Statistical Significance and Advertising

#1
Hi everyone. I'm new here and have only a surface-level knowledge of stats. I know what I want to know, but not how to figure it out. The question is centered around online pay-per-click advertising, where you can track whether a paid advertisement (click) converts into a sale (and its $ amount). Thanks in advance to anyone who takes a shot at this.

I want to know at what Dollar amount of ad spend without a successful conversion (sale) I should cease advertising or change my bid rate. For example, I have the follow data:

A click costs $0.70
A sale is worth $30 of profit, not revenue

At what spending level would I be 95% confident that I am paying more than I will ever get returns on? Is it something like $45, $60, or $90?

I would like to know the mathematical reasoning behind this because I'm interested in it and would like to be able to apply it to many situations.

Further, relevant questions would be "How sensitive is that value to changes in click cost or profit value of a sale?", but I'm initially more interested in how to find that value at which I should stop or change.

Thanks!
Brad
 
Last edited:
#3
I'm by no means an expert, but as I'm looking at your question, I'm thinking that where you say "A sale is worth $30 of profit, not revenue", how do you come up with that? Doesn't your profit ultimately depend on the number of clicks you've paid for up until you make that sale?

It seems to me that profit can only be determined after the fact.