Fred has been posting a wonderful series of MBA Mondays posts that I recommend to everyone who wants to know more about the business side of startups.  If you come from the engineering side, it may also help to think about business using engineering principles.  My favorite one is the idea of a “single binding constraint." 

When the execution of a program runs into a hardware constraint, it tends to be a single binding constraint: the program is either CPU-bound, memory-bound or IO-bound but generally not two or three at the same time.  The same tends to be true for organizations.  At any particular point in the life of a startup, there is most likely a single binding constraint on growth.  For most startups this is the lack of product-market fit.  If you haven’t solved that, investing heavily in a sales force is the equivalent of adding RAM when your program is CPU-bound.

Once you have cracked one binding constraint a new one is likely to emerge.  For instance, with product-market fit in place, your single binding constraint might shift to sales, or to engineering or to customer support.  A key job of the founder/CEO/board is to identify the single binding constraint for the startup at any given time and focus on overcoming it.   I have found this way of thinking to be incredibly powerful.  There tends to be so much going on in a startup that it is all too easy to want to attack many different constraints all at once.  At that point it is critical to remind oneself that only one of those constraints is likely to be binding, in the sense that it is truly impeding growth.