Have a strong vision for your product and go out of your way to do things that help materialize said vision.
Having the courage to subject your grand vision to constant testing and feedback is vital for your startup's growth.
Architect systems in the most simplest manner possible and optimize mainly for flexibility, not performance, because you'll almost certainly not know which apps will make it succeed.
While it can be useful to ask users for feature requests, they themselves do not always know what they want or what is possible. It is your responsibility to try and understand what users want even before they know they want it.
More features does not necessarily equal a better product. You don't need to be good when you're great.
Altering incentive structures to see the behavior you expect might not always be the best way to improve a system. The sytem itself might need to be overhalled. Mechanism design plays an important role in protocol architecture.
Company culture always trickles down from the leaderhip.
It's easy for teams to get lost in the day-to-day aspects of their work. A good leader consistently reminds her/his team of the overarching vision - the reason as to why we do what we do.
Hire only when you need additional resources, not simply because you have the (financial) capacity to do so. Do not hire in expectation of future growth either. Current growth rate by itself is not indicative of future growth.
Know your audience. Calibrate content based on the audience's knowledge of the subject matter at hand.
Don't say something unless it's useful for the reader. Eliminate redundancy.
Simplicity and elegance go a long way.
This might vary from person to person, but the more creative you want to get, the more it might help to procrastinate.
In an age of fleeting attention spans and mindless optimizations for short-term feedback loops, the one's that dare to have a grand vision to change the world, and are ruthlessly focused on executing over long time horizons, will win at scale.