« Brazil facts of the day | Main | Bernanke's bubble laboratory »
Retail loyalty card programs
From some time ago, Kevin Drum reports:
I really loathe retail loyalty card programs.
These programs serve two functions. First, they are a form of price discrimination. Buyers who are willing to collect and show the cards pay lower prices while the "I can't be bothered with this ****" types pay higher prices.
Second, retail loyalty cards enforce partial collusion ex post in an oligopolistic setting. In other words, cards and frequent flyer programs "lock in" buyers to their favored firms. Once that lock-in is accomplished, all firms have weaker incentives to cut price to lure away buyers from their favorites. (The smarty-pants point is to note that firms have to give buyers a better deal upfront in anticipation of this lock-in but still if the company moves first with a non-negotiable offer it still can come out ahead and raise the P/MC ratio.)
The first function is usually welfare-improving, the second function usually is not. Overall you personally benefit from loyalty card programs if you don't mind holding the cards (you have a thick wallet) and you have a strongly favorite company/product anyway. In the latter case you are likely locked in anyway, so the strengthening of the lock-in effect doesn't so much restrict your freedom. This is tricky of course because you might miss out on preemptive price cuts from your favorite firm to keep you, since maybe they don't otherwise know how much you love their stuff. Still, I will stick with this mechanism as a plausible guess of the net effect.
You suffer from loyalty card programs if...you hate them. Not only do the programs and the smiling clerks bug you but you are the kind of person who ends up paying more. Which means you hate the programs even more. Which means...
But wait: the equilibrium seems to converge and so Kevin Drum's anger at retail loyalty card programs remains, in reality, quite low.
Posted by Tyler Cowen on May 16, 2008 at 04:11 AM in Economics | Permalink
Comments
I definitely hate them. And when I get to the register to find that the sticker price is not the price, I feel borderline defrauded. When a Publix was readily available to me, I would go reasonably out of my way to shop there because they didn't use the cards.
The third function of course is that they ease collection of information about your purchases. Annoying, but not too onerous until you start considering ways this info could be funneled to the government.
I would think "good" price discrimination would be better correlated to the amount of resources the customer has rather than their tolerance level for playing card games. It sort of is, I have the resources not to worry about the cards at all but the most frequented stores, but a price discrimination system that makes certain customers not want to shop at your store at all seems not so good, unless you think those customers are particularly bothersome, like shoplifters or civil libertarians.
Posted by: Andrew at May 16, 2008 4:43:23 AM
The collection of information is on average good for the customer, as it allows the retailer to customize the offering, both at the individual (mailings) level and at the aggregate (store) level.
Posted by: londenio at May 16, 2008 4:55:28 AM
The one area where these programs make a huge difference in consumption patterns is in business travel. I know that I stay at a certain hotel chain that I may not have stayed at otherwise becuase they give me free nights (personally), and since my job is paying, I'm not really price sensitive. That's clearly non-optimal from a societal point of view.
Posted by: Joe T at May 16, 2008 5:12:03 AM
There are also a few other factors at play. In fact the behavioral economics of loyalty programmes are immensely complex.
One factor is that the points one earns in rewards can be spent *guiltlessly* on extravagances. Mentally people do not feel the same "I should really be using this to pay my heating bill" pang of guilt when buying leisure flights (or other treats) with points compared to when using cash. Hence offering points may on balance be a more pleasurable experience on balance than reducing prices.
In frequent flyer programmes, there is probably an element of insurance built-in. "Look, I'll fly with your airline whenever I can, and I won't quibble too much about business class fares but, in return, when I need to get to a funeral and the flight's overbooked I damn well expect priority on the waitlist." This is only possible if the airline knows who I am. (In the same way, I negotiate less ferociously with a car dealer if I am planning to have my car serviced at the same dealership, since the extra £200 may buy better service for the next five years. I also tip a little more generously at restaurants to which I plan to return).
A third thought is that large spenders will get their personal preferences better catered for if they use these cards. A UK supermarket chain was planning to de-list Feta cheese, until it learned from card data that, while very few people bought it, those people were disproportionately likely to be the shop's highest spending customers. In some ways, loyalty card data may protect your favourite niche brands: there is a difference between delisting something which lots of people buy rarely and something which few people buy always. I once moved my entire grocery budget away from one store when they stopped selling Grape Nuts.
Posted by: Rory Sutherland at May 16, 2008 6:30:46 AM
Rory,
The Feta cheese story is extremely interesting. Any pointers where I can find out a bit more about it?
Posted by: londenio at May 16, 2008 6:52:54 AM
The cards also allow the stores to target you for junk mail (advertisements, coupons of dubious value, etc.) The info they use for this is also often sold to others (or would seem to be based on increased junk mail.) The annoyance at this (both personally and for the damage done to the environment when CVS prints more glossy fliers to send out that I don't want) was enough to make me stop using the cards. This is so since it's fairly often that a clerk will just swipe a "store" card anyway, at least at a CVS or Duane Reade or the like.
Posted by: Matt at May 16, 2008 7:12:19 AM
...or, you could just take a picture of the card and leave it in your phone.
Posted by: shawn at May 16, 2008 7:48:37 AM
On the collusion front, this recent paper by Mara Lederman (PDF) identifies a 5% price premium at major airports as a result of frequent flyer programs. To distinguish collusion from other effects she looked at frequent-flyer partnerships and found that after the partnerships took effect, airlines' fares increased at specifically those airports where partners dominated the flight schedule. For example, United experienced higher fares on its flights from Atlanta - an airport at which its partner, Delta, operates the majority of flights and where most people collect Delta frequent flyer miles.
Posted by: tom s. at May 16, 2008 8:02:15 AM
The equivalent of the Feta cheese story was around 40-odd years ago among British supermarket managements. The importance of some low turnover items for attracting trade was noted without card data.
Me, I don't like loyalty cards much; but I have one for my local supermarket because it does not change my shopping choices but does result in marginally lower prices. I.e.,it marginally improves my welfare. Some other peole seem to like them, no-one is obliged to take them, so aggregate welfare is probably improved by them.
Posted by: David Heigham at May 16, 2008 8:05:04 AM
"Overall you personally benefit from loyalty card programs if you don't mind holding the cards (you have a thick wallet) and you have a strongly favorite company/product anyway."
You might be able to get round this problem. Using your iphone to store images of your cards so the bar code can be read.
http://www.tisgoud.nl/blog/2008/05/11/WalletCardsOnMyIPhone.aspx
This reduces the amount of loyalty (effort) required to carry a loyalty card.
Posted by: davidc at May 16, 2008 8:07:05 AM
I have a collection of the cards from having travelled throughout the country. I have never used by real name or Address. The hygiene of the data generated from the cards has to be lousy. Also, most stores have a customer non-compliant card at each register that they will use when you tell them that either you get the lower price or you walk out the door.
Posted by: superdestroyer at May 16, 2008 8:28:31 AM
I'll buy the lock-in effects for airlines because of the points accumulation, but not for grocery stores. At least for me, they're only good for instant discounts and since I have a card for every store in the area that has one, I generally just shop at whichever store is most convenient from a time perspective.
As for the pain of carrying them, I also don't get it for grocery stores. For me they just sit right behind my credit cards as they're the second most commonly used things in my wallet, right after the credit cards. The hotel / airline / B&N / Petsmart yadda yadda annoy me because they're used less frequently, but for most of them, they can find me off of my phone # so there's no need to carry the card.
Posted by: Jody at May 16, 2008 8:37:19 AM
I do most of my shopping at Schnucks (since it's two blocks from my house). They have a gas station nearby, and their loyalty program is that when I spend some amount of money on groceries, then I get a few cents off a gallon or something like that. The problem is that I don't have a car, and I can only use the discount to buy gas. Even that would be fine, except they bug me about it every time I go shopping.
Posted by: Lucas at May 16, 2008 8:41:12 AM
Loyalty cards also give stores the ability to restrict special deals to a limited number of items.
Right now, Kroger is letting people with loyalty cards buy up to four $330 gift cards for $300 each. (Obviously trying to capture some of that economic stimulus package.)
They couldn't make this offer at all if they didn't have the loyalty cards as a means of stopping wealthy customers like me from buying forty cards and getting my next few years worth of groceries at 10% off.
Posted by: Brock at May 16, 2008 9:14:10 AM
The Wall Street journal did a study on grocery stores that use "preferred cards." People who used them actually spent more than those who shopped in non-card stores for the same items. The sales items look cheap but everything else is more expensive in the card stores. Stores should be required to compete on price. Wal-Mart forever!!
Posted by: jorod at May 16, 2008 9:38:46 AM
"since my job is paying, I'm not really price sensitive. That's clearly non-optimal from a societal point of view."
Gaming corporate travel policies: Good for airlines, bad for us. Business travelers will buy a $2000 ticket to avoid a layover when a $200 one will do. My wife was complaining about how she was denied a raise of $1700 with her promotion and the next week her new boss had her go with $1800 more expensive plane tickets. Resource allocation by the incentives for the decision makers.
Posted by: Andrew at May 16, 2008 10:02:43 AM
I ignore frequent flier programs unless I am taking the flight anyway. If I worked at a company (and they were paying), I would, of course, ALWAYS use the airline linked to my personal FF miles.
I get cash back on both of my credit cards, so I do not care which I use.
Supermarkets are the most interesting. They will give a card to anyone, for free, so why not take it. If you dislike carrying the card, just give them your ph#.
Interestingly, I heard that one market (Safeway? Sainburys?) accumulated terabytes of loyalty data but couldn't figure out how to turn it into revenue. They tossed it and ended their loyalty card program.
Posted by: David Zetland at May 16, 2008 10:20:58 AM
You must really be desperate for people to comment on your writings! "These programs serve two functions. First, they are a form of price discrimination. Buyers who are willing to collect and show the cards pay lower prices while the "I can't be bothered with this ****" types pay higher prices."
Let me explain it this way. "Students that apply themselves, turn in all their assigments, and pay attention in class, get the best Jobs that pay high salaries while I can't be bothered with this ****! This is discrimination against me!
Bill
Posted by: bill at May 16, 2008 10:30:18 AM
...connectedness/rfid/crowd data could remove this. once we can get data out of supermarkets (which they wouldn't give, but would be easy enough to get in a crowd-source way once we get the tech) about what they're charging for what, you could link your phone to tell you where the cheapest place to buy your shopping list would be, incorporating loyalty cards or not...
eh, nevermind...my brain's going to explode with all the possibilities. too much for a comment, but tech+data kicks ass.
Posted by: shawn at May 16, 2008 10:33:09 AM
Stop & Shop supermarkets in the Northeast have the most lenient policy of all when it comes to loyalty cards. Many of their sale pricess are ostensibly available to cardholders only. As a matter of policy, however, the cashiers will give the cardholder price to anyone who asks. Only if one uses the self-service checkouts is it necessary to have a card in order to get sale prices, and those checkouts work so poorly most people seem to avoid them.
Posted by: Peter at May 16, 2008 10:41:26 AM
How bothersome is it to sign up for a supermarket loyalty program? They'll do it for you instantly upon your first checkout; you often get a keychain fob instead of a full sized card; anytime you forget it you can just use your phone # (or a fake one you first provided); you can even use you iPhone or like device instead if it can capture the barcode image. It's hard to imagine the implementation of any sort of coupon program (mail, email, handout, etc.) that is less onerous than that.
How much lock-in do these programs really create? In practical usage, I'm guessing most customers use them simply for the immediate price discounts, and not for the various "rewards" or "points" they accumulate. So any vendor who offers any kind of similar program is competing on the same level. Nowadays, if you insist on going to a small shop that doesn't have such a program, you're definitely paying higher prices anyway.
We now have retailing giants competing heavily against one another driving the prices to the floor. All you have to do is swipe a fob at checkout. This feels a lot better for the customer than the old days when people would drive around to multiple locations to get a good selection and price (the market, the bakery, the butcher).
Posted by: Mike McLellan at May 16, 2008 10:56:31 AM
One of the more interesting items concerning these programs, was the reported mass data dump one of the major grocery chains was reported to give to the FBI following every republicans favorite day in history.
9-11.
And the potential philosophical implications for "privacy" or lack of it.
Posted by: Lawrence at May 16, 2008 10:58:54 AM
Tyler,
I had not read elsewhere about price discrimination function of these cards, but it makes a lot of sense. I always get new insights like that from this blog.
Supermarkets have demonstrated that possession of an identifying number is what's important, not physical possession of the card. So I expect other retailers to soon follow suit, and the thick wallet syndrome to end.
I can't imagine why brand marketers would not purchase retailer's data on customer purchases. Identifying high potential zip codes - or even high potential individuals - through data mining should lead to more efficient direct mail marketing. Also, marketers should want better data on customer behavior, such as what percentage of sales are made to frequent repeat customers, how often do redeemers of direct mail coupons follow up with additional product purchases, etc.
One can envision marketers mining data from different sources to estimate potential effectiveness of advertising dollars, such as whether subscribers of Conde Nast or Outdoors are more likely to be sunscreen purchasers. I think the potential value of retail purchase data is yet to be realized.
Posted by: John Dewey at May 16, 2008 11:01:04 AM
superdestroyer: "I have a collection of the cards from having travelled throughout the country. I have never used by real name or Address. The hygiene of the data generated from the cards has to be lousy."
Using a false name or address to obtain retail discounts is a form of fraud, isn't it? I may be naive, but I doubt many U.S. consumers engage in fraudulent activity. Of course, I would expect a perpetrator of such fraud to rationalize his or her behavior by arguing that everyone else must be doing it. So someone who commits fraud would be more likely than I am to question the data retailers are accumulating.
Posted by: John Dewey at May 16, 2008 11:06:09 AM
Personally I hate the supermarket cards. I especially hate them when I'm traveling and run to a regional supermarket that wants me to get a card in order to shop there once (since the chain doesn't exist where I live, I won't use it often enough to actually carry it).
My solution is simple. If a product I am considering purchasing there has a "card" price that makes it cheaper, I simply don't buy that type of product there, because obviously if I don't use the card I'm paying more than I should. This means that I can walk into a grocery store intending to purchase $100 worth of items, but due to their frequent shopper card discounts being advertised, only actually buy $25 worth.
Obviously this is costing them revenue and volume. For someone like me without a card, they're essentially getting no return at all on holding a lot of inventory that I'll never consider purchasing from them.
Maybe that's made up for by local buyer lock-in, but I suspect since they likely only analyze ongoing data from those who have their card and use it, their results are going to be biased against recognizing that "non-card" average purchase likely went down after they introduced their card program.
Eventually I've just learned to avoid places I think have a card program and instead find the best prices elsewhere.
Posted by: Sharper at May 16, 2008 11:48:39 AM
John Dewey,
I doubt that you could stretch the criminal laws on fraud to customer discount cards. The Honolulu Advertiser had a story about people using all sorts of names since Safeway in Hawaii had a policy of using the name associated with the card to say thank you? One college student used the name Albert Einstein so that the clerks would say, "Thanks You, Mr. Einstein."
I guess the automatic checkout would be a way of forcing their use. One time I forgot to take my card to Wegman's, the clerk refused to use a non-compliant card. I told her to take everything out of the basket that I had to pay the higher price. She quickly relented.
I different type if the pay $25 dollars at Barnes and Noble and get 10% off each time. If you buy more than $250 a year, it can be worth it.
Posted by: superdestroyer at May 16, 2008 11:58:41 AM
Someone correct me if I'm wrong, but isn't price discrimination welfare-improving only compared to a monopoly? The competitive equilibrium is equally efficient, and additionally maximizes the consumer surplus. So, the loyalty program might be welfare improving, but is not consumer-surplus-maximizing. And as a consumer, that's all I care about.
And an additional reason to hate the cards, at least in grocery stores: they allow the merchant to gather information on the consumer's habits. And the more they know about me, the more they can figure out what's the maximum price they can charge me.
Posted by: pc at May 16, 2008 12:05:04 PM
how much would you pay for knowing the way your customers spend their money? retail loyalty cards collect this information for you.
Posted by: antonio at May 16, 2008 12:05:34 PM
Note that I always get all the 'membership' benefits but don't carry *any* of the membership cards with me - I like keeping a slim wallet. I don't even have accounts at many of these businesses, mainly the grocery stores. Most of these stores allow you to enter your phone number instead of swiping a card (I just use an old number for one of my friends) and I've found that some of them will even accept *any* number, even if it's not a valid phone number. Others have a membership card behind the counter and will just swipe it for you if you say you forgot your card.
This doesn't go against wanting to understand or otherwise discuss this topic, but I thought it worth noting. I personally think the membership cards are very annoying, if for nothing else because I have to spend extra time parsing all the price signs for products that I want to buy and/or compare. Sometimes the 'discounted' price doesn't include a unit price, which is very annoying and I try to avoid those stores.
But I certainly benefit from the system because I get lower prices than I probably otherwise could due to price discrimination - people who perceive the costs of dealing with memberships pay higher prices, which effectively subsidize the occasional advertised 'bargain' deal to get new customers into the store to become members. Sure, there's a time commitment on my part that offsets this advantage but I'd argue it's very, very low since actual participation in these membership programs is trivial (and doesn't even increase the physical size of your wallet!).
-Thomas F.
Posted by: Thracx at May 16, 2008 12:47:32 PM
A lot of TC's analysis seems to presume loyalty card exclusivity - particularly the lock-in component, but of course there's nothing preventing me from having a 'loyalty' card for every merchant I was going to use anyway (not much loyalty there). An interesting question is what happens to the model when every retailer has a loyalty/discount card. I suspect the subsidization of cardholders by noncardholders would taper off and the entire function at that point would be discounts in exchange for identification for targeted marketing purposes.
Posted by: sidereal at May 16, 2008 12:53:39 PM
Things will be a lot nicer when we have digital wallets:
Here's a start http://www.tuaw.com/2008/05/14/scanning-wallet-cards-into-the-iphone/
But we'll be acquiring many more silly memberships then if it is easier. The practice will probably (yech) become more widespread if it is less pain on the user (in the end creating more annoyance).
Posted by: sammymalacky at May 16, 2008 1:15:15 PM
As someone pointed out above, just tell the person at the checkout that you do not have a card but you want the card price. Only once have I gotten a "no" and it was at a supermarket that had just started their card program. "Sorry it is against store policy." A year later, they changed the policy. I have to admit I enjoyed walking out and leaving them a cart full of groceries to re-stock. Yes, I did have to drive somewhere else to shop, but it was my way of sticking it to the man.
One time I did feel out an "application" for a loyalty card because they would give a small kick back to my kids school. According to the application, I live on "123 None-of-your-damn-business Lane." No questions asked.
Posted by: subrosa at May 16, 2008 1:17:14 PM
A few people have made the point about phone numbers. I will admit to leaching off friends' and family loyalty cards by giving their phone numbers. At Petco, this helped us reach the 10 bag discount threshold for Eukanuba dog food. But at PetsMart, this recently made me uneasy because the dude at the register said he'd be calling me the next night at dinner, to which I had to reply that I didn't swing that way.
I have no idea how the stores get any useful data out of loyalty card programs in and of themselves. I agree that they are mostly a mechanism for price discrimination, but I don't see it as a "discount for members". Instead, I see it as letting non-members and loyalty card haters pay more. About the only place I can see that the stores could get something out of the data is in finding optimal price points by conducting price experiments across brands and club discounts. The soda aisle seems hyperactive with such price experimentation.
Posted by: BoscoH at May 16, 2008 1:40:15 PM
I'm surprised no one brought up Vons using their loyalty card records in a court case where a consumer had slipped and injured himself in the store. They wanted to say that he was drunk and and that he had numerous purchases of alcohol. The evidence was never introduced because the plaintiff went to the media and there was a shitstorm. I buy beer at the supermarket during almost every visit, but I don't shop drunk.
Posted by: Mo at May 16, 2008 1:51:35 PM
Londenio, the Feta story came from Safeway's mid 1990s ABC loyalty programme in the UK operated, I think, in partnership with IBM. It was a pioneer scheme, but was abandoned before Safeway itself was acquired by Morrison's.
Sainsbury's abandoned its scheme in a way, but replaced it with the Nectar Card.
http://snipurl.com/296t8 is a wonderful book describing the success of Tesco's loyalty programme, the Clubcard. It contains an extraordinary story about an early privacy breach, of which you'll find details below.
When Tesco introduced internet grocery shopping, they thoughfully prepopulated the favourites list with whatever items the shopper bought most frequently. A sensible idea, except it overlooked the fact that many couples or families often shared cards or a household account.
A phone call came through early on day one: "Why does my Favourites list contain condoms, when my husband never uses them?" asked a distraught caller.
I suppose the unkind answer would have been "Not with you, love." After a hastily convened conference they called back. "There has been a surprise data glitch causing this quite unfortunate error. Please be assured the item has now been removed from the list."
Phew!
Posted by: Rory Sutherland at May 16, 2008 2:32:51 PM
"Posted by: Mo at May 16, 2008 1:51:35 PM"
Obviously the purpose of introducing the evidence or threatening to introduce the evidence was to embarrass the plaintiff into settling cheaply without a trial, not for probative value [which it obviously doesn't have, because it doesn't show him to be drunk in the store or even to have drunk the booze himself -- he might have bought it for his wife].
-dk
Posted by: Dick King at May 16, 2008 3:57:27 PM
what's with all the scanning cards onto your iphone links. are we being spammed?
Posted by: ertgweg at May 17, 2008 5:36:34 AM
ertgweg...not as for me; seems nobody read my initial comment. :)
Posted by: shawn at May 17, 2008 3:14:36 PM
THis is a bizarre comment from Costco fan Kevin Drum.
Posted by: Sanjay at May 18, 2008 11:38:25 AM
I heartily encourage people to trade their loyalty cards as often as possible. I carry a couple and I have no idea who they are registered to.
And given the chance, I'll trade them again. So far as I can tell, the only downside is that somewhere there may be a shopping record of me buying incontinence products. Big deal. Plausible deniability.
Posted by: transiit at May 18, 2008 8:58:19 PM
superdestroyer: "I doubt that you could stretch the criminal laws on fraud to customer discount cards."
Well, I guess I shouldn't be surprised at that attitude: if I cannot be caught - or cannot be prosecuted - then the bahavior is not wrong.
You may be correct that lawmakers probably do not bother writing criminal statutes that cover such minor offenses. Still, providing false information to intentionally deceive in order to realize monetary gain - in order to receive the discount - meets the common definition of fraud. One can always rationalize unethical behavior by arguing that large portions of Americans commit such fraud. But the bahavior is still unethical and it is still fraud.
Posted by: John Dewey at May 19, 2008 5:36:41 AM
This has to rank with the most worthless threads on the internet. I believe about 75% of the people who have posted should be commited for even whinning about these cards. Grow up and quit the whinning, use them if you want, don't use them if you do not want to. Why not complain about something that really matters, rather than wasting internet space.
Posted by: bill at May 19, 2008 11:42:59 AM
