Using online calculators or the formula behind these calculators where each cells has say 1000 names, a standard deviation of 0.5, calculated at 95% confidence and 5% margin of error then my sample size appears to be at least 278.

**However, is that 278 mailed or 278 responses.**

My estimated response rate is only 0.5% at best which would only provide me with 5 responses.

If I need 278 responses I assume I'll need to sacrifice results at recency level to increase volume size ensuring my smallest cell is at least 55,600. Is that correct?

If so the risk is way too high for that kind of volume.

I'm hoping you confirm or correct my thinking and provide any advise please? Thank you.