How do I know if I've surveyed enough people for a market research study?

I did a survey among 700 people in Tianjin and Beijing and was wondering how do I establish if this is enough to represent the population.

How do I know if I've surveyed enough people for a market research study?
It really depends on how the research was collected. By this I mean how did you go about choosing which people you surveyed? The best way is through random sampling (that is the way using a list of randomly listed numbers). I know it is much more tedious and time consuming but it is really the only way to make sure you have a fair representation of the population.
Reply:It really depends on how you're segmenting your data. If you want a respresentative sample of two segments, that number is smaller than a representative sample of four segments. Also, you need to choose your level of confidence and accuracy -- the standard rule of thumb is 95% confidence and this is based on the Normal Curve of results (also known as the Bell Curve). There is a great, FREE software tool you can get from DecisionAnalyst to input your values to determine the best sample size. I use it all of the time.
Reply:1,000 people are the norm for


a survey after it has been tried and


tested and fine tuned.





[u can start with 50 and then 250 as you TWEAK the survey]