I briefly did some math around the capacity curve for said services. I'm not entirely sure how they're able to make it work effectively, and scale revenue superlinearly to cost. Also, the demand curves are less than awesome, given you have to deal with two spikes a day, and then no business at all. Additionally, fault-tolerance was problematic. I had modelled it with the idea of providing valet services that are available in FiDi, at a cost-competitive rate, and using a parking lot in Mission Bay or Hunter's Point during the day. A failure in the arterials roads could be incredibly problematic (i.e. 1st, 2nd, 4th, 5th) to the model, because it would cause service time to unexpectedly go up.
I think there are work arounds to this model, but I fear for Vatler. I'm sure that the Vatler team has done math to make these things make sense for their model. Best of luck to them.
I think there are work arounds to this model, but I fear for Vatler. I'm sure that the Vatler team has done math to make these things make sense for their model. Best of luck to them.