An experiment can continue indefinitely if nobody defines what would end it. Before starting, decide what you want to learn, what you can afford to invest and which result would justify continuing.
For a fictional newsletter test, the useful question might be whether readers request help with a particular reporting problem. Set a limited run and record replies and relevant enquiries. A rising total of messages sent does not answer that question.
Protect the decision
Choose a spending or time boundary that fits your real capacity. Keep a review date. If evidence is too limited for a conclusion, record that uncertainty instead of declaring the idea successful or worthless.
An early stop may be appropriate when the attempt causes harm, breaks an important commitment or reveals that the premise was mistaken. You do not owe a flawed experiment its full planned duration.
Keep the learning
When an attempt ends, preserve what happened and why you stopped. The next version should inherit that knowledge. Renaming the same activity and starting again without a change wastes the result.
Stop rules make persistence more deliberate. They let you distinguish continuing because the evidence supports another attempt from continuing because abandoning the original story feels uncomfortable. Both effort and courage are better used when the decision remains visible.