Software products are increasingly used in critical infrastructures, and verifying the security of these products has become a necessary part of every software development project. Effective and practical methods and processes are needed by software vendors and infrastructure operators to meet the existing extensive demand for security. This article describes a lightweight security risk assessment method that flags security issues as early as possible in the software project, namely during requirements analysis. The method requires minimal training effort, adds low overhead, and makes it possible to show immediate results to affected stakeholders. We present a longitudinal case study of how a large enterprise developing complex telecom products adopted this method all the way from pilot studies to full-scale regular use. Lessons learned from the case study provide knowledge about the impact that upskilling and training of requirements engineers have on reducing the risk of malfunctions or security vulnerabilities in situations where it is not possible to have security experts go through all requirements. The case study highlights the challenges of process changes in large organizations as well as the pros and cons of having centralized, distributed, or semi-distributed workforce for security assurance in requirements engineering.