Tools: Open Source Behavioral Cost Of Personalized Pricing
Recently my wife and I needed to call a car, so out came the apps to compare rideshare prices. There’s always a bit of variation here, but this time was striking.
My wife’s Uber app quoted her $28, while mine gave me $47. Same app, time, and place - but two wildly-different prices.
Who knows why? I’m usually more willing to spend than she is, and I bet that's represented on my user profile. I was paying with a gift card, which surely contributes. Maybe it was a price scraping update, comparison shopping detection, or a system that explores “face-in-the-door” high prices before backing down. From the outside, no one really knows.
That makes me worry. When prices are based on behavior, they incentivize us to behave performatively - after all, the squeaky wheel gets the grease.
Price discrimination has existed for approximately forever - like senior discounts, or coupon books. A business can make a little more money by selectively lowering prices for people who can jump through certain hoops (and might not buy without the discount).
In this simplified demand curve, if the price is $9, then the business only earns $45.
By pricing at $10 for most, with a $1 discount for the price-sensitive blue buyer, the business earns $49 instead.
Behavioral discrimination isn’t a new practice either. When your internet service provider tells you they’re raising rates, do you pay up? Or do you call customer service, wait on hold, threaten to cancel, get transferred to “retention”, and learn about some exciting new discounts that you qualify for? Not everyone has the time or patience to run this gauntlet - and that’s exactly why businesses do it.
When you bring price discrimination into the digital world, there’s nothing different in any individual case. But as the apocryphal Stalin quote goes, quantity has a quality all of its own. Technology gives us not only quantity, but ubiquity - every piece of behavioral context can contribute to your prices.
Unlike in meatspace, every digital action can be cheaply recorded and analyzed.
Source: HackerNews
