« Don’t Know Much About Privacy . . . | Main | An Odd Connection »

Tuesday, June 07, 2005

Soup for Me at $5 but No Soup for You (Or Maybe at $10)

There is still more interesting grist from the national telephone survey by the Annenberg Public Policy Center at the University of Pennsylvania.  The report has an extensive discussion of price discrimination – offering different prices for the same product or service to different customers based on behavioral profiling.


This practice is already happening.  Supermarket discount cards are an example of price discrimination.  The report notes: “[B]eing a loyal customer doesn’t automatically mean getting the lowest prices.  Computer analyses of shopping histories might determine that a person’s allegiance to some products means he or she would buy them even without the discounts, or with smaller discounts than others might get for the same items at the same time.” 

But the future potential for discriminatory pricing is vast.  Ponder this:

Merchants consider the online environment a particularly ripe are for such “dynamic pricing” – that is, for . . . price discrimination driven by behavioral targeting.  Writing in Harvard Business Review, associates from McKinsey & Company chided online companies that they are missing out on a “big opportunity” if they are not tracking customers’ behavior and adjusting prices accordingly.  Consultants urge retailers to tread carefully, though, so as not to alienate customers.  The most public revelation of price discrimination online centered on customer anger at Amazon.com in September 2000 when it offered the same DVDs to different customers at discounts of 30%, 35% or 40% off the manufacturer’s suggested retail price.  Amazon insisted that its discounts were part of a random “price test” and not based on customer profiling.  After weeks of customer criticism, the firm offered to refund the difference to buyers who had paid the higher prices. 

Angeldemon One marketing book suggests that companies begin to treat consumers differently.  The book notes that certain customers (“angel” customers) are very profitable, but that there are other rather unprofitable customers (“demon” customers).  For example, demon customers return items frequently or call customer service a lot.  They can actually cost a company money.  Thus, the book recommends that companies find ways to slough off the demons and retain the angels.

What if, based on profiles, companies started charging higher prices based on people who appeared to be wealthier?  Or people who were determined to have a particular need for a product or service?  Or based on how often people called to complain about products or how often people returned items?  Are such practices uncouth?  Illegal? 

Well . . . not illegal.  But that’s not what most people think:

· 68% of American adults who have used the internet in the past month believe incorrectly that “a site such as Expedia or Orbitz that compares prices on different airlines must include the lowest airline prices.”

· 64% of American adults who have used the internet recently do not know it is legal for “an online store to charge different people different prices at the same time of day.”  71% don’t know it is legal for an offline store to do that.

As for whether such practices are unseemly, that’s for you to decide.  Some might say that price discrimination is fine so long as it makes economic sense.  If you’re willing to pay more than me, why not charge you more?  On the other hand, decisions about who pays what are based on personal information plugged into a profile.  The argument for consumer profiling is that it enables marketers to better target advertisements to interested consumers, thus bringing consumers more information about things that they will find of interest.  This seems quite innocuous.  But what if every little thing we buy, return, or complain about is tracked and then used as a way to treat customers differently, to charge them different rates for products and services?  Things begin to look a bit more problematic.

Posted by Daniel Solove on June 7, 2005 at 03:31 AM in Daniel Solove, Information and Technology | Permalink


TrackBack URL for this entry:

Listed below are links to weblogs that reference Soup for Me at $5 but No Soup for You (Or Maybe at $10):

» The Gifts You Can No Longer Return from Concurring Opinions
In the fun and light documentary, My Date With Drew, an average guy named Brian Herzlinger chronicles his attempt to get a date with Drew Barrymore. The documentary was made on a shoestring budget of just $1100, and Brian cut... [Read More]

Tracked on Dec 29, 2005 12:49:57 AM

» The Gifts You Can No Longer Return from Concurring Opinions
In the fun and light documentary, My Date With Drew, an average guy named Brian Herzlinger chronicles his attempt to get a date with Drew Barrymore. The documentary was made on a shoestring budget of just $1100, and Brian cut... [Read More]

Tracked on Mar 3, 2007 1:06:34 PM


Charging high-cost customers more seems like a basic freedom to contract, but only if the practice is disclosed.

I can't see a problem with selling items for x% more if you want the right to call an 800 customer service line or to return non-defective products at will - that's stating the term up front.

The practice of changing the price on the next purchase based on whether the call or returns were made seems like the same term, and so it should have been disclosed in the original contract to be proper.

Posted by: Not Carl Rove | Jun 8, 2005 9:10:39 AM

Nathan: university scholarships and the dental equivalent ought not be considered price discrimination in the traditional sense because they're lowering their price for those special cases below their cost and noneconomic factors are driving that choice. There's no deception: the students who don't get scholarships know they didn't get scholarships, and know scholarships are available to others (since the schools trumpet them). There's also no anticompetitive aspect to the whole thing, since the ordinary tuition price of the schools is presumptively a normal market rate, it's not an example of one school charging a higher rate because of its monopoly.

Posted by: Paul Gowder | Jun 8, 2005 8:25:43 AM

Paul - maybe, but there would be all sorts of practical problems. It can be very difficult to identify if a company is engaging in price discrimination or there is some real difference in cost. Also, sometimes price discrimination is good for consumers. Imagine a dentist who charges wealthy clients more so she can see poor clients for a much lower fee. Or a university offering a scholarship to a low income student.

Posted by: Nathan | Jun 8, 2005 1:43:26 AM

Christine: re your (1), I'd argue that the person paying the higher price is indeed getting soaked for certain reasonable definitions of "soaked."

A. Soaking = extortion (or things that we consider wrong as analagous to extortion). Extortion can be understood as unfairly imposing extra costs on a person with very low elasticity of demand because they'll be harmed if they don't pay up. For example: if I'm bleeding to death, and you're a doctor, it ought to be considered a form of extortion to charge me $30,000/hour when your normal rate is $300/hour. Do you disagree that this kind of "price discrimination" is immoral and soaking even though the value of the medical services to the person bleeding to death is $30,000/hour and more?

How about airline "price discrimination" against people who have a particular need to fly? What if bereavement fares meant triple the price instead of a third? Would you experience that as a "soaking?"

How about groceries, then? Lets bring it right back around to an example used in the initial post, so I can't be accused of attacking strawmen. Suppose a grocery store (the only one in the neighborhood) uses its frequent-buyer program to discover that Jane Schmoe is a frequent buyer of low-sugar products, whether they're on sale or not. Suppose, then, that store concludes that she values those products at a price at least as high as or higher than the highest price they've ever offered them at. Suppose they charge her that price, all the time, and invisibly (via failing to send her coupons that everyone else gets, for example). Finally -- and here's the predictable punch line -- suppose that Jane has diabetes and needs the low sugar foods to live.

Is Jane being soaked? I think we'd understand a soaking as going on. In fact, we might start using nasty phrases like "disability discrimination" instead of "price discrimination."

B. Soaking = deception. See my comment above re: the necessity of deceiving the price-discriminated-against person be unaware of the price discrimination, in order to prevent him from running to a competitor. This defies the very idea of a competitive market, and surely a victim of deception can be considered a soakee. Moreover, price has a communicative aspect independent of information about the prices offered to others: it makes a statement about the marginal cost of the goods to the merchant. That statement, in price-discrimination-world, is also a lie, and one that might harm the soakee in making her own conscious determinations of value (since of course we're subject to framing effects when determining value of an item for which we're negotiating), or in deciding whether to enter the market as a seller. Again, a soaking.

Posted by: Paul Gowder | Jun 7, 2005 9:20:47 PM

Two points: (1) The person paying the higher price isn't getting soaked. That person is paying the price at which he or she values the product. Just because someone is paying a lower price doesn't mean that the person paying the higher price is getting soaked. (I guess the 1936 authors of the Robinson-Patman act would disagree with me here.)(2) The angel/demon dichotomy is not a price issue, it's a cost issue. A vendor can, and should be able to, charge high transaction cost customers more than low transaction cost customers.

Posted by: Christine Hurt | Jun 7, 2005 7:23:41 PM

Nathan: doesn't that suggest that price discrimination in the formal sense should be banned, as an unfair anticompetitive advantage based in market power and "insider information" about prices?

Posted by: Paul Gowder | Jun 7, 2005 6:38:15 PM

I'm a bit late answering Jeff V.'s first point, but here goes.I might be venturing a little beyond my turf here, but doesn't this practice sort of destroy the proper workings of supply and demand? Theoretically, the supply curve is determined by the amount of people that will be willing to supply a product at a certain price. The price is then set by the intersection of the supply curve with the demand curve -- which is the amount of people that will be willing to buy a product at any given price. But if suppliers can vary their prices based on how much individual customers can pay, doesn't this basically make the supply durve irrelevant?I think you have too simplistic a view of the economics. The standard economic story with supply and demand curves assumes there is perfect competition. This requires that buyers and sellers have perfect information, and that no buyer or seller can individually influence price.

If there is perfect competition, economic theory predicts there will never be price discrimination. And, in practise, you do not see price discrimination in markets that closely approximate perfect competition (e.g. commodities market).

But Amazon does not operate in a perfectly competitive market. Amazon's market is what economics call monopolistically competitive. This means Amazon can raise prices a bit and lose some, but not all, of its business, because its competitors' products are not perfect substitutions for Amazon products. (A book from BN.com is not exactly the same as a book from Amazon.) Probably more importantly, Amazon's customers do not have perfect information about prices elsewhere.

This gives Amazon some room to price discriminate. For example, if it knows some customers always check BN.com before buying from Amazon it will want to offer those customers a competitive price, but it might raise prices a little for customers who do not shop around. Notice this could mean charging some high income people less and some low income people more. But Amazon is still constrained by the competition - they can't set prices based purely on willingness to pay. They can gouge me a bit before I check what a book costs elsewhere, but they can't take my entire consumer surplus.

I would also like to point out that the discussion so far has not distinguished between two different concepts. If Amazon charges me more because I am willing to pay more, that is price discrimination (as economists use the term). If Amazon charges me more because I am a "devil" customer that is NOT price discrimination - it just reflects the fact that Amazon's costs are higher when it deals with me.

Posted by: Nathan | Jun 7, 2005 6:22:22 PM

Dan, thanks for clarifying the issue for me.

My visceral reaction is this: "So, the rich are getting richer." E.g., if someone has a million dollars of disposable income to purchase books, he's going to get the lowest prices on books since, though the margin is low, Amazon knows it's going to make a tidy profit given the rich customer's aggregate purchases. But the poor college kid is going to pay a premium. Of course, these are all voluntary transactions for non-essential goods (well, some would say books are non-essential). If you don't want to pay, then go elsewhere.

One real problem I see is that consumers might become overly risk averse, knowing that in complaining to a corporation, they are going to be "punished." Imagine someone has a "consumer profile," much like we all have credit profiles - high scores for angels and low ones for demons.

I've paid bogus bills (though not since graduating law school) when threatened with the, "If you don't pay, we're going to put this on your credit report." So, an underhanded corporation could say: "If you return this product, we're going to report you to XYZ Database, and ruin your consumer profile."

But there's a solution. Just as various fair debt collection acts prevent abuse, various fair consumer profile act could do the same.

Posted by: Mike | Jun 7, 2005 5:34:23 PM


It's been a while since I read the book, but the point of the book was to suggest to businesses to figure out for themselves how to segment consumers based on their profitability. The "angel" and "demon" labels are just snazzy rhetoric to generalize about a spectrum of behavior.

Clearly, there are demon customers who are obnoxious in their behavior, but even behavior that is completely justified can be unprofitable. As I recall from the book, there is not much of an attempt to introduce a kind of moral judgment into the mix; the analysis goes to profitability, regardless of whether customers are nice or not.

You seem to be in favor of segmentation based on a combination of moral disapproval and unprofitability. Your comments also seem to be focused around offline stores. The Internet makes price discrimination easier. If I walk into Barnes & Noble, it is hard for them to offer me a different price. But if I go to Amazon.com, they can readily offer me a different price based on my customer profile.

Now, I consider myself a good Amazon.com customer. I buy a lot of stuff and don't return much. But suppose a new customer makes a few purchases and they happen to be damaged or defective. The new customer returns the items. A profiling system might identify the new customer as a demon. Now, over time, the customer may not return many more items. But Amazon might still have an incentive to drive the customer away, as the profile thus far signifies a demon and based on statistics, it might be that 80% of such customers with such behavior turn out to be demons in the long run. Not a good gamble, so get rid of that customer.

As I see the potential for price discrimination and other forms of segmentation based on behavioral profiles, it might very well be a cold numbers game for businesses rather than the mom-and-pop "I don't like Mike because he's such an obnoxious customer."

Posted by: Daniel Solove | Jun 7, 2005 2:57:35 PM

You're the expert, so I'm not necessarily arguing here. But I wonder is there isn't a strong correlation between rudeness and general non-profit providing customers.

Thinking back, the rude and nasty people were also the ones more frequently complaining that the merchandise was junk; they were the ones returning obviously used items, etc.

OTOH, I know that I was a demon customer at an organic food store. I purchased organic meat and milk, which has a low profit margin, but not granola, carob, or other things that have a high profit margin. I also know that many credit card companies consider people who don't carry a balance demons.

So it'd be interesting two see the two types of demons separated. Does the author in the book you mention do so?

Posted by: Mike | Jun 7, 2005 2:36:55 PM


Demon customers aren't necessarily ones that make store employees miserable. They are ones that are not as profitable, such as ones that take advantage of a store's promise to match any lower price or ones that actually use customer service lines. These people need not be irascible or annoying. They might simply be exercising their rights as consumers or taking advantage of what a merchant promises to them.

So if Donald Trump were to be amazingly rude to a store's employees, I still bet you he'd be an angel customer in the eyes of the store owners.


Posted by: Daniel Solove | Jun 7, 2005 2:22:25 PM

Another question, beyond the fraudulent concealment angle, is under what conditions can corporations effectively subdivide their customer population? Can one company successfully gather and control the information necessary to effectively price-discriminate, or would it be necessary for industries to do so collectively? And if so, would the other word for that practice be "price-fixing?")

On an ethical level, we need to honestly look at the fact that what's really involved here is deception. The only way a corporation can charge a lower price to customer B for the same product as customer A is getting is if the fact that (s)he's getting soaked is concealed from customer A.

If customer A knows (s)he's being soaked, (s)he ought to be able to turn to a competitor for a lower-cost product. If that's not the case, there's either a market failure (under traditional economics, price should be reduced to marginal cost because of the effects of competition,n non?) or outright price-fixing again.

Posted by: Paul Gowder | Jun 7, 2005 1:22:20 PM

Corporations are going to track the bottom line and try to eke out as much profit as they can. This I accept. I have a problem when the corporations start bleeding out every dollar they can from each individual. It bothers me on some level, but I am not certain that I have legal objections. Keeping "angel" customers is like keeping sheep. While I have suffered at the hands of abusive customers from my time in retail, feedback from customers on what is not working is as important as anything else. Then again, maybe there is a reason I still do the vast majority of my shopping off-line where I can smile at the clerks and ask for help if I need it.

Posted by: Joel | Jun 7, 2005 12:55:07 PM

As a stand-up customer (polite to customer service, don't complain except for good cause, etc.), it's not troubling that "every little thing we buy, return, or complain about is tracked and then used as a way to treat customers differently," since I'll likely benefit.

Generally, I'm not bothered by the privacy implications of private actors compiling detailed information about me. If a private company abuses the information (or loses it and thus causes me some harm), it can be held accountable. What would trouble me is if this information were shared with a government entity - since government actors are mostly unaccountable.

Getting rid of demon customers, which is good for the corporations bottom line, also benefits poor people.

I worked a crappy $7/hr customer service job. Working a demanding job at low pay is hard enough. Dealing with demon customers would make the job almost intolerable. Indeed, I've seen people quit after horrible treatment from customers. I've seen demon customers make people cry and get good employees in a lot of trouble. Mind you, those of us in the lowest-levels of customer service were generally poorer; so getting rid of the demons (assuming no loss of corporate revenue and thus fewer jobs) would ease the poor person's burden quite a bit.

I think that's something people who write about these subjects tend to forget that. Poor people tend to have crappy jobs, which demon customers make almost intolerable. So keeping the demons away would be a concrete benefit for a lot of people whose lives are tough enough as it is.

Posted by: Mike | Jun 7, 2005 12:14:14 PM


What you're talking about is a different phenomenon. In economic theory, the aggregate demand curve is the sum of every single individual's demand curve for a particular product. We usually call this aggregate demand curve the "demand curve."

That's all well and dandy. Different people will pay different prices for a certain product, depending on how bad they want it. However, the market is supposed to take this data and combine it with the (aggregate, again) supply curve, which represents the different prices that suppliers will supply them for. Where these two curves intersect, you are supposed to get ONE price that clears the market. When this price is achieved, people that are willing to pay it (or more) will buy the product.

This then creates a thing called "surplus" -- which represents the fact that people who would have paid more than the market price for the good actually end up paying less than they would have. You can think of this "surplus" in a non-technical sense of having created "value" for the society as a whole [there's also a surplus on the supplier side, but we'll ignore it for now].

If you get rid of this single-price mechanism by doing something like price discrimination, then the surplus goes away, because everyone will end up paying the maximum amount of money that they would be willing to spend on a product.

Price discrimination also nullifies one of the prime values of a market-based economy: forcing people to make rational choices on what products they will by given that they have a limited income. Theoretically, if price-discrimation were pervasive, people would start buying a huge number of products at much lower prices and the market would break down. In this situation, though, since the price discrimination is based on income, the effects would be somewhat more complicated...maybe such price discrimination would in effect be a shift of wealth from the rich to the poor. It doesn't seem like the type of wealth shift that has beneficial effects for the economy as a whole, however.

Posted by: Jeff V. | Jun 7, 2005 11:53:24 AM

I would think that different groups of people have different demand curves. Instead of lumping everyone together into one, you could create separate price points for each group. I was in the market for a bedroom set and went to a major furniture store in Houston. The store had four showrooms. I went to one showroom in The Woodlands, which is fairly affluent. I found one set that I liked. The next day, I went to a different showroom in an area of town that was mixed as to demographics. I could tell from the inventory that they were catering there to a less affluent clientele. The same bedroom furniture (manufactured by Thomasville) was $200 less at that store. I bought it there.

Posted by: Christine Hurt | Jun 7, 2005 10:18:41 AM

Nathan -- yes, it was a typo. I corrected it. Thanks.


Posted by: Daniel Solove | Jun 7, 2005 9:48:36 AM

"Illegal," eh? Well, for expedia and orbitz, maybe not revealing known lower prices is actually illegal. Orbitz, for example, represents its services as characterized by: "Award-winning flight search engine makes it easy to find the lowest fares on more than 455 airlines."

If they know that they're not offering "the lowest fares," might they not be defrauding consumers?

Posted by: Paul Gowder | Jun 7, 2005 8:32:30 AM

I might be venturing a little beyond my turf here, but doesn't this practice sort of destroy the proper workings of supply and demand? Theoretically, the supply curve is determined by the amount of people that will be willing to supply a product at a certain price. The price is then set by the intersection of the supply curve with the demand curve -- which is the amount of people that will be willing to buy a product at any given price. But if suppliers can vary their prices based on how much individual customers can pay, doesn't this basically make the supply durve irrelevant?

I'm sure someone on here has a better backgroud in economics and can figure out the implications of this better than I can, but it seems like this should be a recipe for breaking down the beneficial aspects of the market and should therefore be illegal.

Posted by: Jeff V. | Jun 7, 2005 7:39:11 AM

Is "illegal", in the sentence
71% don’t know it is illegal for an offline store to do that.
a typo?

Posted by: Nathan | Jun 7, 2005 1:47:58 AM

Post a comment