Won’t happen but what if it did?

By the mid-2000s, some economists began wondering whether Big Data could discern every individual’s own personal demand curve—thereby turning the classroom hypothetical of “perfect price discrimination” (a price that’s calibrated precisely to the maximum that you will pay) into an actual possibility.

What if – we’d all be vastly poorer as the consumer surplus would vanish.

24 comments on “Won’t happen but what if it did?

  1. Ah but why would it be one-sided? – i suspect consumers would form buyers’ alliances – in the same way producers form company’s to get back some of that sweet surplus.

  2. Doesn’t competition prevent this from happening? If company X tries to charge me as much as I can bear, company Y steps in and offers me the same deal for less.

    This already happened in the airline sector. By the 1990s, the old airlines (BA etc.) had price discrimination down to a T. Then along came Ryanair and ripped up the rulebook.

    Similar for Tesco’s vs Aldi; and no doubt plenty of other industries.

  3. This then creates arbitrage opportunities. If you have zero interest in Adele’s latest hit and can buy it for $0.01, you can then sell to someone who really wants it and get’s charged $10.00 by the retailer.

  4. @David Moore,

    Aah yes, but this would of course have to be priced in, and so we’d just end up with a very expensive way of arriving back at… the market clearing price. Which spontaneously appears without masses of bureaucrats.

    It’s the old planner’s conundrum: if we could perfectly plan, we could plan the market clearing price. Which would be more or less the same as the free market clearing price, so why bother?

  5. Which would be more or less the same as the free market clearing price, so why bother?

    Which assumes rational planners. And that those planners aren’t rationally interested in power, budget and empire building (which might be counter-productive from the wider viewpoint but could be entirely rational within the local context.)

  6. Isn’t there also an argument that for some people the price would drop? Currently the market clearing price is set to maximize profit (not units sold). Someone may be willing to pay more than the unit cost of production but not the market price so for those people you offer exclusive special offers and consumers end up better off. Particularly if as some people have pointed out competition stops the market clearing price rising.

  7. I suspect this operates in the travel industry world today.

    Have you never noticed that if you start looking at a particular site for flights (say London to Singapore), that some of them have “must go soon” type tickets, usually labelled something like “only 3 seats left at £599” or something similar, even if the flight is several months away.

    Come back 5 minutes later and the ticket price is £699.

    However, if you use another machine and switch to a different IP address by going to the pub next door, all of a sudden the price is £599 again (or possibly even cheaper).

    I think that some of it is them trying to track a specific customer and raising the price to try and maximise that, but I suspect that some airline websites work by having a base price, but monitoring interest in a specific flight and as interest (queries and bookings for that flight) increase, so does the price until interest tails off at which point the price gradually reduces until actual purchases pick up again or if nobody buys it goes down to the base price again.

    I’m not sure if this is how it works, or how it works only for specific airlines, but its how it seems to work for some of them.

  8. This ‘Big Data’ stuff is certainly going to people’s heads. I wonder how many companies will go bust pursuing will-o’-the-wisp correlations from BD?

  9. JG: some of it is them trying to track a specific customer and raising the price to try and maximise that.

    Something like that seems to happen with Amazon. Sometime the price changes quite dramaticly

  10. It’s not possible, certainly not perfectly so. As someone like Hayek would say, such things are discoverable only through the operation of the market itself. And all that is dependent on specific time and place and every other competing option available at the same time (including not buying anything but reading a book under a tree), all channeled through the want parts of the human brains. It is simply not possible to know in advance what James G would pay for x product. Time, place and every other variable within my life and everyone else’s and the market would need to be computed in real time continuously.

  11. JG,

    They first started doing that based on browser, there was quite a stink when Apple users found out the Safari was offering much higher prices. The rest of us just sniggered.

  12. They first started doing that based on browser, there was quite a stink when Apple users found out the Safari was offering much higher prices. The rest of us just sniggered.

    I do hope that is true, the schadenboner would be huge. It does illustrate the point though, since people who pay over the odds for something like an iPhone or iPad when there are cheaper, better and more powerful versions available (I use the Samsung Galaxy Tab S and a Samsung phone both running Android)

    The question comes down to “what factors correlate with being able to charge customers more and them paying it”, clearly using Safari would have a pretty good, but not perfect correlation (for example children using their parents cast off older devices)

    Past performance does not equal future performance may be true in terms of share prices, but as for price discrimination it is a good enough indicator to put money in the bank.

  13. Interesting, although it is not quite as clear cut as I originally envisaged.

    Measuring Price Discrimination and Steering on E-commerce Web Sites (PDF)

    Page 11 talks about Travelocity providing different search results from users of the Safari browser than other types of browser users:

    Travelocity. As shown in Figure 11, Travelocity alters hotel search results for users who browse from iOS devices.

    Figures 11[a] and [b] show that users browsing with Safari on iOS receive slightly different hotels, and in a much different order, than users browsing from Chrome on Android, Safari on OS X, or other desktop browsers. Note that we started our Android treatment at a later date than the other treatments, specifically to determine if the observed changes on Travelocity occurred on all mobile platforms or just iOS.

    Although Figure 11[c] shows that this reordering does not result in price steering, Figures 11[d] and [e] indicate that Travelocity does modify prices for iOS users. Specifically, prices fall by ≈$15 on ≈5% of hotels (or 5 out 50 per page) for iOS users. The noise in Figure 11[e] (e.g., prices increasing by $50 for Chrome and IE 8 users) is not significant: this result is due to a single hotel that changed price.

    The takeaway from Figure 11 is that we observe evidence consistent with price discrimination in favor of iOS users on Travelocity 4.

    Unlike Cheaptickets and Orbitz, which clearly mark sale-price “Members Only” deals, there is no visual cue on Travelocity’s results that indicates prices have been changed for iOS users.

    Online travel retailers have publicly stated that mobile users are a high-growth customer segment, which may explain why Travelocity offers price incentives to iOS users [26].

  14. More likely it would be made illegal to use such information to set prices. Like the EU with male and female car insurance

  15. More likely it would be made illegal to use such information to set prices. Like the EU with male and female car insurance

    The EU…? Is that still a thing?

  16. Interesting fight.
    Big data doing price discrimination on commodity purchases
    vs
    More and more consumers seeking customised purchases.

    My bet is that consumers will win.

  17. Local transit company spent a lot of money on a new electronic fare card system and part of the justification was look at all the wonderful data we will have so we can plan more efficiently, like a lot of big data it seems to ignore the accuracy vs cost side, are the efficiency gains from the extra accuracy going to cover the cost. Among many interesting insights they now know the % of people who get off a bus at a train station and then get on the train, apparently it’s very high, who would have guessed that.

  18. Avoiding price discrimination is one reason why agents exist. This applies whether it’s a travel agent or an estate agent. You might claim that a flight to Singapore costs £599 one day and £699 when you come to book; but if you hide your interest and check prices via an agent (Skyscanner, Kayak, Google Flights, etc.) they simply can’t pull that trick on you.

    Similarly, if a plot of land is worth hundreds of thousands to you because it could be used as an access road to a larger site, but is worth buttons otherwise because it’s too small for a home, employing a buying agent can make a huge difference. (Just don’t employ Iris Robinson.)

    In effect, the presence of an agent tilts the power relationship. Instead of the seller asking the buyer how much they can afford to spend, the buyer asks the seller (via the agent) how cheaply they can afford to sell.

    When the agent gets too big, power shifts to the agent itself. Amazon is effectively an agent for lots of little sellers; and it knows everything about its customers. It has huge price discrimination power; strangely enough they don’t seem to use it.

  19. When Big Data ‘dicerns’ my personal demand curve, I would be obliged if they would tell me as I haven’t a clue what it is.

  20. This will be price offered, not price charged.

    Will there be any way to ‘dicern’ how many consumers offered higher prices were put off purchasing?

  21. Will there be any way to ‘dicern’ how many consumers offered higher prices were put off purchasing?

    But this is exactly where the pricing mechanisms of the stock market and everyday consumer items start to come into play, because instead of looking at the price as a static thing you can look at it as a dynamic based upon the bid/offer spread.

    All you need is the other side of the argument where instead of clicking “Buy”, you click the other button which is “Not at that price” and are then prompted to make an offer.

    The provider doesn’t need to instantly say “Sold” (although they could if their automated mechanism determines that the profit is acceptable), but they might come back to you later (if sales are sluggish) and say “Still interested in that flight at your preferred price of $399?”

    At the very least they are finding out what customers own view of the price should be.

    This is what markets are all about, price discovery.

  22. I’m pretty sure my demand curve wobbles from day to day.

    And it is stable only for big ticket items, like cars and houses, for which it will be very tricky to fluctuate price.

  23. John Galt,

    “The question comes down to “what factors correlate with being able to charge customers more and them paying it”, clearly using Safari would have a pretty good, but not perfect correlation (for example children using their parents cast off older devices)”

    Vogue readers. No-one else is such a mug as these people. Posh, pretty, not very bright girls. Too stupid to figure out that a £500 handbag is madness, but who can access that sort of money.

    Oh, and people who are into designer gin and vodka. Easy prey.

Leave a Reply

Name and email are required. Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.