Testing whether my own emails keep students longer

I've got 14 students on monthly lessons and I lose roughly one a term. I think it's because they only hear from me when something's wrong, so I'm going to send a short personal note every second week — a piece of music they'd like, a question about what they're practising — and see if the ones who get it stay longer. Too early to know anything, but I'd rather change one thing at a time than sign up for another platform.