Paradox of startups: Your startup may solve the most insanely complex and fancy use cases, but most users will use it for such dumb shit that you will say:
"really, you're paying me for that?"
@SplitSoftware did fancy experimentation, but most users used it for on/off feature flagging
@TectonAI could solve online / offline skew, but most users got excited about online serving - just simple serving out of DynamoDB.
@Gong_io can do some fancy stuff in its product, but most people use it for call summary & 2x speed replay.
So why build or market for complexity?
Because startup sales is about the vision of what's possible.
Build the vision, but make sure you know what your "dumb shit" use case is and make it a delight to use your product that way