Hacker News Reader

Comment by neonate | Hacker News Reader

This seems to me the salient part of your comment:

> Taxis are mostly only profitable with extremely low overhead and only servicing high traffic areas like Manhattan or other major metropolitan areas. Uber's thousands of engineers and SF HQ disqualifies them from being low overhead anything.

If I'm understanding correctly, the argument is that taxis are basically a shit business—bad enough to break the "software is eating the world" rule because they have such low margins that you can't pay a software team (or at least, not operate a marquee SV software company) and compete. Better tech isn't enough.

If that's true, that answers my question. I'm surprised if it's true, though. I didn't think the taxi business was that miserable, and I'm not sure how it squares with taxi medallions being the highest-performing investment of the 20th century (or some factoid like that which I read years ago).

View on HN

When he puts it like that, it makes me think Uber has a much better business model than I thought. Taxi efficiency is very heavily dependent on information - if you know exactly where the customers are, where the drivers are, and then can run shortest-path algorithms in real-time to match them up, you can save huge amounts of driver time waiting and finding customers. Figure you save an hour a day * 200 workdays/year * 200K taxi drivers in the U.S. and that's 40 million hours saved in the U.S. a year. At a rate of $10/hour, that's $400M/year. Then add in economies of scale from knowing where all the drivers and customers are, and a better brand and better customer experience from reducing wait times and transaction costs to get a taxi.

Is that worth $80B? I don't know, but I'd guess probably not. But it'd easily pay for a SF HQ and several thousand engineers.

Yes informational efficiencies for transportation at scale could easily be $80 billion worth of value to the world. The real question is whether Uber can capture that amount of value. And whether competition will drive the margins to zero anyways.

Or at least several hundred engineers. Does Uber really need several thousand engineers? It's kinda nuts to me, WhatsApp scaled to a billion with 50, you figure an order of magnitude more to deal with the fact that a large part of Uber is turning legal compliance into code.

When you read their stories about managing 10,000 microservices or whatever.. I don't want to trivialize their domain but it seems like a bit of empire building happened.

Not arguing that there isn't probably a level of bloat at Uber, but the complexity of Uber's product is a lot more than other commenters are implying here. Their ML platform alone seems like a monster: https://eng.uber.com/michelangelo-machine-learning-platform/

I think they have their own ML platform because they hired a lot of engineers, not because the business necessitated one.

I know of no city in the US (although there may be some small towns) where taxi rates are not dictated by either state, county, or local laws. You cannot get a taxi for any rate other than the officially-sanctioned one. Unfortunately, this rate seldom has anything to do with the cost of gas, maintenance, depreciation, pandemics suddenly causing people to use more / less services, etc. This means that for the taxi industry, there are four chief ways of making more money (1) lobbying, which is effective but slow; (2) reducing costs in the extreme, which is also effective, but only to the extent that you are able to reduce your costs further by delaying repairs, doing it yourself, not offering refreshments, not providing great customer service, not developing apps, etc.; (3) working more hours, often in contravention of labor and safety laws. Also effective, but only if you're independent. (4) Breaking the law. A significant number of cab meters are illegally altered to charge more money than allowed by law.

Uber, for all its flaws, provided a service that was better than the taxi companies simply because they didn't require their contractors cut corners, cheat, or provide poor service to make money. People flocked to Uber because it was genuinely a better experience than the local taxi companies. I think in the long run, Uber will compete by becoming as expensive as (or more expensive than) local cab companies, and having a much smaller customer base.

> (4) Breaking the law. A significant number of cab meters are illegally altered to charge more money than allowed by law.

Do you have evidence of this claim, at least insofar as U.S. taxis are concerned?

Certainly the same method used by some unscrupulous gas stations. They added a device to the pump that electronically altered the cost i.e. skipping/adding x cents after y amount pumped. More than one Canadian station owner was nailed a few years ago for adding these devices to their pumps.

My google-foo skills haven't found the relevant story tho.

For taxis the device is called a "pulse multiplier" or "pulse divider". You can make them yourself if you have some basic electronics knowledge.

See this? http://www.taxicabelectronics.com/my_first_taxi.htm

They'll sell you a new taxi cab meter "preprogrammed with your rates". You swap the legit one out yourself. Install this bad boy. Charge whatever you feel like. When it's time for the annual inspection you swap it out for the regular taximeter and you're good to go. Taxi cab drivers share other tips like replacing the tires for k-value tests, swapping out the vss, or installing a "pulse divider" (Google taxi pulse divider / multiplier). Then there's good old-fashioned charging the wrong fare. Many cities have special fares depending on time, pickup / drop-off location, and cabbies will try to "accidentally" charge someone the wrong fare. I've even had taxi drivers show me their "special fares" programmed into their taximeters.

I present no evidence. I have been charged the wrong fare entirely too often in the days back before Uber. It's a real event that happens a significant proportion of the time, and that's not even to touch on the subject of pirate cabs.

This argument is akin to claiming that the availability of sawed-off shotguns implies that a "significant" number of people are buying them and shooting people with them. It's not a substitute for statistical evidence of use and lawbreaking.

Besides, "significant" is a classic weasel word - it could mean one, for all we know.

Yes. I weaseled so hard I literally said "I present no evidence." Thank you for pointing that out.

That made me laugh. Thank you for being a good sport. :-)

If we suspend the Uber/taxi comparison, and assume Uber is a new taxi-like service but better, therefore deserves a premium? I know I’d pay for that premium. Not everyone will, but I can’t be alone?

So that means the addressable market shrinks. Think premium black-car-only service; Uber’s roots). Why can’t such a service find demand/price equilibrium? I think it can, but not in a manner that can justify Uber’s “unicorn” valuation. And I think that’s at the core of Uber’s conundrum/identity crisis: it provides value, but with tech-company cost structures. However, it doesn’t deserve tech-company valuations because it’s just not as scalable — it’s tied to operating in the physical world, with physical constraints.

If the strategy is to burn cash, continue entrenchment, and eventually leapfrog to self-driving Ubers (mitigating the scaling/cost problem to an extent), then that’s a war of attrition. Fingers crossed.

What's premium about Uber ? My personal experience is that taxis are usually much more comfortable cars, they're allowed to drive in dedicated bus lanes, and they don't try to scam you by taking a longer route to rack an overcharge. The advantage and innovation of Uber is the app, but now taxi companies have one too. So what is it ?

My personal experience is taxis are less comfortable and I've only not been scammed by a taxi once (in the us), taxis are fantastic abroad, no idea why.

A large part of Uber is also the rating model. I don't know if taxis have that, but at least for LGBTQ passengers it makes a huge difference. You can't harass your passenger and keep a good rating.

One advantage I get from Uber is a single interface no matter where I am. I travel a lot. I won’t install an app for every city. Uber has taken away the pain of dealing with taxi companies in strange cities.

Unfortunately, it traded one kind of monopoly for another.

Ideally, Uber would have been an app and a network that taxi companies plug into.

Like Hailo?

This product existed, and was profitable until Uber drove them out of business with rides subsidised by VC money.

Taxis absolutely scam you and its harder to prove than an Uber with a digital record of the route taken.

Not so sure about the prevalence of longer route tactic. I’m sure it happens, but with the route plainly visible on the map, it’s hard to cheat here unless the driver is being a bully, which is a different class of problem.

As others have also mentioned below:

- Uber is global, and has a standard interface. I don’t have to fiddle with credit cards and the payment dance. It’s the de-facto global “taxi”. Kind of like the warm homey fuzz one feels at a Starbucks in a foreign land after enduring burnt coffee served in a styrofoam cup on a long flight. (Granted, a bourgeois American-centric view, but I hope it’s illustrative.)

- If you pay a premium (Uber Comfort and higher), the cars are better, the drivers are better (higher rated too), and one can communicate preferences such as temperature and conversation (no, thanks).

- Other long-tail uses: 7 passenger SUVs for large groups, lots of luggage, etc.

- I’ve tried using the taxi apps (Flywheel, etc.), and they’re atrocious/geographically limited. In some places the Uber app can now call taxis. Even then, I’ve never voluntarily wanted to call a taxi for the last decade or so. I especially like the Black Mirror vibe of drivers and passengers having a “rating”. ;)

Better cars and bus lanes aside, I've never seen an Uber driver do anything besides follow whatever route Google Maps, Waze, or Apple Maps suggested. Most of the drivers don't even really know the streets of the city they are driving in.

My complaint would be the exact opposite of yours: taxi drivers often have a few tricks up their sleeve while Uber drivers will follow Google Maps straight into a traffic jam.

What's premium is that you're faced with a better human being. You can easily tell when your driver is a random dude or a former taxi driver. Also, taxi drivers try to claim their credit card machine is broken. They'll tell you one fee then ask you for another at the end. They'll snatch the money out of your hand as you argue over the cost. It brings out some of the worst behavior in human beings.

Uber is a hassle-free experience and saves you from dealing with the one miserable human being you'll encounter that day. Stress has a price.

Drivers are rated and you can cancel if you're handed a bad one.

Maybe I'm missing something, but how would they rack an overcharge? From my experience I've never paid more than the fare calculated before the ride is even accepted by a driver. I haven't heard of any cases of overcharge, or seen any traces of dynamic pricing after the ride has completed, save from tips. Maybe I'm just lucky?

i have no idea what that guy is talking about. the price of an uber ride never changes after youve ordered it

I've had an overage charge once. The driver hadn't selected in the app that he'd dropped me off, and drove for about an hours afterwards. I messaged Uber about it when I saw the overage charge, and it was refunded about 30 minutes later.

It depends on where you are. Germany shut their attempts at ignoring all laws down quite quickly, their Uber Black service is just like any other cab, just with an app instead of a phone call.

They managed to drag our cab companies into the 21st century without even being active almost anywhere, so thanks for that ;)

But in South Africa? Taxis are extremely expensive cars that often look as if they’d fall apart any moment. Uber was/is cheaper, but I’d have taken them even otherwise. Most certainly a premium service.

Those taxi medallion returns were a con from small groups of insiders.[0]

[0]https://www.nytimes.com/2019/05/19/nyregion/nyc-taxis-medall...

Taxis are both a mediocre margin business, and typically cartel restricted. Many people here are likely familiar with the absurd stories of the NYC medallion prices from just before Uber took over (a million dollars for a taxi medallion, before Uber smashed the cartels, tanking the value of the medallions). NYC extracted $855m in medallion fees from just 2004-2014.

So how did the eg NYC medallions become so valuable? The taxi cartel bought into the (mostly unspoken) notion that the medallions represented a government guaranteed blockade on competition. For a city like NYC with rather epic taxi demand, a license to print even just 8% margins perpetually can be an enticing prospect (especially in an age of forever low interest rates). Then, instead, the city governments mostly allowed Uber and Lyft to continue to operate (after their initial move to operate first and seek forgiveness second), rather than shielding the taxi cartel from competition.

> I'm not sure how it squares with taxi medallions being the highest-performing investment of the 20th century

Taxi medallions are an artifact of how NYC regulated its taxi fleet; other places may have had them but the investment factoid probably comes from NYC. Joe’s Cab in Peoria, IL probably doesn’t have a medallion worth $1 million to borrow against. And, of course, note: not all cabbies owned (or currently own) the medallions they drive under: many lease them or drive for a company that owns them, and has sat on them as a speculative asset for decades.

Source: went to school in NYC with plenty of cabbies’ kids. It’s not a job many people eagerly take.

Drivers and vehicles are by far the biggest expense in operating a taxi service, not matching, ordering, and navigation. The only way Uber was ever going to make sense was if software could eat the drivers, but they effectively gave up on that and it's looking like it won't happen any time soon. We've already long had "better tech" for moving large numbers of passengers between popular destinations in a city: trains and buses. But they're still not profitable and require public investment. Uber as a public service could probably work, but then no one could get rich transferring wealth from late-stage investors to early-stage investors.

They were in Manhattan. They weren't in SF, and they certainly weren't in, say, Urbana, IL.

You mean about medallions? That makes sense.

Have I characterized your argument correctly, more or less? If not, I'd like to know what I missed.

I think you got it. Pay a million dollars for a medallion, and then have the right to sit at Times Square or JFK and charge people $50 to drive them for 20 or so minute. The cabs are independent LLCs, so they're on their own for health insurance, 401k, etc. Mostly driven by immigrants, I think being a cabbie can be a tough job.

> taxis are basically a shit business

In efficient market ANY business must be a shit business. If you have high profits -- competitors should squeeze into your profits.

Same logic works for other low-margin businesses, e.g. restaurants, they just saturated their supply.

Reply on news.ycombinator.com