How do you guys currently work to predict customer loss?
Ie how do you zero in on exactly which customers are going to leave you plus how do you calculate exactly when + why the might leave?
Nik - Churnly
So in follow up, (lets use the example of a B2B SaaS company that generates revenue via a subscription model, with churn equalling a subscription cancellation) what metrics do you look at to pre-emptively anticipate a specific customer planning to churn - how do you then transform said insights into an actionable plan?
I ask because at Churnly we are developing a churn prediction tool, I would like to get an idea of what metrics subscription businesses currently value most highly to make their predictions.
This is very vague. Is this a B2B or a B2C or a B2B2C or another variation? Churn is directly related to sales cycles and costs. The longer the sales cycle, usually the slower/lower the churn (because the pain of buying and integrating/adoption makes the buyer far more deliberate in the process). Free B2C models (such as apps) many have very high churn (and with some research you can find the benchmark) but definitionally, do you define churn as a user uninstalling the app or simply not using it for a period of time? In a B2B world, losing a full installation or losing users? In an enterprise installation with many users, number of users not using could be churn even if they are still being charged for the user.
Yes vague and broad question. Manage your customers...get feedback for starters. If you have 10 questions for example and customer service ranks poorly on 90% of those who leave...you clearly have a problem with customer service. Could be quality of product..etc. But again, the question has not been thought through very well, or you are not expressing what it is in your mind...you know but no one viewing the question would know..which happens a lot.
Apologies if my wording makes the question too broad or vague.
Are there any particular metrics that you measure, that give you an indication of when a customer might leave.
For example, do you look at spending patterns, do you look at frequency of purchase or average value of spend.
How do you combine these measurements to produce a predictive analysis of who will leave your company and when?
Or are you guided by intuition on this matter?