When does my corner shop become statistically relevent?

[WillQ]

Hey all,

Here's a nice Wednesday afternoon brain test for you all (I've been pulling my hair out with this one, because I'm sure it's probably straightforward)

If I've got a shop, let's say, and I decide I want to work out how awesome my shop is at getting customers.

I want to work out my awesomeness fraction (let's call it "walk through rate").

I measure:

X, the amount of people that walk past in a given amount of time.

Y, the amount of people that walk past and decide to enter the shop in that given amount of time.

At what point can I say, with a fair level of accuracy, that "Z% of people who walk past my shop come inside"? Obviously I need to wait for a certain number of people to come in or walk past, but I can't work out how to calculate this confidence level... any help would be massively appreciated!

(I was thinking it'd have something to do with calculating the standard deviation, and measuring that as a percentage of the total... but I'm not sure...)

 

[Stephen Tashi]

I think you are looking for a "confidence interval" for sampling from a binomial distribution.
See http://en.wikipedia.org/wiki/Binomial_proportion_confidence_interval