Constraints are a Good Thing

Creativity is a process of making mental connections between disparate ideas.  It's often born out of errors, the creator unintentionally connecting two things that a more objective viewer would rule out as unrelatable.

Being a network process, as opposed to a procedural one, the difficulty of the problem increases as the number of degrees of freedom increase.  The classic example of a network process is the "Traveling Salesman Problem".

This is why art teachers don't assign their students problems like:  "fill this blank page with something beautiful".  Instead, they give them a set of constraints to work within, they dictate a theme, an artistic style, and a time limit.

In another sense, constraints give a creative endeavor a finite set of tractable starting points.  They imply a form for the eventual solution and provide strong foundation points that can be used to underpin a more fanciful idea.  

Twitter's 160 character limit is an example of this.  Initially a mere technical constraint of SMS, it now serves as the boundary condition for a whole new mode of communication, that, without this constraint, might have reverted to yet another blogging service.  

When confronted with an intractable product decision, I first attempt to layout the forced constraints:  

  • How much product investment can we afford?  
  • What kind of timeline should we deliver it in?
  • What is the technology we're using especially good/bad at?
  • What is the user's problem that we're solving?

If I'm still stuck, often because the true constraints are still too broad, then I try to apply artificial constraints that force me to think about different aspects of the problem:

  • What could we do in a single day?
  • What if there was no UI at all?
  • What if our customer's were end-users instead of enterprises?
  • What if we could only charge half as much for the service?

I often discover that such artificial constraints help clarify what the real proposition is, assist in searching for ideal solutions instead of merely adequate ones, and reveal orthogonal ways around obstacles.