A common practice in statistics is to use a 95% confidence level, meaning you can be 95% certain that the true parameter lies within the calculated interval. Some marketers may opt for a 90% or 99% confidence level depending on the stakes and the required precision of the decision.