Moreover, that restriction would motivate people to make one-way reservations farther in advance, which would let the system guide folks to alternative destination lots (“You’re starting in Ballard. That restriction would reduce - potentially avoid - any decrease in local coverage, preserving customer satisfaction. Zipcar would get a ton of press, customers would try to solve one of your most intriguing scale problems, it would go a long way to associate Zipcar with open source and open government (which the Flexcar brand had a stronger association to), and it would cement Zipcar’s justification for a $50 late fee as necessary rather than onerous.Īt least in metro Seattle, I think you have enough cars to pull it off, especially with a few restrictions, like that you can’t return a car to a different destination lot when it leaves the origin lot empty (because all other cars there are already reserved). It could play out like a Netflix Prize for scheduling. With an anonymized rental history (where users are identified by GUIDs) and “home” car locations, I or anyone could make 5 or 10 fake scenarios where some percent of users request different destination lots at time of reservation. The first hard part is getting a machine-parsable, anonymized rental history and lot locations, which is why I’m emailing.
Nothing new there, but here’s why I’m emailing. I signed up for Zipcar and found that it didn’t have the killer feature for me: the ability to return a car to a different lot, even if that spot had to be selected at reservation time. The first car-sharing service to pull it off will have an amazing competitive advantage: paying for only what you actually need and the ability to get a car to the right place at the right time. Zipcar hasn’t and I didn’t receive a reply. As far as I know, no one has tackled this. The email after the jump has lots more details (slightly formatted and edited).
enforced policies, like: must reserve 1 day in advance one-way trip can’t leave a lot empty only offered at certain urban lots (initially).fees: “Returning to this location will cost $4 more”.incentives: “We’ll discount $2 for returning it a mere 5 blocks away from your preferred location”.Like Netflix, The Zipcar Prize would release an anonymized set of trip data, parking space locations, and an evaluation method (which might emphasize certain aspects, as Netflix did). In 2009, I emailed Zipcar about cloning The Netflix Prize for a similar purpose: crowd-sourcing (really, expert-sourcing) a set of algorithms, incentives, and restrictions that would let Zipcar offer one-way rentals. That means customers must rent (and pay for) one specific car for their entire trip, even when it will be parked in an urban area for many hours during the trip. Zipcar customers must return a car to the same parking spot that it was borrowed from.
ZIPCAR SEATTLE UPDATE
Update (December, 2012): Car2Go seems to be trying this and even uses municipal parking spaces, so trips could end essentially anywhere.