An email list that is not growing is shrinking, because the underlying addresses degrade continuously. The rate varies by audience, and the mechanism is the same everywhere.
Addresses stop working without notice
Business addresses are lost whenever someone changes employer, and personal ones are abandoned in favour of newer accounts without being closed.
Neither event generates a notification. The message is delivered, or accepted and never seen, and the list appears intact while its reachable portion falls.
This is why a list that has not been mailed for a long period performs poorly when it is used again, sometimes badly enough to affect the sender's standing with mailbox providers.
Engagement decays faster than addresses do
Subscribers who were interested at the moment they signed up become less so as their circumstances change, and most never unsubscribe.
They stop opening instead, which is invisible in list size and highly visible in results.
The practical consequence is that list count is a poor measure of anything, and the count of people who have engaged recently is the number that predicts performance.
Mailbox providers respond to how recipients behave
Filtering decisions are influenced by whether recipients open, reply, delete without reading or mark messages as unwanted.
Sending to a large inactive segment therefore harms delivery to the active one, because the aggregate signal is what providers evaluate.
This inverts the intuitive approach, since mailing everybody to maximise reach can reduce the number of people who actually receive the message.
Frequency changes the decay rate in both directions
Mailing too rarely allows subscribers to forget why they signed up, and an unexpected message from a forgotten sender is likely to be reported rather than read.
Mailing too often produces fatigue and unsubscribes, though an unsubscribe is a cleaner outcome than a complaint.
The sustainable rate depends on whether each message carries something the recipient wanted, which is a content question rather than a scheduling one.
Acquisition quality determines the starting point
Addresses collected in exchange for an unrelated incentive decay fastest, because the interest was in the incentive rather than the sender.
Subscribers who joined for the content itself remain engaged considerably longer, so the same list size can represent very different assets.
Marketers who understand their decay rate can calculate how much new subscription is required simply to stand still, which is usually more than expected.