The amount of people that you lose early in your relationship with them or early in their account-having status is way worse than you probably think it is.
High level metrics like net churn tell you how many people you lost from the number you gained during a period of time, say a month. It's expressed either in terms of head count, like the number of users, or in terms of revenue — how much subscription money just walked out the door.The idea is that this net churn number should be somewhere in the 2% to 3% range. If you're down to 5%, that's not a great outlook, you would want to run your business differently based off of that. On the other hand, if you can get to 1% or the fabled fantasy lands of net negative churn where your user base is actually growing over time, that's the ideal.The problem is that net churn doesn't factor in when people churn. People churn at very different times in their relationship with you. If you're looking at net churn for the month of August — how many people did we have in August and how many did we keep into September — and you're including your entire user base from people who just walked in the door five minutes ago to people who have been customers of yours for nine years, the likelihood of them churning are very, very different.The net churn number masks that complexity.It gives you the overall performance of how big of a pile of users you have at any given moment, rather than giving you insight into the critical early stages, where you are just haemorrhaging the most amount of people.And it's true for all aggregate metrics, across different levels of scale, because that's just how the unit economics play out.Look at MRR, for example: Look at how many people you keep from month 1 to month 2, it's a bigger drop-off from month 1 to month 2, than month 2 to month 3. And then it's an even smaller drop-off from month 3 to month 4, and then it's even smaller from month 4 to month 5.