Uber's virtuous cycle. Geographic density, hyperlocal marketplaces, and why drivers are key


(Andrew Chen) #1

Originally published at: http://andrewchen.co/ubers-virtuous-cycle-5-important-reads-about-uber/
Uber’s virtuous cycle Back in 2014, David Sacks (ex-Paypal, Yammer, Zenefits) tweeted the above diagram to explain why Uber’s geographic density is the new network effect. It’s an insightful diagram that’s been built upon by Bill Gurley (Benchmark Capital and on Uber’s board) in his essay How to Miss By a Mile: An Alternative Look at Uber’s…


(Ian) #2

Ok, here’s something I’m not getting about Uber and the investment coming at it - and that’s self-driving cars. You talk about drivers being the key, but let’s assume over the next 4-10 yrs we see self-driving cars (google, tesla, apple, whatever) egin to seriously disrupt the market, what place will there be for Uber? Whats your take on this Andrew?


(Todd Emaus) #3

@andrewchen In the “Earnings” charts by city is the y-axis ‘net earnings’ or ‘earnings by driver hour’?

My fear being that it may be the former, which means more earnings, but spread out across even more driver. Is there not a concern that the supply-side is getting squeezed? While I understand the marketplace-dynamics from a business perspective, I (at least in LA) see a growing user-experience problem in Uber’s supply side. The cool, interesting and most importantly safe drivers seem to be few and far between. Just one rider’s observation here…and as a utility perhaps the experience doesn’t matter much if at all. I’m not sure what you can share, but are these real concerns on the team or is the fire burning so hot that there’s only time to add more fuel to the fire?


(harry campbell) #4

Damnit who forgot the y-axis haha?

I bet it’s earnings by driver hour since that’s what they’ve provided in the past but either way, increased utilization means more expenses. So I’d love to see that same graph accounting for expenses and maybe a scale/label.

You raise a much more valid question though and one that I asked in a recent article. Plouffee is right in that there are millions of potential Uber drivers, Andrew just has to find them, that part will be hard but very doable. The real question is will Uber be able to maintain their user-experience. Uber can’t and espically won’t be able to legally provide training (with lawsuits on horizon) and with high churn and lower and lower paid drivers, that seems like a recipe for disaster. And it’s also not something that you would ever gleam from data points until it’s too late.


(Andrew Chen) #5

It’s a big risk, but also a big opportunity for us. That’s why we’re investing so much into it. But I think it’ll take a longer time than you think to get these cars onto the road and working well, there will be a long period where human drivers and self-driving cars will co-exist.


(Omar Albakry) #6

I think if the self driving cars come to full realty and people start to feel safe riding in an autonomous car it will open a totally new and big market for other skilled people who can maintain these cars and it will open a lot of job opportunities to get into the maintenance business which its not as a regular gas station shop ,probably will be a mobile car that have the essential kits to maintain and making sure that the autonomous car is ready for the next rider ,and since most of the people who drive for Uber now are not only driver by core most of them had other skills or can adapt other skills probably we will see a lot of drivers switching to be a maintenance partners with Uber , I thinks it’s a crazy thought but it will be amazing to see this happen .


(Andrew Chen) #7

There’s a footnote on the article that describes the y-axis: “*Based on average net hourly earnings, excluding service fee and Safe Rides Fee.”


(Andrew Chen) #8

The label on the graph indicates:
*Based on average net hourly earnings, excluding service fee and Safe Rides Fee.


(Ian) #9

I’m no expert in this, beyond reading pertinent articles as they pop up in my feed, but my understanding is very low maintenance on electric cars due to the lack of moving parts. In fact, there’s a surprising amount of “repair” that can be done remotely via software update. Think I’d read this in relation to the potential threat to garages (i.e. repairs).
Still, don’t quote me on that as you could well be right!
Do agree with Andrew re: the longer transition phase. It seems the tech itself is not far off - I think the challanges will be legal framework, around insurance, policing, social/cultural acceptance, etc. which will likely create the drag/friction to their introduction.
Personally, I’m (hopefully) confident in the next decade I’ll be jumping into my self-drive car, heading off to wherever, feet up on the back seat and smile on my face


(Will) #10

I think you’ll see a lot of car cleaning type jobs, but your basic maintenance is gonna go down as utilization of cars increases. Self-driving cars enable a bunch of things that are frankly not ready for prime time right now. Really good public transit is one. Not owning a car at all, but just using one on demand for another. Both of those things lead to fewer vehicles with a higher level of utilization, which probably means lower mechanical service requirements due to fewer vehicles on the road (but certainly more work being done on those that are still in service).


(Gen George) #11

This is an old article I am referring to. However do you think Uber will be able to continue to grow and sustain their marketplace while they wait for self driving cars with the continual massive increase in marketing and sales costs to drive the “more drivers” to move their virtuous cycle? Thank you