Hi, I apologize if this question is inappropriate to this group, but I have not been able to find a better option. I am presently surveying a market that has 570 firms to determine which providers they use for particular services. The goal is to construct a breakdown of the marketshare that each service provider holds within the 570 firms. So far we have spoken to about 60% of the firms and have achieved results to the effect of: Brand A: 30% Brand B: 25% Brand C: 20% All others: 25%
My question is: How many firms do we have to survey before we can say that our results represent the service usage of the entire market of 570 firms within a particular range of error? And how do I determine what my margin of error is at a given sample size?
Any help would be greatly appreciated! Please e-mail me at firstname.lastname@example.org