Death of the taxi medallion: SF cab company ponders major change | Hacker News Reader

Death of the taxi medallion: SF cab company ponders major change

sfexaminer.com

81 pointssmalter64 comments

You gotta give Hansu Kim credit, he's doing something different. A lot of companies would complain about Uber and die a slow death, but this guy is like "if they don't have to follow the rules, neither do I."

It shows dash.

> "Uber isn't putting me out of business," Kim said, speaking hypothetically. "What Uber is making me do is retool -- and given the same rules, I'll beat them all day long."

I appreciate his spirit, but I don't think he's fully estimating what it would involve to beat Uber at its game. He'll have to hire more drivers, have lower prices and still be up against a huge network effect wall. He seems jazzed with Flywheel, but they could stop sending him business tomorrow.

One of my favorite parts about Uber is that it's in so many cities and has consistency (easy to see pricing, if there's a problem I know who to contact, etc). I'm never going to seek out DeSoto Cab Co on purpose.

Basically, if the rules tilt towards the Uber model there is no reason for Kim to exist. He's a middle man in a business that is going the opposite direction.

Right, his thinking sounds a lot like what we heard from 2000-2008 or so from the newspaper industry. There's always some new app they think might turn the tide to conveniently mention in interviews, but at the end of the day these taxi companies are all on their way to a different, lower station in the marketplace.

> Basically, if the rules tilt towards the Uber model there is no reason for Kim to exist. He's a middle man in a business that is going the opposite direction.

Not at all. Maybe though, he's a local middle-man in a business which is going toward global middle men.

After all, Uber and Lyft are nothing if not middle men.

Google is just a middleman between you and webpages, Facebook is just a middleman between you and sharing your stuff. Everything is a middleman.

Downvote me all you want, but I just realized is even more true than I through at first, because the main revenue model of Google: Adsense/Adwords, is just a middle men between webmasters and advertisers; the only reason those two groups don't talk directly to each other is because they don't trust each other:

- They don't know each other (lack of contact method)

- They don't know how much they charge.

- They don't know if they are trustworthy.

- They don't know if they will maliciously overcharge or if they won't honor the contract.

The completely exact middleman work that Uber does between drivers and passengers.

I also think that's not always a dead end. There is also an opportunity to grow. For instance, I know Uber exists, but since I only take taxi occasionally I don't own an Uber account.

In NY, the Chinese and the Korean community have their own network knowing how to reach the Chinese/Korean-speaking car service and shuttle service. Uber or alike cannot beat this system. These locale-specific service operate with just telephone. Very school but effective! No app, no Internet access. Pretty reliable and cheap.

So all he need to do is to find a community to target at and find drivers willing to join him.

Translating their app would allow Uber to encroach on such communities. That is likely already in progress.

I will just make my raw observation and prediction that the reach into certain community for certain use case is likely to have little success. For example, if you want to travel between Flushing and Chinatown, or between Brooklyn and Chinatown, or between Queens and Chinatown, there are specific "bus station" for people to wait at. For car service it's also a single phone call and you can get a quote within half a minute. There is usually a single number you dial and you can get a quote on the ETA so you can wait at home and go on the shuttle van when the time comes. The reason technology hasn't beaten this old school system is because of the low cost to run and the low entry to barrier for both the business owners and users. Typing can be quit slow and ineffective and sometimes Internet access is just not available.

It's basically a community lock-in. So a Uber-like clone will have to break this "monopoly practice" not by just translating the app, but actually knows how the community works.

Uber is translated into a ton of languages already, I see that it gets updated very frequently and often times the only changes are various localizations.

Considering you can put your destination address in the app, it should be a huge improvement to anyone who doesn't know the language (or with disabilities).

What do you think Uber is but a middle man?

What do you mean by "Network effect"?

I don't really see how it applies in this situation. (e.g. you can use uber today and a taxi tomorrow and suffer no loss of value).

I think "mindshare" was the term he was looking for.

Uber already has a ton of users and more importantly, drivers. If I visit SF what reason would there be to use this guy's cab service vs Uber? I already have Uber on my phone.

Uber has some drivers in a city, and some passengers - not very valuable. But, Uber does something special, like drop prices, and all of a sudden, there a lot of people who hear about the low cost of UberX, and so they put the App on their iPhone/Android, and start calling vehicles. Drivers, hear about the incredible demand for UberX drivers, and start adding that network to their vehicle. Passengers discover there is always a Uber vehicle close by and so they start using it more often, which creates more demand.

Eventually, you get a network of users/drivers - each new driver is available to many customers, and each customer will make themselves available to many drivers.

It's not a perfect N(N-1)/2 mesh, but then neither were fax machines (not everyone wanted to fax everyone else) - but the same concept of network value exists.

What Uber has going for it is convenience, transparency, and quality. By the time the industry itself settles, who's to say Uber could keep up the high standard of quality? Or if they move toward a monopoly, they might hike the price, allowing smaller, local-focused services to differentiate themselves and flourish.

We have no idea what the future hold, and remember Uber is the new kid on the block here.

I had a friend visit SF from NYC a couple weeks ago, and it blew my mind that he continued to use taxis instead of Ubers. Here, the decision seems clear-cut: calling a car with a phone is cheaper, faster, and more convenient. However, in NYC the streets are so saturated with taxis that waiting 5 to 10 minutes for an Uber instead of hailing a cab immediately seems absurd in most day-to-day situations.

It's pretty much the same in Chicago. You can hail a taxi almost immediately, but uber will take you 5-10 minutes. Can be useful though if you're in the outskirts of downtown.

Yep, I usually wind up using Uber to get from my apartment to the bars, and then there's enough cabs in the street outside the bars that it's faster to just hail one to get home.

Ever tried hailing a cab in NY? I got tired of wasting 15-30 minutes and prefer Uber instead.

Summary: given the current regulatory conditions, you don't need a medallion if you just take bookings and do not get hailed from the street. "Taking a booking" is interpreted more broadly each day, so it includes summoning via apps. The owner, Kim, estimates that giving up the hailed fares in return for not paying for medallion rentals, would let him make more money and outcompete Uber.

My comment: How long until someone converts a street hail into a booked ride by generating a record to make it look booked?

I'm wondering if a "kiosk" in key locations on the street to grab a cab would be considered a booking or a street hail?

What if Uber just gave bar bouncers an app for hailing Ubers and shared a small cut with the bar? They'd just need a Square-like credit card swiper so that customers could pay for the rides individually.

You could even integrate it into the bar tab, given a destination ahead of time.

I think the bigger market is hotels. Most major hotels already have a line of taxis outside.

This is kind of what we had here outside some nightclubs in the UK. There'd be a private hire company (bookable only), with a guy with a notepad and phone standing outside, he'd make the booking on the paper, ring the driver who was in the car next to him, bingo.

Has the market value of taxi medallions fallen after the advent of Uber and Lyft? I think there is a big risk that this way of organizing things is about to go away, which will leave taxi medallion owners with a huge loss.

http://www.washingtonpost.com/blogs/wonkblog/wp/2014/06/20/t...

I don't get it - he's giving up the ability to pick up riders at the cub, so where will his riders come from?

Or does this loophole mean he CAN still pick up at the curb? I did not understand the whole TCP vs TNC business.

Not sure if this is full answer, but from the article:

> Kim said he would do just fine because he has built a loyal customer base over the company's 80-year history and now receives an additional 2,000 hails daily through the hailing app Flywheel, which he hopes would adapt with him.

Thanks, my skimming was poor.

xsmasher - good question :) presumably his rides could come from apps like Uber and Summon (disclosure: founder). I will say that as someone whose platform supports taxi, tncs and tcp cars there is a good chance this is just posturing by the person mentioned to achieve some other goals.

Have you ever tried to hail a taxi in San Francisco? Can't be done anyway.

Can't get one dispatched, either. Uber had a blog post a few years ago [1] where they highlighted some findings from a SFMTA report on cab availability. The report found that 35% of dispatched cabs never showed up, except on weekends, when 72% never showed.

[1] http://blog.uber.com/2011/04/11/uberdata-the-hidden-cost-of-...

Wow those are some horrible numbers. Am I just lucky that cabs in Seattle are reliable? I've never had a no-show, and cabs have always been on time here +/- 5 minutes of the requested pick up time.

Then again my sole need for cabs has always been "I need to go to the airport at 4:30am", a need for which I imagine Uber would not be appropriate.

cabs in sf are an utter shitshow. Uber's original innovation was -- and this is only remarkable in sf -- they where a taxi that came.

I was on crutches for 9 months and took cabs daily to work from near mission and 23rd. All these cab companies are utter assholes: I booked cabs 3-5 days a week, from a population hub, and tipped well, but none of them could provide decent service. Each one would say "10-15 minutes" when I would ask how long the cab would take to arrive. I collected stats for a couple months: the median time was over 30 minutes and the 90th percentile was 1.5 hours. I even had a cabbie come more than 50 minutes after promised; I'd already begged a friend to drive me to work. Dude had the big old brass balls to call my cell and bitch that I hadn't cancelled, after I'd waited over three times their maximum arrival estimate. One dispatcher even admitted on the phone that they just tell everyone near where I live 10-15 minutes.

Uber in other cities may have a harder row to hoe, but in sf, merely being a cab that comes (or -- and this is amazing -- treating your customers with a tiny bit of respect and telling them you won't come if you can't) gets you really far.

Fuck sf taxes -- yellow, desoto, luxor -- I hope all those assholes get unemployed. I'm not sure Uber is a good thing in a city with working transport systems, but that doesn't describe sf. They made my life even harder than it already was; there's no good reason I should have had to call cabs more than an hour and a half before I wanted to be to work (an 18 minute drive) just to have a greater than 90% chance of making it. And when you have doctors appointments or surgeries or something else with a hard start time, I have no idea what you're supposed to do in sf besides use Uber. Hire a black car maybe?

Wow that is terrible. In Seattle, I use an online reservation system for Orange Cab, and they just show up on time the next day!

For Yellow Cab, well, they have a phone dispatch still, I actually switched over to Orange Cab just because Yellow's dispatchers are rather rude.

You may still want to check out Uber... I just checked my last receipt from downtown to SeaTac and it was $23. Orange Cab is $40 without a tip. In town is quite a bit cheaper as well.

I pay about ~60 from Kirkland to SeaTac. Insane rate, but it isn't fixed rate since I am in Kirkland. Costs me 50% more for about 15% more milage, and since it is 405 south bound it is likely a quicker commute anyway.

But again, I doubt Uber runs at 4:30 am in Kirkland!

The per mile rate on UberX is lower, but yes that is a tough time for anything that's not booked ahead of time. The way home could work though, it's easy to get picked up at the airport.

I bet Uber gets there though, I could see advanced booking being added before long.

I think it's better in Seattle. I lived there for seven years and I can't recall a cab ever no-showing like I experienced when I lived in SF.

Basically he's paying $2,200 per month per car for the ability for a car to pick up from the curb (and SFO apparently). He thinks he can not do that and instead rely on fares that were hailed through an app (or dispatcher I'd assume). He'll have lower revenue, but also significantly lower cost.

What I think might be more interesting for Mr. Kim is a hybrid model. He could start shedding medallions, and only use them for the airports or potentially a downtown core.

The rest of his dispatched traffic could be in vehicles without a medallion. If most of the rides don't require the medallion, he could shed a significant cost. It sounds like with 204 of 2000 medallions, his decisions could personally have a fairly large effect on the marginal cost of the medallion (so he is paying less per medallion and for fewer medallions overall).

Interesting that DeSoto fails to mention why people seek out Uber. Kim seems to be focused on the medallions pricing structure completely ignoring user advantages such as ease of access through the app, reliable service, quality, etc, etc.

If cab companies don't realize that new services beat them to the punch in many areas besides pricing they won't survive.

That is, unless, they have a government-granted monopoly that is tremendously unfriendly to disruption (i.e. regulatory capture)

has anyone tried http://www.ztrip.com/ ? it's a car hailing app developed by the company that owns a bunch of yellow cab companies. curious what people's thoughts/impressions of it are...

Interesting name. Ztrip is also an incredibly popular remix artist/DJ.

http://en.wikipedia.org/wiki/DJ_Z-Trip

nice! love old school hip hop dj'ing.

still wondering if anyone has any opinions on the ride summoning app... maybe that's the answer, that no one uses it or has an opinion?

I wonder if we'll see city medallion prices fall as drivers like this choose to not buy/rent medallions. Once that happens, the math might change such that it makes sense for him to keep the medallion.

Sucks to be an investor who bet who bet against technology, ouch!

This won't work everywhere that has a medallion/license system. In Australia taxi licenses are regulated by each department of transport at the state level rather than city/municipal councils.

interesting that right before (a year or two) Uber/Lyft/etc...'s time the prices of medallions hit something like $200K/$600K in SF/NY and there were stories about that in the news. Speaking about dialectics - top blossom of anything carries its undoing with/inside it.

They have cost a fortune for a long time, Uber was born more out of the frustration of the ability to actually get a ride than whatever expenses were involved (though to be fair the reason it's hard to catch a cab in SF is because there are so few medallions which is why they're so pricey).

Taxis are going the way of video rental stores. I wonder what is going next.

Hotels?.. Travel agencies?.. Car dealerships?

There are too many options to know which one is next.

Please let it be car dealerships and mattress salespersons. (crosses fingers)

The big one is going to be schools. They won't die out, but they will be substantially altered and not as dominant as they are today, and that process will take much longer than the uber vs. taxi dynamic will take to play out. So much of modern education is just read, regurgitate, test, even much of a typical undergraduate education falls in that bucket. That sort of education is prime for disruption via interactive educational apps, but the genre hasn't hit it's inflection point moment yet. Nevertheless, sometime in the next 10 to 20 years or so there will be a breakthrough in educational software. Someone will put the pieces together and start making stuff that is universally considered to be excellent, and they will start making a crap ton of money which will fuel a gold rush into making apps using the same ideas. The rest will be history.

Url changed from http://pando.com/2014/08/04/san-francisco-cab-company-admits..., which points to this.

Reply on news.ycombinator.com