Computer users spend the vast majority of their time sitting, and poor posture in the workplace is an endemic issue. This work presents PostureChair, a posture detection system that uses contextual digital feedback to persuade users to improve their sitting posture. Two types of digital feedback, with varying amounts of information, were compared through a repeated measures study to determine how much information is necessary to improve posture and to appeal to the user. The results of the study showed participants' sitting posture improved significantly with both feedback types when compared to their posture with feedback disabled. Participants overwhelmingly preferred the more detailed feedback even though it did not clearly improve users' sitting posture beyond the simpler feedback. The PostureChair system was well received and demonstrates that contextual posture improvement is an effective and much-needed addition to the workplace.