alt Hacker NewsWell I did say "good", which excludes things that don't fit in the context of the wider business.
I think startups have an easier time for 2 reasons: They are innovating on the product side already, and probably were already capacity constrained for work on the top ideas. They also have a high proportion of tech debt which never bubbled to the top of effort vs reward (competing against product work) before but would still be valuable to fix. E.g. https://posthog.com/blog/sql-parser was on the HN frontpage a few months ago, and this work would never have happened in the before times
Even if the idea fits the context of the wider business, there's such a thing as too much change "just because you can", which alienates users and can make the end product an unfocused mess.
If it's easier for startups, it's for the same reason doing a "pivot" is easier: they usually don't have much of a product or revenue or customer base to begin with, they throw things out there and see what sticks.