good vs excellent (with help from AI)
A useful question at the outset of a venture, whether you’re starting a whole company, shipping an early draft of an artifact, or baking a cake for your nephew’s birthday is: what is the difference between a good version and a truly excellent version of this thing?
I like using this as a prompt for Claude or GPT. It’s also a good question to ask customers, collaborators, and mentors or experts in your field. A follow-up prompt for your AI is: “Thanks! Please give me a step by step plan [or recipe, if you are in fact baking a cake] for making the excellent version you just described.”
As you gather information in responses to this prompt, remember that you don’t have to do the excellent thing in every instance. Sometimes that excellence is more costly than it’s worth in the early going. Sometimes you can build an audience or secure a deal that lets you move forward and build more without doing that excellent thing at all. Your cake might not need icing piped in perfect calligraphy across the top. Your podcast might not need an original score that enhances each emotional beat of the pilot episode.
This good-vs-excellent question can also be useful when you’re delegating. The kind of people you want to hire may be predisposed to doing the excellent version of things by default. You may need to help them develop good judgment about when to ship something shitty, or shabby, or shiny. Specifying the difference between those levels of performance can help you and your delegate select a clear and shared definition of success.
-eric
