Churn is a fact of life for a subscription company - like death and taxes. :)

Yet, the biggest surprise as I started studying churn and publishing findings on churn is the mechanical nature of a large portion of your churn. Right now for most subscription businesses roughly 40% of your churn could be solved through mechanical means. Put another way - 60% of your churn requires improvement in product, customer targeting, etc, but the other 40% could be cleaned up by getting your credit card failures in order, optimizing for the proper term for your customers, re-engagement campaigns, etc.

In product, we love to romanticize churn as the eternal pursuit of perfection that will keep the value driving for the customer and therefor keep them around. While there's some truth there, this view clouds growth managers from properly working on churn in most organizations. In reality, those mechanical bits matter. They'll never be the end all be all to your churn (and that 40% obviously fluctuates), but there's a lot more to do that is shorter term/medium impact when it comes to churn and retention.