A recent investigation by Consumer Reports details how Uber and Lyft use AI to charge you more money and charge different prices for the same ride. It’s an interesting read.
Lyft’s patents go much further, describing results and “sense” models that can be used to predict the “importance” or “priority” of a particular trip, arrival or location…and a willingness-to-pay score, defined as “the willingness of the requesting mobile device to pay a higher amount of transport service”.
We discussed in the past about rumors that Uber & Lyft charge more for those with credit, gift cards and promotions in their account. In my opinion, the idea of using the clause “willingness to pay” has serious legal questions.
Obviously, someone with a gift card balance in their account will be more willing to pay for a trip and use their balance. And so if they use the patented method, it would systematically charge more users with a gift/credit balance.
I’m no legal expert, but I don’t see how it can be legal to offer a prepaid option (load/gift card) and then charge more for the trip. They basically charge you $100 for a load and don’t give you $100 of services.
Similarly, imagine someone pays monthly Uber One or Lyft Pink fees for the promised ride discounts (5% from Lyft and 6% from Uber) and then opts for rides that are less than the discount. I don’t see how it can be legal to charge for a subscription and not provide the stated benefits.
Also unethical (but probably not illegal) would be partnering to give a benefit to Chase or AmEx cardholders (Chase for Lyft, Amex for Uber), and then charging more because of their “willingness to pay” score and reducing the benefit. In this case, cardholders are paying an annual fee, but not to Uber or Lyft themselves.
We have no data to know if Uber and Lyft are actually using the patented ‘willingness to pay’ score – all we have is rumors/anecdotal reports of people being charged more when they have a gift card balance. If they use ‘willingness to pay’ as a factor, there are serious legal issues to consider.
Even if they specifically train the algorithms to ignore credit balances, the balance will still affect rider behavior (eg ordering the ride quickly vs checking out the competition) which then results in a higher “willingness to pay” and higher prices.
That’s just my two cents, happy to hear other analysis in the comments below.
