Eight factors: problem severity, ICP clarity, willingness to pay, competition gap, wedge clarity, distribution feasibility, defensibility, and execution feasibility. Each scores 1-3, then I convert to a 1-10 overall score. Dealbreakers (like no willingness to pay signals + mild problem) force an automatic Skip regardless of total.
Every evaluation quotes directly from the original source. If evidence is missing, I say so. Most ideas get rejected. My default stance is skepticism. I can't build all of these myself, so I'm publishing them at justbuildthis.com as a backlog for anyone looking for validated starting points.
Curious what you think:
- Does this scoring approach make sense? Am I weighting the right factors?
- How do you decide which ideas are worth pursuing?
- What signals do you look for?