[1] http://www.businessinsider.com/ubers-revenue-profit-and-loss...

I would assume that would also depend on how much of their fixed costs they can shed in conjunction with that. Uber may not die quickly but it could still die slowly if they lose the ability to move forward because they had to cut too much of their R&D and sales infrastructure to get to that positive number.

Plus such a massive layoff would be a big red flag to customers too. By Uber's nature, a particular ride isn't a long-term commitment, but I think people would start looking for a stronger-looking horse even so. One of those differences between homo economus, who every time they need a ride rationally examines all their options regardless of the future and realizes that as long as Uber can complete this ride the internal state of Uber doesn't matter, and homo sapiens, who will take into account the fact that Uber doesn't seem to be doing well even if doesn't rationally matter at this particular point.

> someone is going to eat the valuation hit

Liquidation preferences mean earlier investors (and employees holding Common Stock) take the hit of a down round.

Let's say Company X has 9,000 shares of common stock outstanding. It raises $1 million at a $10 million post-money valuation with a 1x non-participating liquidation preference. Its cap table is thus 1,000 shares of Series A preferred stock on top of 9,000 shares of common.

It then raises $10 million at a $50 million valuation with similar preference terms. Its cap table is now 2,500 shares of Series B preferred stock on top of 1,000 shares of Series A preferred stock on top of 9,000 shares of common.

Let's contemplate a $100 million exit. Everyone converts to common at $100 million / 12,500 shares, or $8,000 per share. Series A bought at $1,000 and thus sees an 8x return; Series B bought at $4,000 and thus sees 2x.

Let's contemplate a $50 million exit. Everyone converts at $50 million / 12,500 shares, or $4,000 per share. Series A gets 4x; B comes out flat.

Let's contemplate a $25 million exit. Series B does not convert. Instead, it demands its 1x liquidation preference and gets $10 million. This leaves $15 million on the table, or $1,500 per share. Series A converts and sees its 1.5x return; B comes out flat.

Let's contemplate a $15 million exit. Series B does not convert and gets its $10 million. This leaves $5 million on the table, or $500 per share. Series A does not convert and demands its $1 million. Series A and B come out flat; common gets $4 million / 9,000 shares, or about $444.

Let's contemplate a $10 million exit. Series B gets its $10 million and comes out flat; everyone else gets screwed.

Let's contemplate a Theranos exit. Everyone gets screwed. Turtleneck doesn't go to jail.

TL; DR Later stages are least volatile. They get screwed last, but also see upside last. Lower rungs' returns pay for this safety.

> Can you walk me through the math. How does one arrive at 1K of Series A preferred from 9K of common stock? How is that being derived? I'm not following.

At time t=0 (probably at founding and when hiring its first few employees) Company X issued 9,000 shares of common stock. At time t=1 it decides to issue 1,000 shares in a series A offering (most likely to VCs and outside investors). They are separate events.

1000 shares x $1000/share = $1m raised for the company in the series A.

> Also what is meant by "post-money" valuation? I'm assuming there is a corresponding "pre-money"?

https://en.wikipedia.org/wiki/Pre-money_valuation

In this case, pre-money valuation of Company X = $10m - $1m = $9m

> Lastly by "coming out flat" you mean made whole again i.e recouped their initial investment? Thanks.

Yes.

>"Let's say Company X has 9,000 shares of common stock outstanding. It raises $1 million at a $10 million post-money valuation with a 1x non-participating liquidation preference. Its cap table is thus 1,000 shares of Series A preferred stock on top of 9,000 shares of common."

Can you walk me through the math. How does one arrive at 1K of Series A preferred from 9K of common stock? How is that being derived? I'm not following.

Also what is meant by "post-money" valuation? I'm assuming there is a corresponding "pre-money"?

Lastly by "coming out flat" you mean made whole again i.e recouped their initial investment?

Thanks.

And there aren't too many NYCs in the world.

Uber is far more expensive in NYC than other cities such as Chicago.

Now that Uber has allowed tipping and since there is a rating system, everyone must now tip otherwise risk getting low ratings which effectively makes NYC rides even more expensive than they had been.

Most people don't own cars in Manhattan so that they either have to take mass transit or use Uber/Lyft/Taxi. There are many elderly on fixed incomes that have trouble ambulating (moving around) where lower cost taxi service is important.

In NYC at least, Uber doesn't need all of that corporate overhead and perhaps they should spin it off into some low-overhead operation.

- There is already a rating system for drivers, and riders, so this isn't a "will be" thing, right?

- The rating system is not currently blind, as far as I know. Certainly riders can see drivers rating, and I believe drivers can see riders rating as well.

Even if they cost more I have no idea how I'm going to get around when I travel. The other options are so disgusting I'm not going back to that.

I would use them even if prices went up 30%.

There are a few costs like background checks that may go down when they stop expanding, but fewer than you might think.

Uber demand is definitely elastic.

I don't understand why that would decrease overhead?

I also don't understand why the other aspects of overhead (dispatcher, offices, taxi lots) don't add up to more than Uber (whose only cost is servers and support)

More than a team of four can manage certainly.

It does not, however, take a constant number of engineers to maintain the infrastructure that lets a single engineer push a feature to millions of users. (P/I)AAS can help, but you still need people to monitor performance, find regressions and bugs, and track them down.

Uber only works if there are enough drivers so that I can always get a ride. If I get told that I have to wait 30 minutes for a ride more than a tiny handful of times I'm uninstalling the app. Growing organically in region would be very tough since it would take too long to get enough drivers in place that users could rely on the service and conversely drivers would be unwilling to sign up because not enough users where bothering to use the service.

Those 4 can handle everything backend related too?

There is always demand for Uber/Lyft just because there isn't a taxi cab nearby 24/7. How is that difficult to understand?

There's certainly a market for app/phone jailable cabs because there isn't one within street hail range at all times, but regular cabs have been phone hailable forever (it's the only way to get one outside of a few major urban cores) and are increasingly app hailable. Aside from regulatory supply restrictions (flouting which may not be sustainable) there's no real basis for a market for alt-taxis as anything but interchangeable competitors supplying the same substitutable commodity as regular taxis.

For me, this wasn't particularly true, especially in San Francisco. The dispatch was unreliable and slow.

Other areas, sure!

Now that the idea of phone-trackable dispatch has become commonplace, regular taxi companies are starting to do this too.

With Uber you also get know the fare upfront, the time of arrival, no need to tell them your address, there is a driver reputation filter, passenger insurance and it's safer than normal cabs in many countries.

Not the same thing as a phone call.