Using Claude, I analyzed our shared inbox and email history dating back to 2010 to identify patterns in client terminations. For context, I run a digital marketing agency and clients are generally B2B small to medium sized businesses that are digital marketing clients. My goal was to identify a pattern on client terminations, what made them leave and what were some early p[atterns to identify before they left. After Claude produced said analysis - I did some manual verification on most of the data because this peaked my curiousity. When business owners were the puchaser and the point of contact - it could be easily predicted that they will go on to become worst clients. Having been a business owner for over 20 years, I’ve always been struck by a recurring trend: whenever a customer throws a tantrum, threatens legal recourse, or takes things to an extreme, they are almost invariably another business owner.…