I've heard that it's a good rule of thumb to not develop something that is just an extra feature of an existing behemoth (e.g. Atlassian plugins etc., but also the thing you built). If you're failure, then you've wasted time and money and if you're a success, the behemoth will just reimplement your work as a part of their platform.

Pretty much the only way to win it so make something that provides marginal revenue, which makes reimplementation not worth it for the giant (but potentially worth it for you). I suspect a lot of Unity and Unreal plugins and authoring tools reside in that niche.