retention

Retention and churn describe one event from opposite sides, and the arithmetic is trivial: what did not leave, stayed. The difference that matters is what each one is good for. Churn is a loss you report after the fact. Retention names a population you still have and can still write to, which makes it the only one of the two you can act on this week. Everything below is about that action, not about the formula.

Which month decides retention?

The first renewal. In almost any subscription, the largest single drop sits between the first billing cycle and the second, and a fan who has renewed once behaves like a different animal from one who never has. He has proved the thing you were hoping for: that the charge is worth repeating.

The practical consequence is unglamorous. The first weeks of a subscription deserve attention the fourth month does not, and they are the easiest attention to lose: new arrivals wait while the team services whoever is talking loudest today.

What actually lifts it?

Being present, being quick, and not selling too early. Five things move it, and none of them is a new offer:

• Being written to at all. Subscriptions rarely end in an incident. They end in silence: a few weeks with no message, an auto-renew quietly switched off. No complaint, no chance to fix it.
• Answering quickly. A fan who writes in the evening and gets a reply the following afternoon has already half-left, and the evening is when the inbox fills.
• Not selling from the first message. Our data shows that pitching before the sixth message drops conversion by roughly a third, with the optimum after about ten exchanges. A new subscriber whose first three messages are offers has learned what the account is, and he learned it in week one.
• Continuity across writers. A handover that loses what the fan said last week reads as being forgotten. A voice guide exists so that the change of shift is invisible.
• Remembering. A personal callback (something he told you weeks ago) is the strongest thread-reopener there is, with one measured exception noted below.

What does not lift it?

Four habits that feel like retention work and are not:

Habit	Why it fails
A discount offered to someone leaving	Teaches the whole list that leaving is how you get a better price
More mass messages	Volume tells a fan he is a line in a database, not a person
Longer sales sequences	Retention is decided by the conversation, not by the number of offers in it
A callback used as the pitch	Our data’s most counter-intuitive result: the personal detail costs points when it lands at the exact moment of the offer

That last row is worth sitting with. The thing that keeps a fan is the same thing that damages a sale if you spend it at the wrong second. Use it to reopen, not to close.

How do you measure it without fooling yourself?

By arrival month, not in aggregate. An overall retention figure moves whenever recruitment moves, so a good month of new subscribers drags it down while every underlying group improves. Cohort analysis is the only reading that separates the two.

And decide what you are retaining. Retained subscribers and retained buyers are different populations, and only the second one shows up in the money.