United Airlines personalizes online prices based on estimated customer willingness to pay
IN MANY types of face-to-face retailing, it pays to size up your customer and tailor your offering accordingly. In a 2006 study of Fulton fish market in New York, Kathryn Graddy of Brandeis University found that dealers regularly charged Asian buyers less than whites because the Asians had proved, over time, more willing to reject high prices, and readier to band together to boycott dealers who ripped them off.
The internet, by allowing anonymous browsing and rapid price-comparing, was supposed to mean low, and equal, prices for all. Now, however, online retailers are being offered software that helps them detect shoppers who can afford to pay more or are in a hurry to buy, so as to present pricier options to them or simply charge more for the same stuff.
Cookies stored in shoppers’ web browsers may reveal where else they have been looking, giving some clues as to their income bracket and price-sensitivity. A shopper’s internet address may be linked to his physical address, letting sellers offer, say, one price for Bel Air, another for Compton. Doug Bryan of iCrossing, a digital-marketing consultancy, explains that the most up-to-date “price customisation” software can collate such clues with profiles of individual shoppers that internet sellers buy from online-data-aggregation firms. All this is fairly cheap, he says.
One of the few big online firms to admit to using such techniques is Orbitz, a travel website. Its software detects whether people browsing its site are using an Apple Mac or a Windows PC and, since it has found that Mac users tend to choose pricier hotels, that is what it recommends to them. Orbitz stresses that it does not charge people different rates for the same rooms, but some online firms are believed to be doing just that, for instance by charging full whack for those assumed to be willing and able to pay it, while offering promotional prices to the rest.
Allocating discounts with price-customisation software typically brings in two to four times as much money as offering the same discounts at random, claims Ravi Vijayaraghavan of 7, a Bangalore-based firm that develops and operates such software. One way to do this is to monitor how quickly shoppers click through towards the online seller’s payment page: those who already seem set on buying need not be tempted with a special offer. Firms like 7 and RichRelevance, another price-customisation software firm, from San Francisco, are somewhat keener to talk about their software than the internet retailers who are trying it out. Mr Vijayaraghavan names United Airlines, for example, as among his big clients, but the airline declined to comment for this article.
Andrew Fano, a consultant in Accenture’s Chicago office, reckons that at least six of America’s ten biggest web retailers are now customising prices in some way, but it is hard for shoppers to spot when this is going on. If they knew, many would feel that it is “pushing the boundaries” of fairness, notes Werner Reinartz, a University of Cologne marketing professor and consultant to two Fortune 500 companies that use customisation techniques.
Mr Reinartz preaches caution lest companies be dragged through an ordeal pioneered by Amazon in the autumn of 2000. Word broke that the internet giant was selling DVDs at differing prices, to see which web browsers happened to be favoured by shoppers least concerned about cost, former executives say. The resulting backlash prompted it to refund those who paid more, and Amazon now declines to discuss its pricing system.
Users of price-customisation software have so far been reluctant to peep at potential customers’ social-media pages, for fear that this would provoke a privacy backlash. But the operators at the call centres 7 runs for its clients are beginning to scan Twitter for gen on the shoppers they are talking to—and sometimes, says Mr Vijayaraghavan, their tweets give useful hints about whether a discount is needed to clinch the sale.