Many organizations that have moved to more agile ways of working continue to use heavyweight testing tools. These tools have problems that make them unsuitable for fast moving software delivery. Large complex tools have high learning curves and require specialist skills and training, making it hard for the team themselves to test. Often this results in an unnecessary overhead for every release as other teams get involved. Expensive and limiting software licenses makes this problem even worse. Some heavyweight tools use a 'model driven' approach where an attempt is made to accurately model the usage patterns of the application, which leads to costly test script maintenance and development time being lost to 'false positives'. We have seen few situations where simple open source solutions cannot give the required level of confidence for much less time, effort and money.