The thing that I didn't see in the data is while yes, the longer uber is in a city the less likely you are to wait, there is also a flip side of the driver no longer waiting as long when uber becomes more established in a city.
Let me explain...
When Uber was first in Chicago it was normal to wait 10-15 mins for a ride to pick you up. You would pull the app out while you still had a few things to do in the apartment before you left, order the Uber, and then by the time you were ready it was right out front waiting. The time waiting felt very seamless, although there might have been an actual 10 minute wait. When it also first started drivers were totally okay with waiting a minute or two for you because the demand wasn't as high back then. They didn't want to lose a potential customer. However, now that Uber IS more abundant in Chicago you can no longer do that little technique. Most drivers will get pissy with you if they have to wait outside for more than 30 seconds, or if you are not curbside waiting for the ride when they show up. Since drivers now also have more options for people to pick up, they can be more choosy on the customers they decide to take. They can easily cancel rides if they are too far from the pickup or are waiting to get a new customer. So since you can't "queue" up a ride like the "old days" you have order the ride when you are actually ready to go and are already waiting. Now that you do this, the wait is felt 100% by the customer (which was once avoided), if that makes sense. So what used to feel like nothing now feels like you are waiting in the cold for possible 5-10 whole minutes, which obviously often feels a lot longer that. Thus making you more likely to cancel the ride if you see an empty cab or another opportunity. So there is a bit of a catch 22 here, if I'm using that phrase correctly.
It does makes sense that they don't want to expose this side of things though, because that places more blame on their own quality of drivers. It probably shows that as a city is active longer, the ratings/quality of drivers goes down drastically. I've noticed it quite a bit here in Chicago, and it's really disappointing. The average quality of driver I get nowadays is almost 1 to 1.5 stars lower than what I got 2 years ago.
I'm curious to find out if anyone else has had this same experience or not.
To address one of your points, the driver can't start the ride until you're in the car. So you're costing them money by making them wait a minute or two.
Was this always the case? Sure. But when ETAs were longer and drivers generally had more downtime, they were paid more per minute of actual driving. So this wasn't as painful.
If you want, you could try texting the driver saying "Hi, I'll be down in a minute, feel free to start the meter!" They might appreciate that, but it might also be against policy.
> waiting in the cold for possible 5-10 whole minute
Isn't this largely solved by the GPS mapping of your ride from within the Uber app? You can easily see exactly where the driver is, and step outside right as the driver is pulling up, regardless of how long the total wait is.
In my experience (using Uber in Shanghai and Beijing), the latency can be 0 to 2 minutes, so it's hard to time it exactly. It takes me 2 minutes to get from my apartment to the place where a driver would normally pick me up.
One of the reasons I am frustrated about a long waiting time is wrong expectations set by the map. The map in SOMA typically shows that 5+ cars are within 2 minutes. As soon as I request, the closest is 7 minutes away.
AFAIK there isn't a lot they can do about it, because Uber drivers have to manually accept jobs. Those 5+ cars within 2 minutes presumably turned down the job.
Ok, but can they do something about the car that I do get saying "5 minutes" and actually meaning "9 minutes, including the bit where I watch him drive 2 minutes in the other direction"?
This has always seemed to me as a "doesn't understand one way streets" issue. Uber seems to offer the ride to the geographically closest car first, rather than the driving distance closest. While Uber may (rightfully) want to avoid computing a shortest path between you and the driver, it seems silly that they don't do at least something with one way streets.
Are you sure that Uber offers the ride to the closest (geographically) car? I've noticed that most of the time when I see close cars, they are not the ones that snag my ride, which can be frustrating when there are multiple cars 2 minutes away and the car that wins my business is 10+ minutes away (this has happened multiple times to me). The map doesn't seem like it understands one way streets, but it also doesn't seem to prioritize by geography either.
As Uber grows in a city, the average wait time should also decrease pretty dramatically. There are more cars on the road, so a pickup from any given location/time should arrive faster on average than it would two years ago.
The precise methodology is opaque in the post, but there's a chance that these results are driven more by increasing liquidity in the marketplace instead of increasing expectations of Uber over time.
Honestly, I believe the data, but not the conclusion. When I first started using Uber, it was a taxi replacement. I take taxis once every few weeks e.g. to go to the airport and similar. Most of my other transit was either by subway, walking, bicycle, or didn't happen.
When Uber dropped to a couple of minutes, my usage pattern completely changed. It's a casual form of transport. Hungry? I'll take an Uber to where ever I want to eat. Running a few minutes late? I'll Uber it instead of biking. I now meet people more than I used to. When transport became fast, cheap, and convenient, I started using it a lot more.
If I see a 15 minute wait time on an Uber, I'll usually cancel it, but that's not because I'm any less patient than I was a year ago. It's just that Uber now fills a niche in my life that was unfilled a year ago, and indeed, that I didn't know existed. When Uber jumps up to 15 minutes, my usage patterns go back to what they were with taxis, and I cancel most of the casual trips.
It depends on where you live. In New York or Chicago, you just hail a cab, usually in a minute or two. Pulling out an app and trying to get the marker in the right place and waiting 5 minutes is a step backward for the kind of usage you describe. An app is a very clumsy replacement for street hails.
For me this goes doubly so when I happen to be walking somewhere (in NYC) and see a cab driving past with its light lit - I might have normally walked the 15 blocks, but since there a cab right there, and its available right now.... its almost like an impulse-buy on the transportation.
How much does competition account for the "impatience" that Uber's seeing? When I launch Uber and see a "long" wait time, my immediate reaction is to launch Lyft to see how long their wait time is. Not that Uber would go sticking info about their competitors in a blog post, but you wonder if there's any connection between the "patient" cities and how much mindshare competitive services have in those cities (i.e., no competitors outside of legacy taxi companies = people willing to wait longer).
"The bottom line is that we realize we have to continually raise the bar, to get you home from the bar. We are aware of this challenge, from both a technological as well as people perspective, and are doing what we can to fulfill our mission of bringing reliable rides to all who want them."
Yes, because it's really critical to get home from the bar in a state of total inebriation =)
Also, I didn't know giving reliable rides was Uber's mission. I thought it was "evolving the way the world moves." This may sound like nitpicking but if you pause for a moment, the difference, as far as a company mission goes, between "giving reliable rides" and "evolving the way the world moves" is pretty big.
"reliable rides" fits with their stance on surge pricing -- they want to be a service that can always get you a ride, as long as you're willing to pay the price (not saying this is good or bad).
Couple of questions about the post (since the link above doesn't seem to allow comments):
1. Were the numbers for the chart "Willingness to wait in a city 2013 v/s 2014" generated based on estimates from the Uber app or real wait times that were logged once the customer was picked up?
2. Also, looking at the same graph, it seems some of the rides that were not completed because the ETA was < 4.5 minutes (Probability=1) could have trivially been completed had Uber just dispatched the driver a little later (or had the driver delayed him/her-self a bit).
3. What is the volume of rides as a % for each range of waiting times for the ETA?
4. What is the revenue per ride as a % of total revenue for each range of waiting times for the ETA?
Notable point: This blog post was authored by one of Nathan Myhrvold's[1,2] sons.
[1] Yes, that Nathan Myhrvold. Microsoft billionaire(?), founder of Intellectual Ventures, creator of the best cookbook I've ever seen, and notorious patent troll.
Could one of the factors be the difference between users?
Early adopters who really liked the Uber concept for various reasons and wanted to use Uber
VS.
Now a majority of users who heard about Uber because it's supposedly better than other means of transportation , but when they try to book it and see a longer waiting time than just getting a Taxi, or taking a bus/train , they don't want to wait for a Uber car.
Is it just me, or does it seem short sighted to end iOS6 support now? I tried to use the Uber app on my phone the other day and ended up having to call a taxi.
Obviously you have cut your losses with legacy stuff eventually, but surely it dosen't cost much for one more iOS target.
Off the top of my head, iOS 8 adoption is currently like 90%, iOS 7 adoption (before 8) was at like 95% at it's peak.
So current iOS 6 usage would be less than .5% of iOS devices? I would suspect (without any sort of proof), that Uber's demographic doesn't overlap much with people running a 2 year old phone OS.
Let me explain...
When Uber was first in Chicago it was normal to wait 10-15 mins for a ride to pick you up. You would pull the app out while you still had a few things to do in the apartment before you left, order the Uber, and then by the time you were ready it was right out front waiting. The time waiting felt very seamless, although there might have been an actual 10 minute wait. When it also first started drivers were totally okay with waiting a minute or two for you because the demand wasn't as high back then. They didn't want to lose a potential customer. However, now that Uber IS more abundant in Chicago you can no longer do that little technique. Most drivers will get pissy with you if they have to wait outside for more than 30 seconds, or if you are not curbside waiting for the ride when they show up. Since drivers now also have more options for people to pick up, they can be more choosy on the customers they decide to take. They can easily cancel rides if they are too far from the pickup or are waiting to get a new customer. So since you can't "queue" up a ride like the "old days" you have order the ride when you are actually ready to go and are already waiting. Now that you do this, the wait is felt 100% by the customer (which was once avoided), if that makes sense. So what used to feel like nothing now feels like you are waiting in the cold for possible 5-10 whole minutes, which obviously often feels a lot longer that. Thus making you more likely to cancel the ride if you see an empty cab or another opportunity. So there is a bit of a catch 22 here, if I'm using that phrase correctly.
It does makes sense that they don't want to expose this side of things though, because that places more blame on their own quality of drivers. It probably shows that as a city is active longer, the ratings/quality of drivers goes down drastically. I've noticed it quite a bit here in Chicago, and it's really disappointing. The average quality of driver I get nowadays is almost 1 to 1.5 stars lower than what I got 2 years ago.
I'm curious to find out if anyone else has had this same experience or not.