> Uber seem, from the financial data I have seen, to be sticking to that playbook down to the 3rd decimal place

Except, if you check the linked article, you'll find that bookings grew 11%, revenue grew 21%, and losses grew 38%.

And YoY (backing out the Q4 numbers to Q3 numbers), bookings grew 80%, revenue grew 20%, and losses grew 56%. Assuming that "costs" are revenue+loss, costs grew 34%, still faster than revenue.

>The (buzzword warning) hyper-growth startup model is in essence 2X revenue, 1.5X costs. Repeat until inevitably profitable - and I say inevitably because with enough time, 2X revenue, 1.5X costs gets to profitable. The only variable there is having a long enough runway.

it is like gambling with doubling (or even tripling) down each time - with enough time you will inevitably win and as result will reap a huge profit in total. Now if just somebody write me a blank check backed by an unlimited bank account... On practice though, even if got that check - what if the casino can't match my next 3x bet?

The same thing with Uber - the important thing isn't just the Uber's money runway. The other side's "runway" (i.e. total size of the market) is also extremely important - are there enough people on Earth needing enough rides so that that amount of rides would allow for that "2x revenue 1.5x costs" model to reach profitability?

  Now if just somebody write me a blank check backed by an unlimited bank account... On practice though, even if got that check - what if the casino can't match my next 3x bet?

It will never come that far. Even if you truely have no limits (and with doubling and trippling you arrive at mighty high amouts, even if you start at $1) casinos have themselves covered.

It's called a table limit. At one point you just can't raise your bet anymore.

That is exactly his point actually. It's an example of why the model breaks down in reality despite the math working in theory. That's why it's called gambling whether it's a casino or a startup.

Except gambling has the odds stacked in the house's favour. There is no such problem with a startup.

Will revenue double forever? No. Will revenue double for 100 years? No. Will revenue double for 10 years? Hmmm. Will it double for a year? Lets just say yes. Solve for between 1 and 10, and place your bets accordingly.

>"The (buzzword warning) hyper-growth startup model is in essence 2X revenue, 1.5X costs."

Can you elaborate on this? What is the relationship between those figures? I am guessing there is a third component that those are multiples of?

Growth rate... which he implies is large.