There's an old concept from Balaji Srinivasan called the idea maze that's still useful today.

The idea maze shows all permutations of an idea – branches of the decision tree that make up a giant state machine. Open-source vs. closed-source might be one branch. Another might be usage-based or subscription pricing. Targeting developers or another persona.

For each idea, you should be able to plot the historical branches in the idea maze and what ideas have already been tried.

Sometimes one path is only accessible after another has been crossed. For example, the Google founders wanted to create a web browser early on, but CEO Eric Schmidt didn't believe the company was big enough to win the browser wars at the time (Chrome was released ten years after Google was founded).

“Having come through the bruising browser wars, I didn’t want to do that again.” – Eric Schmidt, WSJ (2009)

Circumstances can change, and a previously failed path becomes viable. For example, many dot-com era companies like Webvan and Pets.com failed, but Chewy sold for $3.35 billion in 2017, and Instacart's current valuation is $24 billion.

New paths can become apparent when existing companies reach scale. The shortcomings of the status quo might help identify a new architecture, optimization, or market to enter.

You can read Balaji's original notes from the 2012 class he taught at Stanford here.