Most retailers have the address of the source and destination of every case they've shipped going back many years. Most bus companies have the starting and ending points for every trip that most riders have taken.
I don't work at Uber, but I do work at a competitor of theirs. We have mixed feelings on tracking location data. There are customer advantages:
1. We could monitor drivers to make sure they do not take egregiously long routes.
2. We could use the information to drive supply to the right place (for example, by telling drivers, "Yes, we understand that hails are not as common in AREA, but our data indicates that when they do exist, they tend to be for rides 200% the length of the average ride).
3. And there are all kinds of ways we could tune our service in response to highly detailed data about each ride. ("Avoid this road -- it's a little shorter, but it's massively slower!", or "Driver X really knows part of city Y, let's preferentially hail him there"), though in my experience those kinds of things are hard to act on.
But you're right, it is creepy, and we don't want to be in the business of monitoring the movements of our passengers (or even our drivers, outside of in specific ways directly to do with providing a better service). We particularly do not want to be in the position of being a good target to subpoena about the movements of a customer.
I think it's possible to thread this needle, though there's a fair amount of effort involved. Anonymizing the data, in some cases aggregating it, and/or making it less precise over time (cutting sig digits off it) can preserve passenger and driver privacy while still giving you most or all of the benefits of the data collection.
Well, I can't speak to what's on the roadmaps of our competitors -- they don't share them with us! ;)
Speaking personally, I hadn't heard of kutsuplus, but we have tossed around ideas of doing some kind of "share a taxi" service (as distinct from "ride sharing.") It's an attractive concept, particularly as a way of attacking the basic problem of the taxi (writ broadly) industry: inflexible supply.
(As a digression: the demand for taxis is very spiky, both in a cyclical way (Friday nights are much bigger than Tuesday mid-day) and in special event ways (New Year's Eve and Your Music Festival Here will increase demand by multiples of the usual amount). But it's generally not economically worthwhile to hold a bunch of taxis sitting around in a parking lot only to roll them out a couple of times per year. So if you can find a way to get supply to your customers during high-demand periods, that's super-great).
The thing is, for useful taxi sharing, you need two things:
1. You need to know that person A and person B are fairly near each other and want a taxi at the same time.
2. You need to know that person A and person B want to go someplace fairly near each other.
#1 is pretty easy to know. Though it's not totally trivial. If person A tries to hail, you don't know yet that person B will try to hail 5 minutes later, and if your app is functioning well, it will probably have already dispatched a taxi to person A. Still, in high-demand times, perhaps you haven't yet gotten a taxi to person A by the time person B wants to hail.
#2 is much harder. Passengers do not like typing their destinations into a smartphone keyboard. In general, on every app I'm familiar with, most of the time you're blind to the destination of the passenger. The driver just rolls up and asks them about it.
Even if you solve #2, there are some timing, privacy, and interface issues to be concerned with. For example, imagine you're the second passenger to be picked up. Say you're someone who feels a little physically threatened by some people. Now a taxi rolls up, and the person you're to be sharing the back seat with is a large man who seems inebriated. Obviously, we don't want to share the exact location (pick up or drop off) of our passengers with other passengers -- but how do we inform a passenger of the potential ride share while keeping the other party private? Probably the answer is something like, "There's another passenger within 1/2 mile of you who would like to share a ride -- are you okay with doing it? You'll be dropped off first."
None of these are reasons not to do taxi sharing. I think they can be solved, and I think that the reward for sharing is big, especially in situations like a music festival, where at least one endpoint of the trip is shared. They're tricky enough that our app won't be rolling them out next week or anything, though.
One solution to this problems is asking the required information from users, even in advance(say 20 minutes), and in return offer the relevant discount. Do you think it'll work, and is it possible to sell this to drivers?
I wouldn't call it creepy, but for those that value their privacy, I would call it concerning. In the 'post-Snowden' era, it's not far fetched to imagine that the NSA et all would or could get their hands on the data.
Then again, if your privacy is so essential, you should probably already be wary of using any mobile device/web service to hail a taxi.
