I definitely recommend the slow and measured approach. Started working on my startup as a side project while still employed at BigCo. Launched, waited until $app_income >= $salary_income, then quit to do it full time. I'd argue the only reason to rush and launch something is if you have or are looking to get funding.