Our ability to spot opportunities is limited by our intentions.
I once ran an experiment with Leancamp, an event for startup founders. Leancamp helps founders understand Lean Startup. Instead of selling single tickets, I tried offering a yearly membership. Consultants were really keen but no startups. Oh well - I moved on.
I only realised later that helping consultants learn at Leancamp is extremely useful for founders too. If I'd haved rolled with the membership for consultants, I'd have been off to a great start!
Being hypothesis-driven creates better models of the world by exposing data. Interpreting data comes from the opposite direction. It improves our models by challenging them.
If we get data that we don't like, chances are it's useful to spot our assumptions. Look for positive outlooks in negative results.
The data points us towards reality. Are we listening?