Life, research, and an academic career have one thing in common: they cannot be planned and always turn out di erently than expected. At the same time, it does not mean that they have to be completely random. Using my past research as an example, I will outline how the wrong or right idea paired with random events lead to systems like CrowdDB, Tupleware, and now IDEA/Vizdom, and what I would do di erently in retrospect. Furthermore, I will tell you my most important lessons learned over the course of now 10 years in the research community.