Raising money is never an accomplishment in and of itself—it just raises the stakes for all the hard work you would have had to do anyway: actually building your business. Tell everyone inside the company, over and over, until they can't stand it anymore, and then tell them some more.

Before Product/Market Fit, raise at least enough to get to Product/Market Fit. After Product/Market Fit, raise at least enough to fully exploit the opportunity and get to profitability while doing so. In both cases, add a substantial buffer beyond your default plan—insurance against bad surprises.

The big difference between startups from the late 90s that are doing well today versus the ones that no longer exist is that the former raised a ton of money when they could, and the latter did not. Opsware would have been bankrupt if it hadn't done the same.

Suppose you raise a lot and do well—you'll be happy and rich, even if slightly less rich than if you'd rolled the dice on a smaller round. Suppose you don't raise enough and it backfires—you lose your company. Is the asymmetry really worth it?

The biggest downside of too much money isn't dilution or liquidation preference—it's cultural corrosion: complacency, laziness, arrogance, hiring bloat, lazy management, sales teams pushing an unfinished product, and schedule slippage because there's no urgency.

Act like you haven't raised nearly as much money as you actually have—in how you talk, act, and spend. Stay scrappy on office space and furniture (Ikea all the way); the only things worth splurging on are big monitors and ergonomic chairs.

The easiest way to lose control of spending after a big raise is hiring too many people; the second easiest is paying people too much. Worry far more about the first—headcount multiplies burn much faster than raises do.

Hiring bloat also triggers the Mythical Man Month effect and drives your best engineers to quit.

On liquidation preference: the more you raise, the higher the acquisition price has to be before founders and employees see any money on top of the investor payout. Monitor it, but don't obsess—unless you're explicitly building to flip.