Waze Carpool

» see waze.com/carpool as it lived, on the wayback machine

~ the goodbye ~

The empty seat problem was solved here before anybody wrote software for it. A dolmuş is a shared car on a fixed route that leaves when it fills up, the name means filled, and the fare is split between whoever is in it. It is not an app and it never needed to be. So the pitch Waze Carpool was making in San Francisco in 2016, that strangers going the same direction could ride together and share the cost of the fuel, arrived here as a description of the minibus at the end of the street.

Waze itself is on plenty of phones in Turkey and has been for years, which is what makes this an odd kind of distance. The app was in my pocket the whole time. The feature was in three countries and Turkey was never one of them, so what I had was the navigation half of a product whose other half only existed if you lived somewhere else.

The idea started earlier and closer to home for Waze than California. In July 2015 Google ran a pilot in Tel Aviv called RideWith, deliberately hobbled: a driver could take two trips a day, from home to work and back, and no more. That restriction is the whole design argument in miniature. The moment a driver can do a third trip, the thing stops being a commute and starts being a job, and a job is a taxi, and a taxi is regulated.

~ the chronicle ~

~ the fare that could not be a wage ~

Waze Carpool capped what a rider paid at the American federal mileage rate, which was 54 cents a mile when the service opened. That number is not a price, it is the figure the tax authority uses for what a car costs to run per mile. Pricing a ride at exactly that means a driver ends the trip level, having lost nothing and gained nothing except company.

Everything followed from the cap. There was no surge pricing, because there was no market, and no background checks of the sort a taxi commission demands, because nobody was operating a vehicle for hire. Waze charged the drivers nothing and, for long stretches, charged the riders nothing either, which made the economics of the product itself as flat as the fare. It was Google spending money to take cars off roads that Google was in the business of mapping.

~ the commute that did not come back ~

The service expanded out of the Bay Area across the United States and picked up two more countries, Brazil and Israel, and a standalone app in 2018 so it stopped being a tab inside the navigation software. Then in 2020 the thing it was built on stopped happening. A product that matches two people making the same journey at the same hour every weekday has nothing to work with when the weekday journey is a walk to the kitchen table.

Waze announced the end on 25 August 2022 and began switching it off the following month, saying plainly that its users had gone from mostly commuting to mostly errands. Nobody argued. It is one of the few Google shutdowns where the stated reason is the actual reason, and it is a strange thing to read now, a company explaining that it had built the right product for a world that stopped existing halfway through.

@}-->--

              )
             (_)
            .-'-.
            |   |
            |   |
            |   |
            |   |
          __|   |__   .-.
       .-'  |   |  `-:   :
      :     `---'     :-'
       `-._       _.-'
           '""""""

« previous  ·  random  ·  next »