We report on the design of a study to measure how the introduction of middleware software components could impact confidence in an institution. Middleware is software that runs in the background and thus cannot be perceived directly by end-users of software products. We discuss the concept of confidence as a type of trust, outline existing techniques how to measure it, and discuss the applicability of known techniques on measuring changes in trust caused by software that cannot be perceived by its users. We conclude that the methods proposed are relevant to domains such as e-Health, e-Government, and other domains from the public sector that aim to increase trust by citizens, patients, and other forms of customer roles.