Quote:
Originally Posted by darksoul
|
If you had any grasp of statistics you'd know that there is absolutely no need to use Poisson Distribution when calculating the probability of someone joining your site based on previous factors.
Binomial distribution is a hell of a lot quicker and much more appropriate.
Let's have a rule of thumb that traffic converts at 1:500. Therefore, every time someone hits your site there is a 0.002 of that person becoming a member.
Probability of success on trial: 0.002
Num trial: 2000
Number of successes: 1
Probably that X is 1 \< = 98%
Now in the 4000 hits I've bought from Traffic Holder (I've bought some before) I had gained a total of 2 trial joins. That's a free trial too btw.
That's a ratio of 1:2000 on trials. Or $8 per trial account. The maths doesn't come in to this; I could buy 100k traffic, check the results and create a normal distribution and calculate the probability of a person signing up but it's pretty clear that the traffic from Traffic Holder is shit. If I can only get 1 trial in every 2000 hits, what are the chances of this actually being profitable for webmasters?
TrafficHolder; either I've had a Z score of 2.9* in respect to chance or you give poor traffic to those who do not buy a lot of clicks. I'd really love to talk to any of the webmasters who buy bulk traffic from you - I know for a fact that the traffic I buy from you isn't even close to being on a 90% loss margin.