Ask HN: How do you experiment to determine the right model w/o alienating users?
We learn a ton by making tweaks and seeing how our potential customers react (pricing, how products are bundled, how trials work, terms of the subscription, etc). At some point though this experimentation process has diminishing returns in the form of alienating users (they get confused or pissed off and we're perceived as wishywashy).
So my question is how do other startups juggle the tradeoff of wanting to rapidly iterate and get to the optimal "configuration of dials" vs. keeping consistency of the offering to earn the trust of users? TIA sean