One of the most common questions I have been asked is, "What sample size do I need for ... ."
The "for ..." can mean "confidence limits of a certain width,"" or 99% reliability with 95% confidence,"
or "to run a designed experiment with 5 factors and a full quadratic model."