Ask HN: How does your research lab prioritize projects?

I've worked in industry and have a decent grasp of how projects get prioritized for the benefit of the business. Balancing short-term cash flow with long-term vision and whatnot. In a lot of cases I've seen, Product will decide which projects have highest impact on the business.

I'm curious how project prioritization work in research labs. From what I've heard, it's usually the case that the Principle Investigator simply has too many ideas and too few grad students/technicians to work on everything.

What factors do you consider when deciding to commit resources towards a research project? Is there an "agile" way to do research?