass surgery is a major procedure, not something doctors suggest lightly. But here, Babu and team found that doctors were more likely to suggest the surgery for someone in their very late 70s than someone in their very early 80s, even if those people were born weeks apart. Why?
Well, if you're in your 80s, you're an octogenarian, too old to make it through that tough surgery, right? But if you're still in your 70s, well, that feels young enough to take it. Of course, though, that's a bias, focusing too much on the leftmost digit of a person's age being a seven or an eight instead of appreciating that a 79-year-old who turns 80 next week is not meaningfully younger than someone who turned 80 last week.
One takeaway from this is that if you change the way the information is presented or you make doctors aware of the issue, it could potentially save lives.
Babu Jena is an economist, physician, and professor at Harvard Medical School. He was also the host of the podcast Freakonomics MD and is the co-author of the book Random Acts of Medicine. You can find a link to his book and Melina Palmer's book, The Truth About Pricing, in the show notes and at schwab.com/podcast.
What we see in pricing and in doctors' offices and emergency rooms is that we tend to overweight the importance of the leftmost digit of a number. It's a phenomenon called left-digit bias. My next guest, Devon Pope, is a professor of behavioral science and economics at the Booth School of Business at the University of Chicago. The research I invited him to discuss today on left-digit bias happens to primarily feature cars, what we pay for them and what we pay to ride in them.
Hi Devon, welcome back to the show.
Thank you, it's good to be here.
I was hoping we could start with a definition. Could you just tell us what left digit bias is?
Yes, so when, when people are looking at a number, especially if the number is large, it can be very hard to process. We see lots of numbers in our lives. And so, the human mind, as you've discussed on your show many times, likes to be efficient. And sometimes we'll use simple rules of thumb or heuristics to make it easy. And one of the things that we do when we see numbers is we focus mostly on the leftmost digit. It tends to be the part of the number that's most informative, right? If it's a $2 and something item versus an $8 and something item, it's good to know that left digit. And so, we tend to do that, but that can, as we'll discuss, create some interesting issues.
Could you tell us a little bit about the project you worked on with Lyft to implement 99 cent pricing and what you learned from that?
Yeah, so it started when my co-authors and I recognized that Uber and Lyft were offering a lot of people a ride that was just over a dollar value. For example, I literally got a ride offer from Uber for $14.02 one day. And I had been studying left digit bias for other projects and things, and it seemed like a very strange offer to make. They should be offering me a ride for either $13.99, because I'd be way more likely to accept that, or it might as well offer $14.99 and capture some more value, and I probably won't think of $14.99 and $14.02 as very different. And so, I sent a note and buddied up with John List, who's an economist here at Chicago, who at the time was the chief economist at Lyft.
And he's a former Choiceology guest too, I should say.
Nice, a good choice. And we decided, along with Greg's son and Ian Muir, to go ahead and write a paper about this.
Yeah, tell me about the paper, I love this paper.
So what we did next is we started with historical data at Lyft. So we grabbed 600 million observations from 2019.
Just a little bit of data.
Yeah, and looked at the prices that were offered to consumers for those potential rides.
And we verified that they weren't doing any sort of left-digit bias pricing. They were just as likely to offer that $14.02 ride as a $13.98 ride. And then we were really interested in knowing whether people were more likely to accept rides that were just under the dollar value versus just over the dollar value. And we find super strong evidence of this. So you're about one to two percentage points less likely to accept the ride as soon as you hop over that next dollar value, suggesting that Lyft was leaving a lot of money on the table by pricing a lot of rides at just over dollar values.
Yeah, I love that finding. And I think what I remember is that you estimated the magnitude of the effect is something like Lyft was losing $160 million a year by ignoring this in their pricing. Am I remembering that right?
That's right.
Does it surprise you that that was such a big amount of money they were leaving on the table?
It surprised them. And it was certainly of interest to them and Uber and every other rideshare company that has now changed their pricing because of this paper. It's, I think one thing that makes it surprising is that one could argue that Lyft and Uber and other rideshare apps are some of the most sophisticated pricing algorithms that we have on the planet, right? They're incredibly dynamic. They adjust based on supply and demand factors. There's different prices across different geographies at different times of day. So it's an incredibly sophisticated pricing system. And yet they were ignoring a very simple psychological thing that's been around actually a long time.
I know that after you did that analysis of the historical data, you did an experiment. Could you talk a little bit about that experiment and why that was the next follow-up step after you'd figured this out?
Yeah, so John more than me, but went to the Lyft team and showed them our analysis using historical data.
And argued that they were leaving a lot of money on the table by not adopting a 99 cent pricing type strategy, or something at least that approximates that. And they weren't totally convinced, to be honest. They were super interested. I mean, it was, we were claiming that, you know, this was the most profitable potential change in their app that they had ever done. I mean, it was a really big deal, but they were worried. The thing they were most worried about is that it's possible if all of the prices that you see on Lyft or Uber end in 99 cents, the app can start to maybe feel gimmicky or cheap. They were worried about kind of some of the long run reputation costs that might come from implementing this.
And so for that reason, along with other things, they wanted to run an experiment. And so within a couple of months, they were able to roll out a large experiment with all of the platform's 21 million riders on Lyft and 50% were put in a control group and others were put in these different treatment groups. And part of the experiment was to see if someone kept seeing 99 cent pricing over and over, would they maybe stop showing up to the app because they felt it was gimmicky or something like that? And the results of that experiment were that when someone saw a 99 cent price, they were more likely to accept the ride, just like we had shown. But they were also actually more likely to show up to the app again sooner.
They remembered their experience as being a pretty cheap, good option. And so we were actually underestimating how much money they were leaving on the table because they could not only get more rides accepted, but also people would be more loyal to the app, at least in the medium term. They are still testing really long-term holdout samples because maybe after three years, you eventually get sick of those 99 cent numbers. But that's why they ran the experiment before actually implementing.
That's great, and I love experiments. I'm curious if you could talk a little bit about places we see this bias, or you think this bias matters besides the market for diamond rings and getting into an Uber or Lyft. Where else does left digit bias matter?
The most obvious application is in pricing, right? So you see this anytime you walk into a grocery store or you buy gasoline. And now you see it on Uber and Lyft and other places as well. But in addition to prices, I think this is something that is super universal. Anytime you're looking at a number, you can fall prey to left digit bias. So one paper that Justin Sidner and Nico LaSaitara and I wrote together was about how people process the odometer values on used cars. Again, it's a big number. You're shopping for a car. There's lots to pay attention to. And we find that, for example, cars that are just under around value
that have, say, 69,000 miles on them sell for a lot more than cars that have 70,000 miles. And you see that at every 10,000 mile mark. So that would be another example. But I think it goes beyond just odometer values. You would probably see it on a car that gets 30 miles per gallon probably sells better than a car that gets 29 miles per gallon, because that just feels like a lot more miles per gallon to people.
I love that. If you were to sort of step back from the specific examples we've been talking about, what advice would you offer to people so they could avoid making bad decisions as a result of left digit bias?
That's a really good question. So the truth is, is I probably actually wouldn't worry about it too much. So take the Lyft example. We find that Lyft is able to make an extra 25 cents on average per ride that they do. So if they're doing a billion rides a year or something, that's where you can get these numbers of a couple hundred million dollars of profit extra per year. Now, for an individual, though, it's only 25 cents, right? 25 cents per ride. And we have to pay attention to a lot of numbers in our lives. And it's not obvious to me that paying attention to all of those numbers actually is the right thing to do. I think using shortcuts makes sense. And in fact, I'll go even further. If you're really interested and want to go all the way to section six
of our paper that we wrote, you'll read about what's called consumer welfare. So clearly, this idea is something that made Lyft a lot of money. Did it hurt consumers? And the answer is, is we don't know. It's not clear whether consumers were helped or hurt. And the intuition is, is that perhaps they actually were not accepting Lyft rides enough because of their left digit bias. And we were actually helping them get more in line with an optimal package of consumption that they were doing in their lives. And so it's, it's a little bit unintuitive, I think. But for the individual consumer, if everyone is kind of pricing at 99 cents,
it's actually smart and okay to start to ignore anything except the left digit.
I love that as an economist, you're thinking about the world as in an equilibrium. But since you showed that the most sophisticated pricing algorithm in the world wasn't incorporating left digit bias, my guess is that even with this amazing paper you've written, which now has Lyft and Uber in line, there's still lots of places, mom and pop shops where this isn't being incorporated. And we are encountering pricing that doesn't factor in left digit bias.
Yes.
Still, I take your point that we shouldn't sweat the small stuff, right? Who cares if you're leaving 25 cents on the table when you take a taxi ride or buy toilet paper? But I will say I messed up when buying a house with left digit bias in a negotiation, which is to this day, one of the most humiliating things in my past. We were outbid by someone who knew we were going to pick a round number and picked a little above a round number. And I'm like, God, I played that game in game theory. How did I mess that up? But I did, right? So I like the house I got, confirmation bias and all that. So I do think there are places where we make big decisions and left digit bias matters.
I think you're exactly right. So the reason why this is important for Lyft is they're doing so much volume and it's a little tiny bit times a really big number. So you're right, though, as a consumer, you might want to be a little careful when you're making a really big purchase that involves a number, a diamond, or a house, or something like that, could be in that camp. And yeah, you might be able to save a fair bit by being a little careful in those cases.
Or get the house in the first place.
Yes, exactly.
Yeah. Okay. Devin, this is so helpful. I'm curious if you do anything differently or think about anything differently after doing all this research on left-digit bias in your own life? Has it led you to shop differently? Or are you not sweating the small stuff?
Yeah, I try not to sweat it too much. I think it makes me notice. Or let me share one thing that is interesting. So here is what Uber is doing right now with their pricing. I just pulled up my activity, and let me just read you the prices for my last, say, 10 rides. So I took a ride for $28.95, $31.93, $44.96, $31.95, $5.93, $31.96, et cetera. So what they're doing is they're not pricing at the 99 cent mark. They're pricing it somewhere between 90 cents and 98 cents, is what it appears that they're doing. I've never talked
to Uber or anything, but it's very clear from just my activity level alone what they're doing. My guess is there's basically no consumer out there that has ever noticed this. Now, I notice it every time I take an Uber ride, because every single time I'm looking at those digits, and it's somewhere in the 90-something range, and I'm like, they're trying to capture a little of extra value from each customer, and then I don't worry about it. So, yeah, so I walk around noticing left-digit bias all over the place, and then I just try to convince myself to focus on the things that maybe matter the most.
Devin, I so appreciate you taking the time to talk to me today. Thank you for doing this. It was really interesting.
Thank you, Katie. It was fun to be here.
Devin Pope is the Stephen G. Rothmeier Professor of Behavioral Science and Economics at the Booth School of Business at the University of Chicago. You can find links to his research on left-digit bias in the show notes and at schwab.com/podcast.
Any major purchase can be fraught with emotional and psychological pitfalls. For tips on navigating the especially tricky decision-making process of car buying, check out the recent Financial Decoder episode, "Are You Making One of These Car Buying Mistakes?" You can find it at schwab.com/financialdecoder, or just search for Financial Decoder in your podcast app.
Left-digit bias is a bias to look out for when you're making a big purchase. If you're buying an engagement ring, a used car, a house, or making another large outlay, then what we've shared today could help you get a real deal. You can get a bargain on a ring if you're willing to buy that 1.99-karat diamond instead of insisting on two carats. You can get a bargain on a car if you opt for 80,000 miles on the odometer instead of trying to stay below that round number. And if you're in a bidding war for a house and get just one final bid, for goodness sakes, bid just above a round number that other bidders are likely to select. You might also think about this bias if you find yourself getting important medical advice
and have just celebrated a round number birthday. Consider asking your doctor if they think someone a few years younger or older might be advised to pursue a different course of treatment. And if the answer is yes, keep that option on the table. Or maybe you've taken an important test, say, the SAT or the GRE, and gotten a score just below a round number, say, a 1390. It might be useful to retake the test and aim to bump yourself above 1,400 so you can benefit from left-digit bias. But when you're making small decisions, like whether to buy a new song or hop in an Uber or Lyft, Devin's advice is spot on. Don't sweat it.
You've been listening to Choiceology, an original podcast from Charles Schwab. If you've enjoyed the show, we'd be really grateful if you'd leave us a review on Apple Podcasts, a comment or a rating on Spotify or YouTube, or feedback wherever you listen. You can also follow us for free in your favorite podcasting app. And if you want more of the kinds of insights we bring you on Choiceology about how to improve your decisions, you can order my book, How to Change, or sign up for my monthly newsletter, Milkman Delivers, on Substack. Next time, I'll speak with Berkeley Haas professor Ellen Evers about a clever life hack designed to help you move on from difficult or painful events while getting more joy from positive surprises. I'm Dr. Katie Milkman. Talk to you soon.
For important disclosures, see the show notes or visit schwab.com/podcast.
Page updated November 3, 2026 · transcript and summary are machine-generated