Mob Programming, or "mobbing", is a relatively new collaborative programming practice being experimented with in different organizational contexts. There are a number of claimed benefits to this way of working, but it is not clear if these are realized in practice and under what circumstances. This paper describes the experience of one team's experiences experimenting with Mob Programming over an 18-month period. The context is programming in a software product organization in the Financial Services sector. The paper details the benefits and challenges observed as well as lessons learned from these experiences. It also reports some early work on understanding others' experiences and perceptions of mobbing through a preliminary international survey of 82 practitioners of Mob Programming. The findings from the case and the survey generally align well, as well as suggesting several fruitful areas for further research into Mob Programming. Practitioners should find this useful to extract learnings to inform their own mobbing experiments and its potential impact on collaborative software development.