The app lived or died on two conversions: a user finishing a resume, and a user paying for the subscription. Both happen on a phone screen, usually in a hurry. The product needed flows that didn't leak users, and a subscription page that could be optimized with evidence instead of opinions.