4 ms·
I've gotta think one category he's talking about are the delivery services. I still have not seen a decent articulation of the end-to-end economics. But they ca
by pbreit 11y ago
I've gotta think one category he's talking about are the delivery services. I still have not seen a decent articulation of the end-to-end economics. But they can't be terrific.
I know, for example, that Sidecar pays drivers up to $20 for a single delivery and just can't see how this is viable outside of rare exceptions. At least Instacart prices the item being delivered (and I suspect never gives you the yellow ticket price which pretty much every Safeway item gets nowadays).
- exelius 11y agoI agree; delivery services will only make sense with autonomous delivery (which I suspect is the long-term plan for all of them - Uber included). But realistically, the R&D required there is crazy if you're trying to be first to market - Amazon and Uber are investing billions in this, and they're the competition. Once drones or autonomous cars become widespread, this business becomes a commodity and anyone without a legacy business they have to migrate will be at an advantage. So I have no idea why these companies are getting funded other than it's a problem that's simple enough for a really smart university student to tackle.
- _sentient 11y agoUber is profitable in many of its mature markets. This is while charging less than half what you would pay a legacy taxi service, while paying drivers more than they would make at the same. This is the direct byproduct of strong network effects on both the supply/demand side, paired with operational excellence on their part. Self driving cars would bring the marginal cost down even further, but they have built a stunningly good business with or without a quantum leap in automation. The same probably isn't true for many other Uber-for-X companies, particularly those with underwater unit economics.
- exelius 11y agoYeah; Uber had a unique opportunity because the taxi business was stuck in the 20th century and had a bad case of regulatory capture as a result. That said, in Uber's more "mature" markets you're starting to see a lot of the problems that caused the taxi business to regulate in the first place: too many drivers, no way to guarantee steady revenue as a driver, etc.
- hueving 11y agoThe purpose of the industry isn't to provide steady revenue to drivers. That's a non issue that will be solved by drivers quitting when the price gets too low.
- pbreit 11y agoThat's not quite true. Regulation frequently aims to smooth out the booms and busts and I would argue that happens in the current taxi industry.
- BurningFrog 11y agoRegulation is marketed by nice sounding goals like that, while it is actually motivated by protecting the established players and keeping competition out.
- Daishiman 11y agoYes, all regulations, all the time, and so much evidence to boot!
- pbreit 11y agoEither way, it smoothes out the booms and busts.
- hueving 11y agoThere isn't really any correlation between smoothness of booms and busts and regulation. Quite frequently regulation causes massive swings because it prevents the market from pricing risk correctly (e.g. The affordable housing act leading to banks not needing to care about mortgage payback risk).
- lifeisstillgood 11y agoUber is profitable in many of its mature markets. This is while charging less than half what you would pay a legacy taxi service, while paying drivers more than they would make at the same. What? Seriously? Are you saying that say the cost of a "medallion" is more than twice the cost of all other costs in a taxi service? Can you give any examples?
- wxs 11y agoWouldn't the statement also be consistent with an increased percentage of car time with a passenger? That is, if all other costs were equal, and occupied time per driver-hour doubled despite per-ride profit halving?
- DMac87 11y agoThis doesn't seem that unreasonable - weren't/aren't medallions over $1M in NYC, which is a factor more expensive than labor, fuel, or maintenance of a vehicle...
- exelius 11y agoYep, and they were all effectively owned by banks who leased them to cab drivers on a per-shift basis. Your cab driver has to pay ~$150 (I don't know the actual number) per day to drive a cab, so until he makes that $150 back he is losing money. If you had to do this every day, knowing there is a chance you might lose money by going to work, well, you'd be a pretty miserable human being -- just like most NYC taxi drivers.
- DanielDent 11y agoIn Vancouver (Canada), until recently the market price for a "medallion" was $800,000-$1,400,000. Taxis are often unavailable when you want one. A significant portion of the taxi fare does not got to the driver - it goes to the person they are renting the taxi/medallion from. Uber briefly entered the market using licensed limousine drivers. They were pushed out because we have a crazy rule that limousine rides can't cost less than $80-90 - the only reason for that rule is so that limousines don't compete with taxis. It's insane.
- Pxtl 11y agoUber's success has nothing to do with their technology and everything to do with the fact that they straight-up ignored the regulations on their industry by claiming to operate in a different one. The taxi-industry is overregulated and corrupt, so Uber created a "car service" that works exactly like a taxicab.
- cortesoft 11y agoIt doesn't work just like a taxicab. I can't hail an Uber on the streetcorner by raising my arm.
- Bjartr 11y agoI wonder if you could mount one phone in the windshield of an Uber driver's car that could recognize a flashing light pattern from a held aloft cellphone of a person who needs a ride, such that the backend system routes that user's request to that driver. So the actual 'reservation' is functionally identical to how it is today, it just so happened to be facilitated by line of sight. Although honestly this becomes legal masturbation as it's quite clear the intent is one of circumvention.
- jcater 11y agoIn most places, you don't hail a cab on the streetcorner by raising your hand. In most US cities, you call dispatch.
- ghaff 11y agoThe logistics industry has actually done a pretty good job of getting deliveries in the 2-4 day range (depending on warehouse locations) across the continental US to the few $$ range whether paid for explicitly or bundled. But routine same-day delivery even in relatively dense urban areas is still pretty clearly a premium service (and pretty much unavailable if you go 15 miles out of the urban core). Grocery delivery for $10 isn't that interesting to most people if they're in a position to hop in a car or walk 5 blocks. And single items like something from the hardware store don't work at all.
- avip 11y agoI'm in Mumbai, and can day-deliver pretty much everything from everywhere. I can order 5 eggs, 1L milk, or a hammer. Not sure how this system works though.
- ghaff 11y ago>Not sure how this system works though. I'm sure the basic answer is very low labor costs. There are lots of personal services available in countries with low labor costs that aren't nearly as affordable in the US.
- ismail 11y agoI think there are a few aspects to making delivery service economics work, marginal costs go down the more you can group deliveries and route better in areas. In other words, network effects. 1. Do I need it right this minute ( example pizza) 2. Is it purchased in advance or planned? Like weekly groceries 3. Are you just doing delivery, and that is your only revenue? 4. Will your earn a commission, transaction fee or add a markup? (i.e Instacart) 4. Are you going to follow an uber type model or integrated by hiring staff? 5. Can you group customers orders together and do delivery runs using optimization algorithms (tsp, vrp). Lowering your cost? Here is an example (Using ZAR values / 10 for approx USD values): Say you start a diaper delivery business. On average a parents will need it once a month. To make it simple we will assume 1 customer initially, and then scale up to 10. We will compare 2 scenerios. Here is the example in Google Docs: https://docs.google.com/spreadsheets/d/1fpdh9BPWJMqWXV-wWRl2W1hraWrBrl74ehAPHD5lpyI/edit?usp=sharing https://docs.google.com/spreadsheets/d/1fpdh9BPWJMqWXV-wWRl2... Cost of diaper pack: $30 Avg cost of delivery per KM: $ 0.40 Distance to Deliver: 20 KM Time to deliver & return to Hub: 1 Hour Cost Per Hour for Driver (South Africa): $5 Delivery Fee: $3 Retail of Diaper with Margin: $35 Scenario 1, No Margin on Diapers, with 1 order Total Cost: 30 + (20 * 0.4) + ( 1 * 5) = 43 Total Revenue: 30 + 3 = 33 Loss: $10 Scenario 1, No Margin on Diapers, with 10 orders For simplicity: Driver will drive to furthest point, and each point is 1KM away in direction of the hub. No change in distance driven. 1 Additional hour as more stops. Total Cost: (30 * 10) + (0.4 * 20) + (5 * 2) = 318 Total Revenue: (30 * 10) + (0.4 * 20) + (10 *3) = 330 GP: $ 12 GP %: 3.64 Scenario 2, Margin on Diapers ($5), with 1 order Total Cost: 30 + (20 * 0.4) + ( 1 * 5) = 43 Total Revenue: 35 + 3 = 38 Loss: $5 Scenario 2, Margin on Diapers, with 10 orders For simplicity: Driver will drive to furthest point, and each point is 1KM away in direction of the hub. No change in distance driven. 1 Additional hour as more stops. Total Cost: (30 * 10) + (0.4 * 20) + (5 * 2) = 318 Total Revenue: (35 * 10) + (0.4 * 20) + (10 *3) = 380 GP: $ 62 GP %: 16.32 % https://docs.google.com/spreadsheets/d/1fpdh9BPWJMqWXV-wWRl2W1hraWrBrl74ehAPHD5lpyI/edit?usp=sharing https://docs.google.com/spreadsheets/d/1fpdh9BPWJMqWXV-wWRl2...
- pbreit 11y agoThat does make some sense but I'm not sure if the assumptions totally work, primarily the "no extra miles" and "only 1 extra hour for 9 more stops".