I am certain I saw this somewhere and that I didn't come up with it, but I can't recall where so I wanted to recreate it for posterity.
You will mostly hear about the Eisenhower matrix in the context of prioritization of work, and this one is the same just in the context of digital product design.
If clarity is high, and risk is low: measure, and iterate.
If clarity is high, and risk is high: you need to make sure you are attacking from the right angle.
If clarity is low, and risk is high: do a lot of research before you make anything tangible.
If clarity is low, and risk is low: just execute iterate, as fast, as cheap, and as much as possible.
2022-01-08: Found the original post via Christopher Kuhrt.
2023-01-18: I've revised the image, and the last point of the matrix interpretation for clarity. Since this post blew up thanks to Vitaly Friedman's posts many people were left confused and asking about the low clarity-low risk situation. The point is that if your problem is in this quadrant it means it's difficult to make a decision, but it's cheap to experiment.