- 8 minutes ago
For generations, posted prices made shopping feel predictable: Shoppers generally saw the same price for the same product and could compare it across stores. Online, prices for rides, tickets, and groceries can shift while you shop.
Category
🎈
FunTranscript
00:00I gathered six colleagues to pull up a ticket on StubHub for a New York Knicks game.
00:04The exact same seat at the exact same time.
00:08And we got vastly different prices.
00:12On my laptop and my phone, I see $1,420.
00:17The total price on my phone was $1,491.
00:21I'm seeing on my laptop $1,385.
00:26Look at this now.
00:27But on my phone, I got $1,538.
00:31That's crazy.
00:33That is so weird.
00:35There was a $153 gap between my price and Nikita's for this seat at a New York Knicks game on
00:42StubHub.
00:43It was suspicious, so I took a closer look at our receipts.
00:48The base price of both tickets was exactly the same.
00:51But there was an unexplainable difference in the fees, which are the part of the price StubHub sets.
00:57Mine feels like it's exponentially worse than anyone else's.
01:03Yeah, I'm kind of like, what did I do wrong?
01:06This didn't look like the kind of dynamic pricing we're used to seeing with tickets.
01:11When a price suddenly jumps, it's easy to blame dynamic pricing.
01:16Prices that shift with demand or availability.
01:19You're doing dynamic pricing on all of us.
01:23Dynamic pricing.
01:24My God, that pisses me off.
01:27But there's another kind of pricing drawing scrutiny.
01:30Critics call it surveillance pricing, where companies can charge you the highest price they think you're willing to pay.
01:37Which at that point I was like, does the list price ever actually exist?
01:42Your frustration is really valid.
01:45Something has shifted.
01:47Companies have gotten a lot better at overcharging you.
01:51After running more experiments and digging through patents, I'll show you how companies create prices just for you.
01:59And if this could mean you're paying more.
02:07Okay, so it's October 29th and I'm going to check this round trip plane ticket from New York to LA
02:14in November.
02:15So it's $976.80 per person.
02:20And now it's October 30th.
02:23And that same flight is $891.80.
02:30Before this became what buying a flight looked like, everyday shopping felt a lot more straightforward.
02:37You looked at the price tag.
02:38But before price tags came along, shoppers often had to haggle.
02:43In the US context, it was actually the Quakers who started to feel like haggling was unfair.
02:49And so they stopped using haggling and instituted fixed prices.
02:53In the mid-1800s, American department stores brought that principle to mass retail.
02:59For, you know, nearly 150 years, American consumers have been able to rely on a posted price.
03:06Posted prices allow for transparency.
03:09You can walk across the street to another department store and compare prices.
03:13Lindsay Owens studies how companies set prices.
03:16The title of her book, Gouged, offers a hint about what she concluded.
03:22With the posted price, shoppers generally saw the same price for the same product.
03:27Airlines and live events worked a little differently.
03:30They sold seats and tiers.
03:32The trade-off was pretty clear.
03:35Pay more, get a better seat.
03:37By the late 1980s, airlines pioneered an early form of dynamic pricing.
03:44Using seat availability and expected demand to determine which fares remained available.
03:50Today, versions of that same logic show up far beyond airlines.
03:56Take a ticket to Disneyland, for example.
03:59I'm checking for a one-day adult ticket to Disneyland for October 30th.
04:05It costs $174.
04:09And now, I'm going to check that same ticket on Christmas Day.
04:13And let's see what we get.
04:16Okay, it's $224.
04:19That's date-based pricing.
04:21So maybe it doesn't feel great knowing that you have to pay more for essentially the same experience.
04:27But it's pretty transparent.
04:31Uber took dynamic pricing a step further by popularizing surge pricing.
04:36The price can move in real-time as rider demand and driver availability change.
04:42I don't have a problem with surge pricing.
04:44To me, that's something that's pretty clearly defined.
04:46Even if it costs $100, sometimes I might be willing to pay $100 for that ride.
04:50If it means I can truly get one.
04:52Concert ticket prices can move quickly, too.
04:55On primary and resale sites, artists and sellers can raise prices as demand grows.
05:02In retrospect, it was really a tipping point for what has come next.
05:09Which is dynamic pricing seeping into so many areas of our lives.
05:15Where more transparent, fixed pricing used to be the norm.
05:20By 2012, dynamic pricing was no longer confined to flights, hotels, and event tickets.
05:27Companies were using algorithms to change prices of all sorts of ordinary products online, like video games, refrigerators, and even
05:39toilet paper.
05:41In 2012, a computer scientist told the Wall Street Journal,
05:45In the age of the internet, fixed prices are a thing of the past.
05:50I wanted to run a series of experiments to really figure out how far dynamic pricing has gone.
05:57Before I did that, I hopped on a call with my colleague Juliana, who conducted her own experiment a few
06:04months ago.
06:05All right.
06:07It is recording.
06:10Yay!
06:11I really like Old Navy.
06:14That's why I was doing this.
06:16The shirt is actually from Old Navy.
06:17First, she tracked a cart full of basics.
06:21Socks, leggings, a t-shirt, and jeans.
06:24Every day for two weeks on the Old Navy app.
06:28Over those weeks, her total changed five times.
06:32I think the thing that was sort of like crazy making is that the prices would not move in concert
06:39or like in tandem.
06:41For instance, Juliana found t-shirt prices would go up, but jean prices would go down.
06:46In another example, jean prices would go up, but legging prices would go down.
06:52Then, she took that same cart into two Old Navy stores.
06:56Some of them are almost twice as much in the store.
06:59Like, the store really did default to the list price, the price that was actually on the price tag.
07:07Granted, the Old Navy Juliana went to says it will match the online prices.
07:12Old Navy's price tags even tell you to download the app, but…
07:17I noticed a lot of older folks in there, too, and less likely to be tech-savvy.
07:23And to me, I was like, well, technically this information would be available to them if they knew to seek
07:32it out.
07:33But who's going to tell them to seek it out?
07:35Okay, so I added similar items on my digital cart as Juliana's, starting with these ankle jeans I found in
07:45the store.
07:47And, you know, taking a look at the price here, it's 50% off.
07:52And it's $22.
07:56It's almost a $5 difference online versus in store.
08:01Old Navy didn't respond to either of our requests for comment.
08:07Caroline fairs are pretty front-facing of fluctuating prices, but brick-and-mortar retail stores aren't.
08:14Sticker above my head, we're also doing fluctuating prices.
08:19It's something that you kind of notice by yourself.
08:24Which at that point I was like, does the list price ever actually exist if there's always some sort of
08:30price fluctuation on it?
08:31That's when I realized dynamic pricing isn't the whole story.
08:37This isn't just about supply and demand moving the price.
08:41It points to another question.
08:43Who gets what price?
08:46The idea of giving different people different prices isn't new.
08:51Economists call it price discrimination.
08:54The basic idea is companies will set different prices for different groups.
08:59That's price discrimination.
09:00One form of price discrimination is a student discount, right?
09:04Where students are offered a better price, you know, at a restaurant or at a movie theater.
09:09Companies use customer groups as clues for what people might pay.
09:13And sometimes we do the sorting ourselves.
09:16You can see this at play during happy hour.
09:19Same drink, same bar, just cheaper for the group of customers willing to come when demand is low.
09:24In these examples, you can see what group you're a part of.
09:29Plus, they're advertised as discounts.
09:31And who doesn't love a good deal?
09:33But the internet and shopping online changes that.
09:37The group you're in can be much harder to see.
09:40I asked our senior retail reporter about these invisible groups when I reported on Uber earlier this year.
09:47They have all this data on a bunch of people like you or people who have certain similar habits.
09:54And so, you know, if we send that group of people a specific price, you're likely to respond to that
10:01in a certain way.
10:02Essentially, Alex is saying companies like Uber can sort you into these invisible pricing groups.
10:08He also mentioned something interesting about how they do it.
10:12They have all this data.
10:14I got a glimpse of this when I asked McDonald's to send all the data it had collected about me.
10:21I got...
10:25431 pages back.
10:27McDonald's sorted me into a customer group called CV1 and scored me zero on its measure of how likely I
10:35am to stop visiting.
10:37I have to admit, that one called me out.
10:40The file doesn't show whether any of those predictions changes the prices I paid,
10:45but it does illustrate how companies can sort customers into groups you'd never know about.
10:51The internet made pricing less transparent.
10:56Companies having access to consumer data has taken it to the next level.
11:00And part of what fueled that growth is sitting right in our pockets.
11:05Our phones can generate a detailed trail of how we browse and shop.
11:11You have mobile devices in the 2010s, which really increased the level of location granularity.
11:18That's Stephanie Nguyen, former chief technologist for the Federal Trade Commission.
11:23She explained that personal data collection goes beyond tracking your tops and building a profile about you.
11:30The rise of algorithms and AI made it easier to connect the dots and make predictions about us.
11:37You also have now the ability for companies to collect that data in near real time.
11:44Now, of course, with AI, increasingly companies can also track sort of conversational data in addition to behavioral data.
11:55They can collect and store your confessions, what you tell your chatbot.
12:00All right, so I'm going to try chatting with Amazon's Alexa for Shopping,
12:07which is Amazon's AI shopping assistant, and see what answer we get.
12:14I can't deny that these are pretty accurate.
12:18Amazon describes Alexa for Shopping as helping customers make shopping decisions and understand product prices.
12:25Just from purchase history alone, it paints a pretty, as it says, vivid picture of what I like to buy
12:35and what I need.
12:37The thing is, companies tell us very little about what happens between the data they collect and the price we
12:44see.
12:44But their patents can show us what they may be trying to build.
12:48A researcher out of MIT tracked more than 3,000 U.S. patent applications filed by 23 major companies from
12:572015 to 2025.
12:59This was the most common type of patent — consumer data collection and market modeling.
13:06The patents themselves reveal a lot, like this one assigned to DoorDash in 2021.
13:12It describes tracking a shopper's internet traffic to figure out what they know about a product's price.
13:20A 2018 Amazon patent describes using location data to predict where people will gather, then picking the timing and discount
13:29for a deal.
13:30Lindsay's team specifically looked at Walmart's patents.
13:34She found one from 2021 that describes changing the online prices you see, based partly on who you are, how
13:41you've shopped, and how close you are to a store.
13:45Now, the patents the MIT researcher and Lindsay's team cite don't prove that these technologies are being used.
13:52DoorDash told me that the patent referenced was never used and has been withdrawn and that the company does not
13:59use customers' emotional state, stress level, or biometric data to charge customers more.
14:06Amazon also shared with me that the patent is unrelated to its online store and it does not use personal
14:13information to charge customers higher prices.
14:17Meanwhile, Walmart said, the patent has been issued, but we will not use it or renew it.
14:24Together, their patent reviews show where technology may be headed.
14:28Pricing built around shopper data.
14:31Data is the bedrock foundation for commercial surveillance and for AI models.
14:39And consent at this point effectively looks like a tiny little pop-up window with legalese.
14:46The ability for someone to know that any of this is going on is effectively opaque.
14:52Your data can make pricing much more precise.
14:55Rather than stopping at what a group might pay, experts worry companies will try to estimate what you might pay.
15:06That's personalized pricing.
15:08And when personal data helps determine the price, critics call it surveillance pricing.
15:13Surveillance pricing is not the same as dynamic pricing.
15:16This is pricing that is based on the company's assessment of your willingness to pay.
15:22These days, talk of surveillance and personalized pricing is hard to avoid.
15:28It's in headlines and people online are putting it to the test.
15:32Go to Target and buy spam.
15:34Okay.
15:35How much is it?
15:36$4.99.
15:37My price is $4.59.
15:40That's bulls**t, dude.
15:42The FTC is now examining this too.
15:46It's ordered nine companies from airlines and ride hailing apps to retailers to explain how they set prices and whether
15:55personal data plays a role.
15:57Instead of student and adult, there's a million different bins and you're getting very, at that point, you're getting very
16:04close to personalized pricing where they're giving basically individual prices, right?
16:08That brings us back to our StubHub experiment from the beginning of the video.
16:14Nikita and I were shown totals $153 apart for the same Nix seat, even though the base price of the
16:21ticket itself was exactly the same.
16:24So we tried again, this time with a different event.
16:28Olivia Rodrigo at Barclays Center on February 23rd.
16:32We ended up having very similar results.
16:36My phone, the ticket was also $252, but the fees were $68 for a total price of $321.
16:46Similarly to everyone, the base ticket is $252, but the fees on my computer were $66.
16:53And $51 on my laptop.
16:55And on my phone, the fees were $78, which brought my total to $331.
17:01For the second time, Nikita's phone showed the highest fees in our group.
17:07My laptop showed the lowest.
17:10This time, the gap was $27.
17:13All of it still coming from the fee.
17:16So it almost feels like you just have to hope that you're lucky enough to get a fair ticket price.
17:21And we wouldn't have known if we didn't all do it at the same time.
17:26To help interpret our results, I spoke to Mark Tremblay, an economics professor who studies digital platforms and fee transparency.
17:35My initial reaction to this is like, oh, this could be the mechanism where they implement personalization, right?
17:44So now because everything's all inclusive, you can also kind of maybe hide where price variations are coming from.
17:54When I asked StubHub about the results, they said these are randomized, anonymous tests to find which discounts best balance
18:02demand and access.
18:04And that these tests don't use data like demographics, location, or purchase history.
18:11StubHub shared a similar response to my colleague Emily after she documented the same kind of fee variation in June.
18:19I invited her to join our experiment.
18:22With me, they're just like it's experiments.
18:24And they do point out that like other platforms do this as well.
18:28I mean, at least when I looked at it, like StubHub had more like variation, wider variation.
18:34But it's like, it just kind of feels like sucky and unfair that you don't get to know.
18:40This is where the economic theory kind of butts heads a bit with some consumer groups.
18:46So let's say they were personalized pricing or even adding randomness to the prices.
18:51You could ban this and then everyone would get the same price.
18:55And that price is going to be in the middle between the high and the low.
18:58So the winner is the person that had the higher price and the loser is the person that had the
19:02lower price, which now has to pay a higher price.
19:04Can a pricing test or experiment or, you know, still create fairness or transparency concerns, even if it's random or
19:15if it's, you know, even if it's not personalized?
19:20I mean, I could talk for hours on this.
19:22But you're right. I mean, I don't think they're disclosing this to the consumers that, hey, you're a part of
19:26a pricing experiment.
19:27Our experiment was limited to a few data points, but Groundwork Collaborative, Consumer Reports and More Perfect Union asked a
19:36similar question on a much larger scale to 437 Instacart shoppers across four cities.
19:44Lindy's team was at the forefront of the study.
19:47For almost 75% of the items in the cart, different shoppers were being charged different amounts.
19:53Instacart called them short term randomized tests, not prices based on personal, demographic or behavioral data.
20:02It later shut down all item price testing on the platform.
20:07In the e-commerce setting, we're all just guinea pigs in this big pricing lab.
20:12It's the best excuse, right? Like, if you don't want to say personalization, you say randomization.
20:18If they do give me a call, I'll stop doing a lot of my research of trying to detect personalization
20:23if they're just going to start admitting it.
20:25One place companies openly acknowledge using personal or behavioral data is for personalized discounts and promotions.
20:34We know this because Uber says so.
20:37Instacart, Target and Starbucks all disclose versions of this practice.
20:42These deals aren't always as great as they seem.
20:45In an Uber experiment we did, comparing prices for the same ride at the exact same time,
20:51we found that discounts didn't always lead to lower prices.
20:56Okay, my Uber X is $88.05.
20:58My Uber X is $89.29.
21:02Yeah, $88.10.
21:03Oh, they're out.
21:04Mine's $88.04.
21:05But I got a promotion.
21:06Mine's $87.
21:07Mine's $97.
21:08$90?
21:09Who has a promotion? Everyone has a promotion?
21:12Okay, that was a lot.
21:14My friends like to talk over each other.
21:17Essentially, Hamza's reference price was $90.17.
21:22With his promotion, his price went down to $89.29.
21:28That discounted price was still higher than what others in the group, without a discount, were quoted for the exact
21:35same ride.
21:36So Uber says they may use personal data to dole out discounts, and that discounted price isn't guaranteed to be
21:44cheaper.
21:45Consumer Reports found a similar pattern with Uber and Lyft, with 174 participants.
21:51In 12.4% of advertised discounts, the final fare was actually higher than the median undiscounted quote for the
22:01same route.
22:02Both companies disputed the findings and denied using fictitious discounts or surveillance pricing.
22:07When we asked Uber about this, it said that each participant had slightly different pickup routes and that different pickup
22:15estimates can lead to small variations in price.
22:19If you and I are shopping in an app and both of us are looking for a ride and you're
22:26offered a discount and I'm not, I'm charged more than you.
22:29That's personalized pricing.
22:31Whether it was done by increasing the base price or whether it was done by offering one of us a
22:37discount.
22:37Maybe the difference is random, as companies say.
22:41Maybe it's personal, as Lindsay suspects.
22:44Either way, consumers see different prices without knowing why.
22:49It felt like almost a loss of agency.
22:53As pricing becomes more and more of a black box, I realized no matter how long I gamed it, I
23:01would never win.
23:03So this is kind of the challenge with personalized pricing that I think economists like myself want to push back
23:10a little bit on is it can be used to do good things, right?
23:14You can offer discounts and kind of expand the market.
23:17That being said, I think seeing all of the pushback we've seen around this issue, I've been somewhat surprised.
23:25It seems like a lot of consumers, even if they're getting the lower price, don't necessarily feel good about that.
23:31Okay, so shopping can feel like a losing game.
23:35Is there anything you can do?
23:38One obvious solution would be to join a company's loyalty program.
23:42Sign up, get the deals.
23:45Well, critics say it's yet another way companies can get more of your data.
23:52Washington Post reporter Jeffrey Fowler shared his Starbucks rewards records with Stephanie.
23:58She found that the months Fowler bought more coffee, Starbucks sent him fewer promotions.
24:04Stephanie says it suggests that Starbucks used the data they had on a customer to limit their discounts.
24:11This American deal is being degraded.
24:14The benefits that we used to receive are not as good.
24:19Starbucks confirmed that it does use purchase history to choose discounts.
24:23Stephanie points out that shrinking rewards are part of a broader pattern that she calls the hike.
24:30The hike is what happens when they have all this data, they can be able to, you know, understand who
24:37you are and prices might go up.
24:39Perhaps the benefits that you get as a loyalty user goes down.
24:46As Stephanie describes in her paper, companies learn more and more about their customers through loyalty programs while using that
24:55information to restrict what members get in return.
24:58I wanted to find out how much data companies are gathering through loyalty programs, so I checked for myself.
25:06That brings us back to the 431 pages McDonald sent about me.
25:12There were hundreds of pages logging my interactions on the app over my year of being a member.
25:18Near the end came its reading on my eating habits.
25:23It ranked 20 piece McNuggets as my top product, calling my visits dinner on the go and late night.
25:31Which I can't really argue with.
25:34Then it predicted that I'll visit 3.56 times and spend $70.08 over the next six weeks.
25:42The file doesn't show whether any of those predictions influenced the prices I paid or the deals I got.
25:50McDonald's also shared with me that it doesn't use any personal information, such as purchase history, to set individual prices
25:58or menu prices.
26:00Given all of the commercial surveillance and targeting infrastructure that has developed over time,
26:07is that loyalty programs have become a Trojan horse for that surveillance.
26:13Some worry that even in-store prices could soon be under attack.
26:18At this Walmart in New Jersey, they've installed digital price labels.
26:23Walmart's CEO said that they price the products, not the person.
26:28The company also told me that a digital shelf label isn't a pricing strategy.
26:34Lindsay's team found a patent that describes changing a shelf price based on what a shopper scans into their cart.
26:41Scans Uno, for example, in Mayo could be a different price.
26:46Walmart shared with me that a patent isn't a plan and that they will not use or renew this specific
26:52patent.
26:53Is there something consumers can realistically do to, you know, mitigate this while it's happening?
27:01I think the burden should not be on consumers to try and navigate this complex infrastructure.
27:10It has taken hundreds of reporters and researchers and people whose job it is to actually hunt down pricing abuses
27:21to unveil what these things are.
27:23Stephanie says it's lawmakers who should intervene.
27:27And they are starting to take notice.
27:30In August, surveillance pricing was the subject of a hearing on Capitol Hill.
27:35Members of both parties expressed concern.
27:38AI surveillance pricing is the unholy trinity of everything Americans hate.
27:43Spying on people, ripping them off, and taking away jobs.
27:46Meanwhile, four states have enacted laws aimed directly at the practice.
27:51New York requires businesses to disclose when personal data is used to help set prices.
27:58Maryland now restricts personal data pricing for groceries.
28:02Connecticut's broader retail ban begins in July 2027.
28:06And New Jersey's ban on groceries starts in August 2027.
28:10But these laws still exempt certain discounts and loyalty programs.
28:15All encompassing policies are really important here.
28:18It's not just prices have many components of, you know, booking fees, base prices, all these things.
28:25It's also promotions.
28:26I think we need to return to ensuring that we actually see real prices.
28:32And that we can compare real prices in the marketplace.
28:35Without that, we have no functioning marketplace.
28:39Yeah, we can put a stop to this.
28:41We can fight back.
28:41We can restore a fair price in this country.
28:43COVID-19 delivery feel free.
29:09COVID-19 delivery
29:10You