Target employees 410k people according to Wikipedia. A sample size of 384 could represent a larger 500k population with 95% confidence and a 5% margin of error. I don't think the sample size is all that bad in the realm of surveys. How the sample was chosen is the critical piece here.