Cross-posted from my Quora answer to: What are people’s experiences with Uber in SF, NY, or DC?
I got to use Uber (formerly UberCab) while on a short trip to San Francisco this past week and I am completely sold on this service! (When is Uber coming to Southern California???) My experience has shown that existing taxi cab operations are technological dinosaurs, and in the age of GPS, smartphones and mobile payments, having to deal with their backwards system is simply intolerable. It’s clear to me that Uber is a company that’s harnessed all these new tools to tackle hard dispatch/logistics problems and deliver Apple-type “magic” to their customers.
I’ve got quite the anecdote for an experience, so hear me out…
On the morning of my flight out of SF, I managed to squeeze in breakfast with a friend from college before catching a Caltrain ride down the peninsula to the airport. Breakfast was set for 9:20 AM and I had to be at the nearby station by 10:05 AM, so missing the meeting time with my friend was not an option!
It was 8:55 AM when I exited my friend’s apartment in The Castro. I had a cool 25 minutes to get to the far edge of SOMA near AT&T Park. It was much more time than I needed to get there with a cab.
the ride (direct trace by Uber)
Totally underestimating the cab-seeking problem that plagues SF , I thought it would take a minute to call up Yellow Cab and hail a taxi. 
How wrong I was.
Standing under torrential rain, I soon realized after the third minute on hold that this wouldn’t be a quick and painless task. I quickly scurried to an awning of a nearby bank, doing my best to avoid the rain and figure things out (didn’t help much).
the awning (Google Street View)
Five minutes on hold with Yellow Cab, still no word…
Then seven minutes…
“Aw jeez! My friend’s going to be there in 15 minutes! What do I do?!”
Finally someone picks up!
“Yellow Cab, how can I help you?”
“Yes I need a cab at 15th and Market”
“Oh… well, sir, I just want to warn you that we may not be able to service you.” 
“What do you mean?“
“We’re totally spread thin and we–”
<”ee ee ee” : the iPhone disconnected tone>
“Ohhh man!!!! What the hell?!? Damn you AT&T! Damn you Yellow Cab!! Damn you rain!”
So I call Yellow Cab again…
On hold once more, the minutes started passing again.
“OK screw this, I’m taking action!”
Thankfully, I had read Bradley Voytek’s awesome posts about slicing and dicing Uber’s usage data , and I recalled it that instant: MUST TRY UBER NOW!
I knew it would be expensive, especially for the non-optimal scenario I was in , but I had such little time and was sick of being rained on that I threw caution to the wind and gave it a shot. 
Despite the spotty service, AT&T had a saving grace for me: simultaneous data and voice connectivity.
As I’m still on hold with Yellow Cab:
- I went to the iPhone App Store
- Downloaded the Uber app
- Launched the app and signed up for an account
- Entered my credit card information
- Verified my cell number by way of confirmatory SMS text message
- Verified my e-mail address by way of confirmatory email and HTML page
- Inputted my pick-up location automatically through GPS geolocation
- Received a push notification that my driver was four minutes away
- Received a quick phone call from my driver saying, “Get on the other side of the street so it’s easier for me to get you. Please cross legally, sir!” 
Yes, you understood this correctly: I did all of this while still on hold with Yellow Cab. So…
- Yellow Cab is an incredibly horrible service.
- Uber is amazing, especially their quick and painless on-boarding process.
My Uber town car arrived within 3 minutes and I was able to make it by 9:25 AM. It was five minutes later than planned, but without Uber, I’d still be stuck on that street corner, drenched in rain, frustrated and pissed off that I wouldn’t be able to meet my friend I hadn’t seen in over a year.
trip details (from Uber's user dashboard)
Using the pickup time from the trip details above, I want to run down the chronology in more ordered fashion:
8:55 AM – Walk out onto 15th street
8:56 AM – Stop under awning and call Yellow Cab
9:03 AM – Yellow Cab dispatcher picks up phone
9:04 AM – Call disconnected, redial Yellow Cab
9:06 AM – Desperation kicks in, start using Uber
9:08 AM – Complete Uber download, sign up, confirmation, and ride dispatch
9:11 AM – Pick-up by Uber town car
9:25 AM – Arrive at destination, proceed to have breakfast
Whoa… wait a second. Did it take about 2 minutes to go from download to ride dispatch with Uber? YES! Amazing. I’ve become familiar with the on-boarding process for countless apps and web services and this was one of the fastest, most fluid, painless experiences I’ve ever had.
Yes, it cost nearly $30, but what’s the point of paying about half that amount for a cab that will never show up?
I hope my experience has made it clear that Uber really is the future of personal transportation. Right now, it only appeals to a certain type of user and income level, but over time, this product may scale into something far more versatile. Clearly the team behind it has done everything to give the service that type of “magic” Apple products have, and with skills like that anything is possible.
In conclusion: Uber is uber-awesome.
 The SF cab-seeking problem was something I should have been more cognizant of. The “taxi cab problem” was one of the first things discussed in one of my Berkeley economics courses to demonstrate how certain regulatory regimes are detrimental to consumer welfare. (Wait, what? At Berkeley? What about the Marxists? You thought wrong, pal!) Medallion regimes introduce price ceilings that lead to a gross undersupply of cabs, and cities like SF are deeply plagued by this.
I said this in the intro and it bears repeating: existing taxi cab operations are technological dinosaurs, and in the age of GPS, smartphones and mobile payments, having to deal with their backwards system is simply intolerable. It’s even worse that government regulation keeps these firms in power, completely distorting the market to the detriment of consumers.
For more information on the cab supply problem, see:
San Francisco Taxi Services: Why is it so difficult to get a cab in SF when you need one?
San Francisco Taxi Services: What’s the average wait time for a taxi in San Francisco?
 Flagging a cab down off the street wasn’t an option: there are too few to be reliable and it was raining hard (and I had no umbrella).
 Full disclosure: there was also the Oracle Openworld conference taking place that week as well, which was probably what accounted for the unbearable load on Yellow Cab. But still: there’s plenty of conferences taking place in SF (though most not as large as Oracle’s) and the cab companies need to be able to scale to meet these demands. Anything short of that is logistical failure.
 Bradley’s posts for Uber are written under the #uberdata category here.
When you look at Bradley’s data dissection, you can’t help but feel the same way Jack Dorsey must have felt in his earlier years, intrigued by local couriers and how to develop a short-form messaging system to help make them more efficient. What lies at the core of both Uber and the early Twitter product is the realization that dispatch is uniquely suited to disruption. It’s an interesting logistics problem with lots of application and potential, and I can’t wait to see what other startups like Postmates do in this space.
If you’re interested in the early Twitter story and Jack Dorsey’s logistics/dispatch fascinations, read this Vanity Fair profile of him.
 By non-optimal scenario, I mean to say that this particular ride didn’t conform to the optimal Uber scenario some friends had told me about. There’s apparently a “Goldilocks condition” where the distance and number of people sharing the ride make Uber economically on par with a regular cab. If you happen to be in that condition, there’s no reason not to use Uber.
 I’m also a big startup enthusiast and love using new products and technologies, seeing what’s great about them, critiquing what’s deficient. This was a perfect time to finally try Uber.
In addition, I was always interested in Uber because a friend and I had experimented with a similar concept using Twitter and its embedded geolocation feature for taxi dispatch. We couldn’t really scale well because we couldn’t see past the authentication/trust/verifiability problem. Apparently nobody liked our proposed names either: Twaxi, TweetCab, etc. To maintain product simplicity and “magic,” we didn’t want to force users through an API for a second layer of login credentialing. Uber went the route we refused to go, but hey, it’s clearly working quite well for them.
UPDATE: There’s apparently a company already doing this in London called Green Tomato Cars.
 Having gone to Berkeley, I’m majorly prone to jaywalking, as are many other Bay Area denizens. Thumbs-up to the driver for dropping this line, haha!
A note on the weather: in the interest of providing as much proof as possible about this whole experience, here’s a screenshot from Wolfram|Alpha showing the weather on that day (October 6, 2011). Note the precipitation rate during the 9 AM – 10 AM time: