We argue that the first undergraduate courses are the most important ones on the student's path towards becoming a computer scientist. Therefore, during 2010-2012, we have exercised extensive effort in order to improve the first-semester Computer Science (CS) courses. We have been able to use a learning-by-doing approach called the Extreme Apprenticeship (XA) method accompanied by personal advising even for courses with hundreds of students. We claim that when high demands are met with sufficient support, students learn valuable programming skills that become a foundation that carries them in their further CS courses. In this paper, we analyze how the effects of a three-year effort of renovating our introductory programming courses propagate to further studies. Compared to the control cohorts of 2007-2009, we observe a carry-on-effect caused by the XA method in student success that is visible in the per-student average accumulation of credits after 7 and 13 months after the start of studies. In addition, we can see the effect propagating to mandatory subsequent courses, even without the XA method.